project.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "a64cbc05-9988-4261-93c5-4af7ce59aee1",
      "metadata": {
        "id": "a64cbc05-9988-4261-93c5-4af7ce59aee1",
        "scrolled": false
      },
      "outputs": [],
      "source": [
        "# requirements\n",
        "!pip install -q rdflib sparqlwrapper owlrl gdown pandas wikipedia-api\n",
        "!pip install --upgrade wikipedia-api\n",
        "!apt install graphviz libgraphviz-dev > /dev/null\n",
        "!pip install --quiet rdflib networkx pygraphviz\n",
        "!mkdir -p images\n",
        "!pip freeze > requirements.txt"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "a247a8de",
      "metadata": {
        "id": "a247a8de"
      },
      "outputs": [],
      "source": [
        "conda install --channel conda-forge pygraphviz"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "2fa54075-b30d-4e94-9d41-5111cb90aaa3",
      "metadata": {
        "id": "2fa54075-b30d-4e94-9d41-5111cb90aaa3"
      },
      "outputs": [],
      "source": [
        "import gdown\n",
        "import os\n",
        "import pandas as pd\n",
        "from rdflib import Graph, URIRef, BNode, Literal, FOAF, RDF, RDFS, OWL, XSD, Namespace\n",
        "from SPARQLWrapper import SPARQLWrapper\n",
        "\n",
        "from datetime import datetime\n",
        "import numpy as np\n",
        "import math\n",
        "import wikipediaapi\n",
        "import owlrl\n",
        "from rdflib.extras.external_graph_libs import rdflib_to_networkx_multidigraph, rdflib_to_networkx_digraph\n",
        "import networkx as nx\n",
        "import matplotlib.pyplot as plt\n",
        "from IPython.display import Image"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "c86133b5-6fe3-4223-90e8-7349d8d4f6c8",
      "metadata": {
        "id": "c86133b5-6fe3-4223-90e8-7349d8d4f6c8"
      },
      "outputs": [],
      "source": [
        "# create directory for data\n",
        "if not os.path.exists(\"./data/\"):\n",
        "    os.mkdir(\"./data/\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "21a73c00-3fbe-4a50-9f5b-f8d6e077bb67",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 36
        },
        "id": "21a73c00-3fbe-4a50-9f5b-f8d6e077bb67",
        "outputId": "07243e9f-aad7-44ec-ecf4-e210afcb04fe"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "'./data/albumlist.csv'"
            ]
          },
          "execution_count": 5,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Download data\n",
        "gdown.download(id=\"1UILSMenvqzFMsIhV6l4ZV1opJpRQZnwC\", output=\"./data/rym_top_5000_all_time.csv\", quiet=True)\n",
        "gdown.download(id=\"1UNJ2iDX-xoWIekSA0ZwoqNOk3eAZjqrp\", output=\"./data/albumlist.csv\", quiet=True)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "yb3gB6VfgLRT",
      "metadata": {
        "id": "yb3gB6VfgLRT"
      },
      "outputs": [],
      "source": [
        "# Move the checkpoint graphs to the folder 'data'\n",
        "! mv /content/graph_01.ttl /content/data\n",
        "! mv /content/graph_02.ttl /content/data\n",
        "! mv /content/graph_03.ttl /content/data\n",
        "! mv /content/graph_04.ttl /content/data\n",
        "! mv /content/graph_05.ttl /content/data\n",
        "! mv /content/graph_06.ttl /content/data\n",
        "! mv /content/graph_07.ttl /content/data\n",
        "! mv /content/graph_08.ttl /content/data\n",
        "! mv /content/graph_09.ttl /content/data\n",
        "! mv /content/graph_10.ttl /content/data"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "0715d141-bac3-4529-946e-4776452e88fe",
      "metadata": {
        "id": "0715d141-bac3-4529-946e-4776452e88fe"
      },
      "source": [
        "# **KRR Module 2 2022/23 Project**\n",
        "\n",
        "The project consists in the creation of an RDF knowledge graph starting from tabular data (two CSV files)."
      ]
    },
    {
      "cell_type": "markdown",
      "id": "64fdd5ae-67d2-4ce3-88af-6373858335e4",
      "metadata": {
        "id": "64fdd5ae-67d2-4ce3-88af-6373858335e4"
      },
      "source": [
        "## **`1. Create an RDFLib Graph from data contained in CSV files`**"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "mnWNAyeFgS58",
      "metadata": {
        "id": "mnWNAyeFgS58"
      },
      "source": [
        "### 1.1 ) Load the CSV files in two pandas DataFrames and join them"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "530f8ed4-604e-4489-93c5-425372ae6745",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "530f8ed4-604e-4489-93c5-425372ae6745",
        "outputId": "634d338e-d4a6-41af-fbeb-c187df0acc6a"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Number</th>\n",
              "      <th>Year</th>\n",
              "      <th>Album</th>\n",
              "      <th>Artist</th>\n",
              "      <th>Genre</th>\n",
              "      <th>Subgenre</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>1967</td>\n",
              "      <td>Sgt. Pepper's Lonely Hearts Club Band</td>\n",
              "      <td>The Beatles</td>\n",
              "      <td>Rock</td>\n",
              "      <td>Rock &amp; Roll, Psychedelic Rock</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2</td>\n",
              "      <td>1966</td>\n",
              "      <td>Pet Sounds</td>\n",
              "      <td>The Beach Boys</td>\n",
              "      <td>Rock</td>\n",
              "      <td>Pop Rock, Psychedelic Rock</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3</td>\n",
              "      <td>1966</td>\n",
              "      <td>Revolver</td>\n",
              "      <td>The Beatles</td>\n",
              "      <td>Rock</td>\n",
              "      <td>Psychedelic Rock, Pop Rock</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4</td>\n",
              "      <td>1965</td>\n",
              "      <td>Highway 61 Revisited</td>\n",
              "      <td>Bob Dylan</td>\n",
              "      <td>Rock</td>\n",
              "      <td>Folk Rock, Blues Rock</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5</td>\n",
              "      <td>1965</td>\n",
              "      <td>Rubber Soul</td>\n",
              "      <td>The Beatles</td>\n",
              "      <td>Rock, Pop</td>\n",
              "      <td>Pop Rock</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Number  Year                                  Album          Artist  \\\n",
              "0       1  1967  Sgt. Pepper's Lonely Hearts Club Band     The Beatles   \n",
              "1       2  1966                             Pet Sounds  The Beach Boys   \n",
              "2       3  1966                               Revolver     The Beatles   \n",
              "3       4  1965                   Highway 61 Revisited       Bob Dylan   \n",
              "4       5  1965                            Rubber Soul     The Beatles   \n",
              "\n",
              "       Genre                       Subgenre  \n",
              "0       Rock  Rock & Roll, Psychedelic Rock  \n",
              "1       Rock     Pop Rock, Psychedelic Rock  \n",
              "2       Rock     Psychedelic Rock, Pop Rock  \n",
              "3       Rock          Folk Rock, Blues Rock  \n",
              "4  Rock, Pop                       Pop Rock  "
            ]
          },
          "execution_count": 7,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df1 = pd.read_csv('data/albumlist.csv', encoding='mac_roman')\n",
        "df1.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "4c614dab-619c-455c-bd63-565ff79619e9",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 354
        },
        "id": "4c614dab-619c-455c-bd63-565ff79619e9",
        "outputId": "1bcbde9b-7e60-4f56-f87d-34d5aaac39c7"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Ranking</th>\n",
              "      <th>Album</th>\n",
              "      <th>Artist Name</th>\n",
              "      <th>Release Date</th>\n",
              "      <th>Genres</th>\n",
              "      <th>Descriptors</th>\n",
              "      <th>Average Rating</th>\n",
              "      <th>Number of Ratings</th>\n",
              "      <th>Number of Reviews</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1.0</td>\n",
              "      <td>OK Computer</td>\n",
              "      <td>Radiohead</td>\n",
              "      <td>16 June 1997</td>\n",
              "      <td>Alternative Rock, Art Rock</td>\n",
              "      <td>melancholic, anxious, futuristic, alienation, ...</td>\n",
              "      <td>4.23</td>\n",
              "      <td>70,382</td>\n",
              "      <td>1531</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>2.0</td>\n",
              "      <td>Wish You Were Here</td>\n",
              "      <td>Pink Floyd</td>\n",
              "      <td>12 September 1975</td>\n",
              "      <td>Progressive Rock, Art Rock</td>\n",
              "      <td>melancholic, atmospheric, progressive, male vo...</td>\n",
              "      <td>4.29</td>\n",
              "      <td>48,662</td>\n",
              "      <td>983</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>3.0</td>\n",
              "      <td>In the Court of the Crimson King</td>\n",
              "      <td>King Crimson</td>\n",
              "      <td>10 October 1969</td>\n",
              "      <td>Progressive Rock, Art Rock</td>\n",
              "      <td>fantasy, epic, progressive, philosophical, com...</td>\n",
              "      <td>4.30</td>\n",
              "      <td>44,943</td>\n",
              "      <td>870</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>4.0</td>\n",
              "      <td>Kid A</td>\n",
              "      <td>Radiohead</td>\n",
              "      <td>3 October 2000</td>\n",
              "      <td>Art Rock, Experimental Rock, Electronic</td>\n",
              "      <td>cold, melancholic, futuristic, atmospheric, an...</td>\n",
              "      <td>4.21</td>\n",
              "      <td>58,590</td>\n",
              "      <td>734</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>5.0</td>\n",
              "      <td>To Pimp a Butterfly</td>\n",
              "      <td>Kendrick Lamar</td>\n",
              "      <td>15 March 2015</td>\n",
              "      <td>Conscious Hip Hop, West Coast Hip Hop, Jazz Rap</td>\n",
              "      <td>political, conscious, poetic, protest, concept...</td>\n",
              "      <td>4.27</td>\n",
              "      <td>44,206</td>\n",
              "      <td>379</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Ranking                             Album     Artist Name  \\\n",
              "0      1.0                       OK Computer       Radiohead   \n",
              "1      2.0                Wish You Were Here      Pink Floyd   \n",
              "2      3.0  In the Court of the Crimson King    King Crimson   \n",
              "3      4.0                             Kid A       Radiohead   \n",
              "4      5.0               To Pimp a Butterfly  Kendrick Lamar   \n",
              "\n",
              "        Release Date                                           Genres  \\\n",
              "0       16 June 1997                       Alternative Rock, Art Rock   \n",
              "1  12 September 1975                       Progressive Rock, Art Rock   \n",
              "2    10 October 1969                       Progressive Rock, Art Rock   \n",
              "3     3 October 2000          Art Rock, Experimental Rock, Electronic   \n",
              "4      15 March 2015  Conscious Hip Hop, West Coast Hip Hop, Jazz Rap   \n",
              "\n",
              "                                         Descriptors  Average Rating  \\\n",
              "0  melancholic, anxious, futuristic, alienation, ...            4.23   \n",
              "1  melancholic, atmospheric, progressive, male vo...            4.29   \n",
              "2  fantasy, epic, progressive, philosophical, com...            4.30   \n",
              "3  cold, melancholic, futuristic, atmospheric, an...            4.21   \n",
              "4  political, conscious, poetic, protest, concept...            4.27   \n",
              "\n",
              "  Number of Ratings  Number of Reviews  \n",
              "0            70,382               1531  \n",
              "1            48,662                983  \n",
              "2            44,943                870  \n",
              "3            58,590                734  \n",
              "4            44,206                379  "
            ]
          },
          "execution_count": 8,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df2 = pd.read_csv('data/rym_top_5000_all_time.csv')\n",
        "df2.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "621c566f-64b4-47a7-be9b-f4869942bfcf",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "621c566f-64b4-47a7-be9b-f4869942bfcf",
        "outputId": "6bb02c57-3cbd-49d3-f782-5d9a24be89b6"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "((500, 6), (5000, 9))"
            ]
          },
          "execution_count": 9,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df1.shape, df2.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "HvP9_4Y1kudm",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HvP9_4Y1kudm",
        "outputId": "5d5fbefd-d5f2-4c93-9bed-1531b1ee406d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "All elements of df1['Album'] are unique: False \n",
            "All elements of df2['Album'] are unique: False\n"
          ]
        }
      ],
      "source": [
        "# Check if all elements in 'Album' column are unique\n",
        "is_unique_df1 = not df1['Album'].duplicated().any()\n",
        "is_unique_df2 = not df2['Album'].duplicated().any()\n",
        "\n",
        "print(f\"All elements of df1['Album'] are unique: {is_unique_df1} \\nAll elements of df2['Album'] are unique: {is_unique_df2}\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "N0MVDutJ7Cbl",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "N0MVDutJ7Cbl",
        "outputId": "c4d03d49-a04c-40f8-eb11-28a5b3a6505d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "All elements of df1['Album'] are present in df2['Album']: False\n"
          ]
        }
      ],
      "source": [
        "# Check if all elements of df1['Album'] are present in df2['Album']\n",
        "all_present = df1['Album'].isin(df2['Album']).all()\n",
        "\n",
        "print(f\"All elements of df1['Album'] are present in df2['Album']: {all_present}\")"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "rKKQ1kZlly3k",
      "metadata": {
        "id": "rKKQ1kZlly3k"
      },
      "source": [
        "The elements in the `Album` column are not unique, in both the dataframes, so if there are duplicates I can't join on the `Album` column.\n",
        "\n",
        "Since I don't have that df2 contains each album in df1, I can't apply a left or right join, but I should apply an outer join."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "QrW9H1LTu3Sm",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "QrW9H1LTu3Sm",
        "outputId": "d051b66e-4f62-462a-da49-cb5e61543f54"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "All pairs of elements df1['Album', 'Artist'] are unique: True \n",
            "All pairs of elements df2['Album', 'Artist Name'] are unique: False\n"
          ]
        }
      ],
      "source": [
        "# Check if all pairs of elements from 'Name' and 'Age' columns are unique\n",
        "is_unique_df1 = not df1.duplicated(subset=['Album', 'Artist']).any()\n",
        "is_unique_df2 = not df2.duplicated(subset=['Album', 'Artist Name']).any()\n",
        "\n",
        "print(f\"All pairs of elements df1['Album', 'Artist'] are unique: {is_unique_df1} \\nAll pairs of elements df2['Album', 'Artist Name'] are unique: {is_unique_df2}\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "sonunO6cGfQy",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sonunO6cGfQy",
        "outputId": "94a58649-2578-4042-fc1d-d5896733f0bf"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "(23, 9)"
            ]
          },
          "execution_count": 13,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# DataFrame with the elements that have a duplicated pair df2['Album', 'Artist Name']\n",
        "duplicated_pairs_df2 = df2[df2.duplicated(subset=['Album', 'Artist Name'], keep=False)]\n",
        "\n",
        "duplicated_pairs_df2.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "PybH5A1OwAH3",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PybH5A1OwAH3",
        "outputId": "2b4d6dd6-ed4e-4603-d23b-b85942b722c9"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "All pairs of elements df2['Album', 'Artist Name'] that are not unique, are not present in df1.\n"
          ]
        }
      ],
      "source": [
        "# Find the common pairs between df1 and duplicated_pairs_df2\n",
        "# (All pairs of elements ['Album', 'Artist Name'] in df2 that are not unique, that we have also in df1)\n",
        "common_pairs = df1.merge(duplicated_pairs_df2, left_on=['Album', 'Artist'], right_on=['Album', 'Artist Name'])\n",
        "\n",
        "if common_pairs.empty:\n",
        "  print(\"All pairs of elements df2['Album', 'Artist Name'] that are not unique, are not present in df1.\")"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "bkmOYY5IFh77",
      "metadata": {
        "id": "bkmOYY5IFh77"
      },
      "source": [
        "After some further analysis, I know that in df2 I have some row that have the same pair `['Album', 'Artist Name']`; but fortunately none of this pairs appear in any row of df1.\n",
        "\n",
        "Now the last step that remains is to find a way to solve this issue with the duplicated elements in df2, so that at the end I can merge the two DataFrames in the best way possible.\n",
        "\n",
        "A possible way is to assign a general name, like `Album{N}`, so that I can create a new index based on the couple of columns `['Album', 'Artist Name']`, that after our changes doesn't have any duplicate."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "ZREtmg7kUlPv",
      "metadata": {
        "id": "ZREtmg7kUlPv"
      },
      "outputs": [],
      "source": [
        "# In \"duplicated_pairs_df2\" I have the elements that have a duplicated pair df2['Album', 'Artist Name']\n",
        "# duplicated_pairs_df2 = df2[df2.duplicated(subset=['Album', 'Artist Name'], keep=False)]    (I have already run it in the previous cell)\n",
        "\n",
        "# Now I retrieve the indexes of the rows that I have to change\n",
        "row_indexes = duplicated_pairs_df2.index.tolist()\n",
        "\n",
        "# Change the values of specific rows in the 'Album' column\n",
        "# As new values for the corresponding rows I will use ['Album0', 'Album1'...]\n",
        "new_names = list()\n",
        "for N in range(len(row_indexes)):\n",
        "  new_names.append(f'Album{N}')\n",
        "\n",
        "df2.loc[row_indexes, 'Album'] = new_names"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "Ag2MuUzx9Qd0",
      "metadata": {
        "id": "Ag2MuUzx9Qd0"
      },
      "source": [
        "Finally I merge the two DataFrames.\n",
        "\n",
        "I set the couple `['Album', 'Artist']` as index, since after our changes each of this couples is unique. Then I merge on the index, that is faster then merging on simple columns.\n",
        "\n",
        "The result is the outer join of the two DataFrames."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "Yl-u6BL89YnG",
      "metadata": {
        "id": "Yl-u6BL89YnG"
      },
      "outputs": [],
      "source": [
        "# I rename the 'Artist Name' column of df2 to 'Artist' name, so it is equal to the DataFrame df1\n",
        "df2 = df2.rename(columns={'Artist Name': 'Artist'})\n",
        "\n",
        "df1.set_index(['Album', 'Artist'], inplace=True)\n",
        "df2.set_index(['Album', 'Artist'], inplace=True)\n",
        "\n",
        "df_merged = pd.merge(df1, df2, left_index=True, right_index=True, how='outer')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "kz3ZZUYfLBlV",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 385
        },
        "id": "kz3ZZUYfLBlV",
        "outputId": "f405c96d-e397-4e51-f1a5-1d9b5a44d4b0"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th>Number</th>\n",
              "      <th>Year</th>\n",
              "      <th>Genre</th>\n",
              "      <th>Subgenre</th>\n",
              "      <th>Ranking</th>\n",
              "      <th>Release Date</th>\n",
              "      <th>Genres</th>\n",
              "      <th>Descriptors</th>\n",
              "      <th>Average Rating</th>\n",
              "      <th>Number of Ratings</th>\n",
              "      <th>Number of Reviews</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Album</th>\n",
              "      <th>Artist</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>!!Destroy-Oh-Boy!!</th>\n",
              "      <th>New Bomb Turks</th>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>4872.0</td>\n",
              "      <td>1993</td>\n",
              "      <td>Garage Punk, Punk Rock</td>\n",
              "      <td>energetic, humorous, sarcastic, male vocals, p...</td>\n",
              "      <td>3.69</td>\n",
              "      <td>676</td>\n",
              "      <td>14.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>\"Al final de este viaje...\"</th>\n",
              "      <th>Silvio Rodríguez</th>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>682.0</td>\n",
              "      <td>1978</td>\n",
              "      <td>Nueva trova</td>\n",
              "      <td>acoustic, passionate, male vocals, poetic, con...</td>\n",
              "      <td>3.91</td>\n",
              "      <td>1,489</td>\n",
              "      <td>33.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>\"Born Into Trouble as the Sparks Fly Upward.\"</th>\n",
              "      <th>The Silver Mt. Zion Memorial Orchestra &amp; Tra-La-La Band</th>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>4763.0</td>\n",
              "      <td>23 October 2001</td>\n",
              "      <td>Post-Rock</td>\n",
              "      <td>melancholic, apocalyptic, suspenseful, atmosph...</td>\n",
              "      <td>3.65</td>\n",
              "      <td>3,982</td>\n",
              "      <td>51.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>\"Chrono Trigger\"</th>\n",
              "      <th>Yasunori Mitsuda</th>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>240.0</td>\n",
              "      <td>25 March 1995</td>\n",
              "      <td>Video Game Music, 16-bit</td>\n",
              "      <td>epic, fantasy, instrumental, eclectic, atmosph...</td>\n",
              "      <td>4.01</td>\n",
              "      <td>1,689</td>\n",
              "      <td>25.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>\"Heroes\"</th>\n",
              "      <th>David Bowie</th>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>155.0</td>\n",
              "      <td>14 October 1977</td>\n",
              "      <td>Art Rock</td>\n",
              "      <td>futuristic, atmospheric, male vocals, anxious,...</td>\n",
              "      <td>3.97</td>\n",
              "      <td>22,443</td>\n",
              "      <td>292.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                                                                                  Number  \\\n",
              "Album                                         Artist                                                       \n",
              "!!Destroy-Oh-Boy!!                            New Bomb Turks                                         NaN   \n",
              "\"Al final de este viaje...\"                   Silvio Rodríguez                                       NaN   \n",
              "\"Born Into Trouble as the Sparks Fly Upward.\" The Silver Mt. Zion Memorial Orchestra & Tra-La...     NaN   \n",
              "\"Chrono Trigger\"                              Yasunori Mitsuda                                       NaN   \n",
              "\"Heroes\"                                      David Bowie                                            NaN   \n",
              "\n",
              "                                                                                                  Year  \\\n",
              "Album                                         Artist                                                     \n",
              "!!Destroy-Oh-Boy!!                            New Bomb Turks                                       NaN   \n",
              "\"Al final de este viaje...\"                   Silvio Rodríguez                                     NaN   \n",
              "\"Born Into Trouble as the Sparks Fly Upward.\" The Silver Mt. Zion Memorial Orchestra & Tra-La...   NaN   \n",
              "\"Chrono Trigger\"                              Yasunori Mitsuda                                     NaN   \n",
              "\"Heroes\"                                      David Bowie                                          NaN   \n",
              "\n",
              "                                                                                                 Genre  \\\n",
              "Album                                         Artist                                                     \n",
              "!!Destroy-Oh-Boy!!                            New Bomb Turks                                       NaN   \n",
              "\"Al final de este viaje...\"                   Silvio Rodríguez                                     NaN   \n",
              "\"Born Into Trouble as the Sparks Fly Upward.\" The Silver Mt. Zion Memorial Orchestra & Tra-La...   NaN   \n",
              "\"Chrono Trigger\"                              Yasunori Mitsuda                                     NaN   \n",
              "\"Heroes\"                                      David Bowie                                          NaN   \n",
              "\n",
              "                                                                                                 Subgenre  \\\n",
              "Album                                         Artist                                                        \n",
              "!!Destroy-Oh-Boy!!                            New Bomb Turks                                          NaN   \n",
              "\"Al final de este viaje...\"                   Silvio Rodríguez                                        NaN   \n",
              "\"Born Into Trouble as the Sparks Fly Upward.\" The Silver Mt. Zion Memorial Orchestra & Tra-La...      NaN   \n",
              "\"Chrono Trigger\"                              Yasunori Mitsuda                                        NaN   \n",
              "\"Heroes\"                                      David Bowie                                             NaN   \n",
              "\n",
              "                                                                                                  Ranking  \\\n",
              "Album                                         Artist                                                        \n",
              "!!Destroy-Oh-Boy!!                            New Bomb Turks                                       4872.0   \n",
              "\"Al final de este viaje...\"                   Silvio Rodríguez                                      682.0   \n",
              "\"Born Into Trouble as the Sparks Fly Upward.\" The Silver Mt. Zion Memorial Orchestra & Tra-La...   4763.0   \n",
              "\"Chrono Trigger\"                              Yasunori Mitsuda                                      240.0   \n",
              "\"Heroes\"                                      David Bowie                                           155.0   \n",
              "\n",
              "                                                                                                     Release Date  \\\n",
              "Album                                         Artist                                                                \n",
              "!!Destroy-Oh-Boy!!                            New Bomb Turks                                                 1993   \n",
              "\"Al final de este viaje...\"                   Silvio Rodríguez                                               1978   \n",
              "\"Born Into Trouble as the Sparks Fly Upward.\" The Silver Mt. Zion Memorial Orchestra & Tra-La...  23 October 2001   \n",
              "\"Chrono Trigger\"                              Yasunori Mitsuda                                      25 March 1995   \n",
              "\"Heroes\"                                      David Bowie                                         14 October 1977   \n",
              "\n",
              "                                                                                                                    Genres  \\\n",
              "Album                                         Artist                                                                         \n",
              "!!Destroy-Oh-Boy!!                            New Bomb Turks                                        Garage Punk, Punk Rock   \n",
              "\"Al final de este viaje...\"                   Silvio Rodríguez                                                 Nueva trova   \n",
              "\"Born Into Trouble as the Sparks Fly Upward.\" The Silver Mt. Zion Memorial Orchestra & Tra-La...                 Post-Rock   \n",
              "\"Chrono Trigger\"                              Yasunori Mitsuda                                    Video Game Music, 16-bit   \n",
              "\"Heroes\"                                      David Bowie                                                         Art Rock   \n",
              "\n",
              "                                                                                                                                        Descriptors  \\\n",
              "Album                                         Artist                                                                                                  \n",
              "!!Destroy-Oh-Boy!!                            New Bomb Turks                                      energetic, humorous, sarcastic, male vocals, p...   \n",
              "\"Al final de este viaje...\"                   Silvio Rodríguez                                    acoustic, passionate, male vocals, poetic, con...   \n",
              "\"Born Into Trouble as the Sparks Fly Upward.\" The Silver Mt. Zion Memorial Orchestra & Tra-La...  melancholic, apocalyptic, suspenseful, atmosph...   \n",
              "\"Chrono Trigger\"                              Yasunori Mitsuda                                    epic, fantasy, instrumental, eclectic, atmosph...   \n",
              "\"Heroes\"                                      David Bowie                                         futuristic, atmospheric, male vocals, anxious,...   \n",
              "\n",
              "                                                                                                  Average Rating  \\\n",
              "Album                                         Artist                                                               \n",
              "!!Destroy-Oh-Boy!!                            New Bomb Turks                                                3.69   \n",
              "\"Al final de este viaje...\"                   Silvio Rodríguez                                              3.91   \n",
              "\"Born Into Trouble as the Sparks Fly Upward.\" The Silver Mt. Zion Memorial Orchestra & Tra-La...            3.65   \n",
              "\"Chrono Trigger\"                              Yasunori Mitsuda                                              4.01   \n",
              "\"Heroes\"                                      David Bowie                                                   3.97   \n",
              "\n",
              "                                                                                                 Number of Ratings  \\\n",
              "Album                                         Artist                                                                 \n",
              "!!Destroy-Oh-Boy!!                            New Bomb Turks                                                   676   \n",
              "\"Al final de este viaje...\"                   Silvio Rodríguez                                               1,489   \n",
              "\"Born Into Trouble as the Sparks Fly Upward.\" The Silver Mt. Zion Memorial Orchestra & Tra-La...             3,982   \n",
              "\"Chrono Trigger\"                              Yasunori Mitsuda                                               1,689   \n",
              "\"Heroes\"                                      David Bowie                                                   22,443   \n",
              "\n",
              "                                                                                                  Number of Reviews  \n",
              "Album                                         Artist                                                                 \n",
              "!!Destroy-Oh-Boy!!                            New Bomb Turks                                                   14.0  \n",
              "\"Al final de este viaje...\"                   Silvio Rodríguez                                                 33.0  \n",
              "\"Born Into Trouble as the Sparks Fly Upward.\" The Silver Mt. Zion Memorial Orchestra & Tra-La...               51.0  \n",
              "\"Chrono Trigger\"                              Yasunori Mitsuda                                                 25.0  \n",
              "\"Heroes\"                                      David Bowie                                                     292.0  "
            ]
          },
          "execution_count": 17,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df_merged.head()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "cmBWTvHdNfYm",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cmBWTvHdNfYm",
        "outputId": "1e6ab056-0417-454e-824d-520181dbcbe7"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "(5205, 11)"
            ]
          },
          "execution_count": 18,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df_merged.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "yPonqAwH_fci",
      "metadata": {
        "id": "yPonqAwH_fci"
      },
      "outputs": [],
      "source": [
        "df_merged.to_csv('data/merged_dataframe.csv', index=True)"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "SBhmD19mnVMx",
      "metadata": {
        "id": "SBhmD19mnVMx"
      },
      "source": [
        "### 1.2 ) Create an RDFLib `Graph` and add triples from the data of the dataframes"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "r8bXKb3N99Tm",
      "metadata": {
        "id": "r8bXKb3N99Tm"
      },
      "source": [
        "I start creating:\n",
        "  \n",
        "*   an RDFLib `Graph`,\n",
        "*   the namespace `df` that we will use to add the elements of our merged dataframe to the graph,\n",
        "*   the URIs that we will use as predicate and class.\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "bpjqxNNnG9HK",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "id": "bpjqxNNnG9HK",
        "outputId": "00758e3a-43d2-4e3b-a5b2-f9002d9b4f6e"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "rdflib.term.URIRef('http://example.org/DataFrame/numberOfReviews')"
            ]
          },
          "execution_count": 20,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Create our graph\n",
        "rdf_graph = Graph()\n",
        "df = Namespace('http://example.org/DataFrame/')\n",
        "rdf_graph.namespace_manager.bind('df', 'http://example.org/DataFrame/')\n",
        "\n",
        "# Create some URI that I will use as predicate or object for my triples\n",
        "df['Album']\n",
        "df['Artist']\n",
        "df['Genre']\n",
        "df['artist']\n",
        "df['releaseDate']\n",
        "df['year']\n",
        "df['genre']\n",
        "df['subgenre']\n",
        "df['descriptors']\n",
        "df['rsRanking']\n",
        "df['rymRanking']\n",
        "df['averageRating']\n",
        "df['numberOfRatings']\n",
        "df['numberOfReviews']"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "PzzSI7SIc4f4",
      "metadata": {
        "id": "PzzSI7SIc4f4"
      },
      "source": [
        "I will have something of this kind:"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "XZPG7-_Hc7xT",
      "metadata": {
        "id": "XZPG7-_Hc7xT",
        "outputId": "1877d968-a8c2-4fc7-fdb1-cc07c233d333"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "execution_count": 21,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Visualizing a sample graph\n",
        "g_sample = Graph()\n",
        "df = Namespace('http://example.org/DataFrame/')\n",
        "g_sample.namespace_manager.bind('df', 'http://example.org/DataFrame/')\n",
        "\n",
        "album_uri = Literal(\"_:album\")\n",
        "data = (\n",
        "    (album_uri, RDF.type, df.Album),\n",
        "    (album_uri, df.artist, Literal(\"_:artist\")),\n",
        "    (Literal(\"_:artist\"), RDF.type, df.Artist),\n",
        "    (album_uri, df.averageRating, Literal(\"3.88\")),\n",
        "    (album_uri, df.numberOfRatings, Literal(\"8442\")),\n",
        "    (album_uri, df.numberOfReviews, Literal(\"25.0\")),\n",
        "    (album_uri, df.rymRanking, Literal(\"1930.0\")),\n",
        "    (album_uri, df.rsRanking, Literal(\"279\")),\n",
        "    (album_uri, df.releaseDate, Literal(\"1989-12-07\", datatype=XSD.date)),\n",
        "    (album_uri, df.year, Literal(\"1989\", datatype=XSD.gYear)),\n",
        "    (album_uri, df.genre, df.Rock),\n",
        "    (album_uri, df.subgenre, df.JRock),\n",
        "    (album_uri, df.subgenre, df.Alternative_Rock),\n",
        "    (album_uri, df.descriptors, Literal(\"energetic, humorous, sarcastic\"))\n",
        ")\n",
        "for triple in data:\n",
        "    g_sample.add(triple)\n",
        "\n",
        "G_sample = rdflib_to_networkx_multidigraph(g_sample, edge_attrs = lambda s, p, o: {\"label\" : p.n3(g_sample.namespace_manager)})\n",
        "mapping = {node: node.n3(g_sample.namespace_manager) for node in G_sample.nodes}\n",
        "G_sample = nx.relabel_nodes(G_sample, mapping)\n",
        "A = nx.nx_agraph.to_agraph(G_sample)\n",
        "A.draw(\"images/example.png\", prog='dot', args=\"-Gdpi=1000 -Gsize=5\")\n",
        "Image(\"images/example.png\")"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "5ckMAcpBAaKn",
      "metadata": {
        "id": "5ckMAcpBAaKn"
      },
      "source": [
        "**FUNCTIONS**"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "Ll48hmXGdN-a",
      "metadata": {
        "id": "Ll48hmXGdN-a"
      },
      "source": [
        "I create some function that will transform the simple literals and integers that I have in the tables in a form that will be useful for adding them in the graph.\n",
        "\n",
        "In the URIs for the album and the artist I will replace spaces with underscore `_`, removes the quotes `\"` and the accents. And I will link those URIs with an `RDFS.label`, which contains again the artist and the album, but as `datatype=XSD.string`.\n",
        "\n",
        "For the floats values I have to remove the comma `,` since in the column `Number of Ratings` it is used for representing the thousands.\n",
        "\n",
        "Language tagged strings are simply added as strings and `lang=\"en\"`, and will be used in particular for the descriptors"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "V_YNO_fcuk2U",
      "metadata": {
        "id": "V_YNO_fcuk2U"
      },
      "outputs": [],
      "source": [
        "# Here I create the functions that will transform the values in the dataframe\n",
        "### URI:\n",
        "def value_to_uri(value):\n",
        "    modified_string = value.replace(\" \", \"_\")\n",
        "    modified_string = modified_string.replace('\"', '')\n",
        "    modified_string = modified_string.replace('à', 'a')\n",
        "    modified_string = modified_string.replace('ã', 'a')\n",
        "    modified_string = modified_string.replace('ä', 'a')\n",
        "    modified_string = modified_string.replace('é', 'e')\n",
        "    modified_string = modified_string.replace('è', 'e')\n",
        "    modified_string = modified_string.replace('ï', 'i')\n",
        "    modified_string = modified_string.replace('í', 'i')\n",
        "    modified_string = modified_string.replace('ò', 'o')\n",
        "    modified_string = modified_string.replace('ö', 'o')\n",
        "    modified_string = modified_string.replace('ù', 'u')\n",
        "    modified_string = modified_string.replace('ü', 'u')\n",
        "    modified_string = modified_string.replace('ç', 'c')\n",
        "    return df[ modified_string ]\n",
        "\n",
        "### Float:\n",
        "def value_to_float(value):\n",
        "    # I have to remove the comma (representing the thousands),\n",
        "    # so I first transform into a string so that I can use the replace() method\n",
        "    value_to_str = str(value)\n",
        "    value_whitout_comma = value_to_str.replace(\",\", \"\")\n",
        "    value_float = float(value_whitout_comma)\n",
        "    return Literal(value_float, datatype=XSD.float)\n",
        "\n",
        "\n",
        "### Language tagged string:\n",
        "def value_to_tagged_string(value):\n",
        "    value_string = str(value)\n",
        "    return Literal(value_string, lang=\"en\")\n"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "QybG36-pulWF",
      "metadata": {
        "id": "QybG36-pulWF"
      },
      "source": [
        "The function that will transform the date and the years take into considerations more cases, since the the column `'Release Date'`, the dates appears in many formats, therefore to transfer this data in the cleanest way into my graph I will:\n",
        "\n",
        "\n",
        "*   transforme the dates in the format YYYY-MM-DD and use them as `datatype=XSD.date`,\n",
        "*   if I have simply the year, I will use the format YYYY, as `datatype=XSD.gYear`.\n",
        "\n",
        "In particular in the column `['Date']` I can have the date in the format 'DD month YYYY' or 'month YYYY' or 'YYYY', while in `['Year']` simply 'YYYY' (as a float). In the rows in which I don't have a value for the year, I will extract the information from the date column, and use that one."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "3UCkiZ0kC5H2",
      "metadata": {
        "id": "3UCkiZ0kC5H2"
      },
      "outputs": [],
      "source": [
        "# Check if a value is \"NaN\"\n",
        "def is_nan(value):\n",
        "    if isinstance(value, str) or isinstance(value, int):\n",
        "        return False\n",
        "    return math.isnan(value)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "m6XnzIqJHNIc",
      "metadata": {
        "id": "m6XnzIqJHNIc"
      },
      "outputs": [],
      "source": [
        "### Date and Year, with xsd:date and xsd:gYear:\n",
        "# The following function check if the format of date_string matches the date_format\n",
        "def is_valid_date_format(date_string, date_format):\n",
        "    try:\n",
        "        datetime.strptime(date_string, date_format)\n",
        "        return True\n",
        "    except:\n",
        "        return False\n",
        "\n",
        "# Input 1: Date, Input 2: Year\n",
        "# Output 1: Date, Output 2: Year\n",
        "def value_to_date_year(date_value, year_value):\n",
        "\n",
        "    # If I have only the year\n",
        "    if is_nan(date_value):\n",
        "        year_value = int(year_value)\n",
        "        return None, Literal(year_value, datatype=XSD.gYear)\n",
        "\n",
        "    # If I have only the date\n",
        "    # I extract the year from it\n",
        "    elif is_nan(year_value):\n",
        "        # check each format one by one\n",
        "        # Day, month and year\n",
        "        if is_valid_date_format(date_value, '%d %B %Y'):\n",
        "            # Convert string to datetime object\n",
        "            date_object = datetime.strptime(date_value, '%d %B %Y')\n",
        "            # Convert datetime object to desired string format 'yyyy-mm-dd'\n",
        "            output_date = date_object.strftime('%Y-%m-%d')\n",
        "            # Extract the year in the format 'yyyy'\n",
        "            output_year = date_object.strftime('%Y')\n",
        "            return Literal(output_date, datatype=XSD.date), Literal(output_year, datatype=XSD.gYear)\n",
        "        # Repeat with the other formats\n",
        "        # Month and year\n",
        "        elif is_valid_date_format(date_value, '%B %Y'):\n",
        "            date_object = datetime.strptime(date_value, '%B %Y')\n",
        "            output_date = date_object.strftime('%Y-%m-%d')\n",
        "            output_year = date_object.strftime('%Y')\n",
        "            return Literal(output_date, datatype=XSD.date), Literal(output_year, datatype=XSD.gYear)\n",
        "        # If I have only the year, I will use it as datatype=XSD.gYear\n",
        "        elif is_valid_date_format(date_value, '%Y'):\n",
        "            return None, Literal(date_value, datatype=XSD.gYear)\n",
        "\n",
        "    # If I have both, I do similarly to the previous point, but I give as result also the year value, taked from ['Year'] column\n",
        "    else:\n",
        "        year_value = int(year_value)\n",
        "        if is_valid_date_format(date_value, '%d %B %Y'):\n",
        "            date_object = datetime.strptime(date_value, '%d %B %Y')\n",
        "            output_date = date_object.strftime('%Y-%m-%d')\n",
        "            return Literal(output_date, datatype=XSD.date), Literal(year_value, datatype=XSD.gYear)\n",
        "\n",
        "        elif is_valid_date_format(date_value, '%B %Y'):\n",
        "            date_object = datetime.strptime(date_value, '%B %Y')\n",
        "            output_date = date_object.strftime('%Y-%m-%d')\n",
        "            return Literal(output_date, datatype=XSD.date), Literal(year_value, datatype=XSD.gYear)\n",
        "\n",
        "        # In this case I will maintain only the value from the year column\n",
        "        elif is_valid_date_format(date_value, '%Y'):\n",
        "            return None, Literal(year_value, datatype=XSD.gYear)"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "vflbpjeD9txV",
      "metadata": {
        "id": "vflbpjeD9txV"
      },
      "source": [
        "The function that will handle the columns `['Genre', 'Subgenre']` from df1 and `['Genres']` from df2 will work in the following way.\n",
        "\n",
        "I will use only two predicate, `df.genre` and `df.subgenre`\n",
        "\n",
        "For each album:\n",
        "\n",
        "*   if I have only `['Genre', 'Subgenre']`, use directly them as `df.genre` and `df.subgenre`,\n",
        "*   if I have only `['Genres']`, I will extract the main genre for `df.genre` and put everything that I have in `['Genres']` as `df.subgenre`,\n",
        "*   if I have both I will integrate the elements in `['Genre', 'Subgenre']` by adding the ones in `['Genres']`.\n",
        "\n",
        "For each element I need to parse it, to obtain each genre separatly. My idea is to collect every genre in a dictionary, so that I can check if one of this value is present in `['Genres']`, so that I can extract it as main genre, and then put the others as subgenres."
      ]
    },
    {
      "cell_type": "markdown",
      "id": "CWko_8AzQMH_",
      "metadata": {
        "id": "CWko_8AzQMH_"
      },
      "source": [
        "\n",
        "\n",
        "----------\n",
        "\n",
        "\n",
        "Build a dictionary which collect every genre that I have in the column `['Genre']`:\n",
        "\n",
        "I create a duplicate column of `['Genre']` as a separate series, remove all the `NaN` values and do some operations:\n",
        "\n",
        "after I want to split each list-value of my dictionary (since some element of my series have more than one genre), so I have to change some special character that appears in my values\n",
        "\n",
        "*   Replace `', & '` with `', '`, since it appears for ex. `'Folk, World, & Country'`, but also `'Stage & Screen'` which is a unique genre that I don't want to split,\n",
        "*   Replace `' / '` with `', '`, since it appears for ex. `'Funk / Soul'`,\n",
        "*   Replace `'\\xa0'` with `' '`, since it appears for ex. `'Rock,\\xa0Pop'`,\n",
        "\n",
        "Then `split(', ')`, convert my modified series into a dictionary, split each list which appear as value in my dictionary as multiple key-value (with as value we have a simple string), finally eliminate duplicated values and rename the keys.\n",
        "\n",
        "For each genre I will attach an `RDFS.label` with the name of the genre as `datatype=XSD.string`"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "9AKyQ4VKIIMC",
      "metadata": {
        "id": "9AKyQ4VKIIMC"
      },
      "outputs": [],
      "source": [
        "# Build A dictionary with the genres\n",
        "\n",
        "# Create a duplicate column as a separated Series\n",
        "genre_column_copy = df_merged['Genre'].copy()\n",
        "\n",
        "# Remove NaN values from the Series\n",
        "genre_column_without_nan = genre_column_copy.dropna()\n",
        "\n",
        "# Replace ', & ' with ', '\n",
        "# Replace ' / ' with ', '\n",
        "# in each value of the Series\n",
        "genre_column_modified = genre_column_without_nan.apply(lambda x: x.replace(', & ', ', '))\n",
        "genre_column_modified = genre_column_modified.apply(lambda x: x.replace(' / ', ', '))\n",
        "genre_column_modified = genre_column_modified.apply(lambda x: x.replace('\\xa0', ' '))\n",
        "\n",
        "# Split the values in the duplicated column\n",
        "genre_column_splitted = genre_column_modified.str.split(', ')\n",
        "\n",
        "# Convert the duplicated column to a dictionary\n",
        "genre_column_dict_list = genre_column_splitted.to_dict()\n",
        "\n",
        "# Split each value in the original dictionary\n",
        "# and store it in a new dictionary, with each value splitted (insted of having lists of values)\n",
        "genre_dict_with_duplicates = {}\n",
        "\n",
        "for key, values in genre_column_dict_list.items():\n",
        "    for value in values:\n",
        "        new_key = f'{key}_{value}'\n",
        "        genre_dict_with_duplicates[new_key] = value\n",
        "\n",
        "# Create a set to store unique values\n",
        "unique_values = set()\n",
        "\n",
        "# Filter out duplicate values from the dictionary\n",
        "genre_dict_with_index_as_keys = {key: value for key, value in genre_dict_with_duplicates.items() if value not in unique_values and not unique_values.add(value)}\n",
        "\n",
        "# Create a new dictionary with renamed keys\n",
        "genre_dict = {str(i+1): value for i, (_, value) in enumerate(genre_dict_with_index_as_keys.items())}\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "1aEEWRzzMLlq",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1aEEWRzzMLlq",
        "outputId": "9a294674-e5d9-41ec-d792-4ce1c2a25c63"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Here I have my genres:\n",
            "1: Rock\n",
            "2: Pop\n",
            "3: Funk\n",
            "4: Soul\n",
            "5: Hip Hop\n",
            "6: Folk\n",
            "7: World\n",
            "8: Country\n",
            "9: Jazz\n",
            "10: Electronic\n",
            "11: Stage & Screen\n",
            "12: Latin\n",
            "13: Blues\n",
            "14: Reggae\n",
            "15: Classical\n"
          ]
        }
      ],
      "source": [
        "print(\"Here I have my genres:\")\n",
        "for key, value in genre_dict.items():\n",
        "    print(f'{key}: {value}')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "bg8X7X0gJuzM",
      "metadata": {
        "id": "bg8X7X0gJuzM"
      },
      "outputs": [],
      "source": [
        "# 'genre 1, genre 2' -> 'genre_1,_genre_2' -> ['genre_1', 'genre_2']\n",
        "def value_to_correct_format(value):\n",
        "    # Replace /\n",
        "    value_modified = value.replace(\" / \", \", \")\n",
        "    # Replace &\n",
        "    value_modified = value_modified.replace(', & ', ', ')\n",
        "    # Replace spaces\n",
        "    value_modified = value_modified.replace(\" \", \"_\")\n",
        "    # Split the string into individual words\n",
        "    value_list = value_modified.split(',_')\n",
        "    return value_list"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "bKvVvb4aAS-D",
      "metadata": {
        "id": "bKvVvb4aAS-D"
      },
      "outputs": [],
      "source": [
        "# Input 1: Genre_df1, Input 2: Subgenre_df1, Input 3: Genres_df2\n",
        "# As output I return lists, than I will get each value by doing an iteraton when I build the graph\n",
        "# Output 1: Genre_list, Output 2: Subgenre_list\n",
        "def value_to_genre_subgenre(genre_df1_value, subgenre_df1_value, genres_df2_value):\n",
        "    # If I have only genre and subgenre\n",
        "    if is_nan(genres_df2_value):\n",
        "        # Put in the correct format\n",
        "        genre_list = value_to_correct_format(genre_df1_value)\n",
        "        subgenre_list = value_to_correct_format(subgenre_df1_value)\n",
        "        return genre_list, subgenre_list\n",
        "\n",
        "    # If I have only genres\n",
        "    elif is_nan(genre_df1_value):\n",
        "        # Find all values from the dictionary that appear in the string\n",
        "        genre_list = [value for value in genre_dict.values() if str(value) in genres_df2_value]\n",
        "\n",
        "        # Store each genres as subgenre\n",
        "        subgenre_list = value_to_correct_format(genres_df2_value)\n",
        "        return genre_list, subgenre_list\n",
        "\n",
        "    # If I have both\n",
        "    # I do as in the two previous cases, but I sum the two results\n",
        "    else:\n",
        "        genre_list_1 = value_to_correct_format(genre_df1_value)\n",
        "        subgenre_list_1 = value_to_correct_format(subgenre_df1_value)\n",
        "\n",
        "        genre_list_2 = [value for value in genre_dict.values() if str(value) in genres_df2_value]\n",
        "        subgenre_list_2 = value_to_correct_format(genres_df2_value)\n",
        "\n",
        "        # I sum them doing the union of two sets, so that I don't have duplicates\n",
        "        genre_list = list(set(genre_list_1) | set(genre_list_2))\n",
        "        subgenre_list = list(set(subgenre_list_1) | set(subgenre_list_2))\n",
        "\n",
        "        return genre_list, subgenre_list"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "sowPHgWHAqdL",
      "metadata": {
        "id": "sowPHgWHAqdL"
      },
      "source": [
        "**BUILD THE GRAPH**"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "vgE9TQoVgTOW",
      "metadata": {
        "id": "vgE9TQoVgTOW"
      },
      "source": [
        "Then I iterate over each row using the method `itertuples() `, and for each row I retrieve all the values, transform them with the functions I have just created, and finally build the triples that I will add to the graph.\n",
        "\n",
        "I will use `itertuples()` instead of `iterrows()`, since doing an iteration on tuples should be faster.\n",
        "\n",
        "For convinience I will do a simple check on the `['Number']` and `['Ranking']` columns to see if this row have both elements from the starting dataframes df1 and df2 (respectively), so I can easily see if I have each value from this row (or if some of them are `NaN`). By doing this I can easily recognise what triples I can add to my graph and what are useless since doesn't hold any value."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "q03XgUs4rnl2",
      "metadata": {
        "id": "q03XgUs4rnl2"
      },
      "outputs": [],
      "source": [
        "# Iterate over each row with index using itertuples()\n",
        "for row in df_merged.itertuples():\n",
        "    # Access the index value\n",
        "    album_value = row.Index[0]\n",
        "    artist_value = row.Index[1]\n",
        "\n",
        "    # Access the value of each column separately\n",
        "    # df1\n",
        "    number_value = row[1]\n",
        "    year_value = row[2]\n",
        "    genre_value = row[3]\n",
        "    subgenre_value = row[4]\n",
        "    # df2\n",
        "    ranking_value = row[5]\n",
        "    release_date_value = row[6]\n",
        "    genres_value = row[7]\n",
        "    descriptors_value = row[8]\n",
        "    average_rating_value = row[9]\n",
        "    number_of_ratings_value = row[10]\n",
        "    number_of_reviews_value = row[11]\n",
        "\n",
        "    ############################################################################\n",
        "\n",
        "    # Transform every information that I have obtained from this row and add them to the graph\n",
        "    # simple URIs + label\n",
        "    album = value_to_uri(album_value)\n",
        "    artist = value_to_uri(artist_value)\n",
        "    rdf_graph.add((album, df.artist, artist))\n",
        "    rdf_graph.add((album, RDF.type, df.Album))\n",
        "    rdf_graph.add((artist, RDF.type, df.Artist))\n",
        "    # label\n",
        "    album_label = Literal(album_value, datatype=XSD.string)\n",
        "    artist_label = Literal(artist_value, datatype=XSD.string)\n",
        "    rdf_graph.add((album, RDFS.label, album_label))\n",
        "    rdf_graph.add((artist, RDFS.label, artist_label))\n",
        "\n",
        "    # Date and Year\n",
        "    output_date, output_year = value_to_date_year(release_date_value, year_value)\n",
        "    if output_date:\n",
        "        rdf_graph.add((album, df.releaseDate, output_date))\n",
        "        rdf_graph.add((album, df.year, output_year))\n",
        "    else:\n",
        "        rdf_graph.add((album, df.year, output_year))\n",
        "\n",
        "    # Genre and Subgenre (which are URIs)\n",
        "    output_genre_list, output_subgenre_list = value_to_genre_subgenre(genre_value, subgenre_value, genres_value)\n",
        "    for genre in output_genre_list:\n",
        "        uri_genre = value_to_uri(genre)\n",
        "        literal_genre = value_to_tagged_string(genre)\n",
        "        rdf_graph.add((album, df.genre, uri_genre))\n",
        "        rdf_graph.add((uri_genre, RDF.type, df.Genre))\n",
        "        rdf_graph.add((uri_genre, RDFS.label, literal_genre))\n",
        "    for subgenre in output_subgenre_list:\n",
        "        uri_subgenre = value_to_uri(subgenre)\n",
        "        literal_subgenre = value_to_tagged_string(subgenre)\n",
        "        rdf_graph.add((album, df.subgenre, uri_subgenre))\n",
        "        rdf_graph.add((uri_subgenre, RDF.type, df.Genre))\n",
        "        rdf_graph.add((uri_subgenre, RDFS.label, literal_subgenre))\n",
        "\n",
        "    ############################################################################\n",
        "\n",
        "    # here the simpler datas\n",
        "    # df1\n",
        "    if not is_nan(number_value):\n",
        "        # rollingStones ranking\n",
        "        rs_Ranking = value_to_float(number_value)\n",
        "        rdf_graph.add((album, df.rsRanking, rs_Ranking))\n",
        "\n",
        "    # df2\n",
        "    if not is_nan(ranking_value):\n",
        "        # rateYourMusic ranking\n",
        "        rym_Ranking = value_to_float(ranking_value)\n",
        "        rdf_graph.add((album, df.rymRanking, rym_Ranking))\n",
        "        # descriptors\n",
        "        descriptors = value_to_tagged_string(descriptors_value)\n",
        "        rdf_graph.add((album, df.descriptors, descriptors))\n",
        "        # average rating\n",
        "        average_rating = value_to_float(average_rating_value)\n",
        "        rdf_graph.add((album, df.averageRating, average_rating))\n",
        "        # number of ratings\n",
        "        number_of_ratings = value_to_float(number_of_ratings_value)\n",
        "        rdf_graph.add((album, df.numberOfRatings, number_of_ratings))\n",
        "        # number of reviews\n",
        "        number_of_reviews = value_to_float(number_of_reviews_value)\n",
        "        rdf_graph.add((album, df.numberOfReviews, number_of_reviews))\n"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "8UHEfZjXjgHI",
      "metadata": {
        "id": "8UHEfZjXjgHI"
      },
      "source": [
        " I write the RDF graph as a file on disk."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "nnsVg6bRjcvI",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nnsVg6bRjcvI",
        "outputId": "82ee3345-50c3-4961-f5b7-e7d640051263"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<Graph identifier=Nfc54fab0d8f748199ae38c6283c308d6 (<class 'rdflib.graph.Graph'>)>"
            ]
          },
          "execution_count": 30,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# SAVE ON DISK\n",
        "rdf_graph.serialize(destination=\"./data/graph_01.ttl\")"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "b65cd562",
      "metadata": {
        "id": "b65cd562"
      },
      "source": [
        "## **`2. Integrate your data with DBpedia's`**"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "p3fyV2fQqWGQ",
      "metadata": {
        "id": "p3fyV2fQqWGQ"
      },
      "source": [
        "### 2.1 ) Link the URIs of my current graph with DBpedia"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "Od3h66ogqwwZ",
      "metadata": {
        "id": "Od3h66ogqwwZ"
      },
      "source": [
        "Before of extending my current graph with other information that I have in DBpedia, I want to link my URIs extracted from the dataframes to the mains one used in DBpedia, by employing `OWL.sameAs`\n",
        "\n",
        "Because my aim is to complete at the end the graph using the OWL reasoner, that will link by means of `OWL.sameAs` what I have in my initial graph, with the informations that I have in DBpedia. But still maintaining the original form and informations that I have in the dataframes, as base of the graph."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "NY5StSteWYKe",
      "metadata": {
        "id": "NY5StSteWYKe"
      },
      "outputs": [],
      "source": [
        "# Load the graph of the previous stage, without having to redo everything\n",
        "rdf_graph = Graph().parse(\"./data/graph_01.ttl\")\n",
        "\n",
        "df = Namespace('http://example.org/DataFrame/')\n",
        "rdf_graph.bind('df', 'http://example.org/DataFrame/')\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "NraCKRW6w7K0",
      "metadata": {
        "id": "NraCKRW6w7K0"
      },
      "outputs": [],
      "source": [
        "# Define the DBpedia namespaces\n",
        "dbp = Namespace('http://dbpedia.org/property/')\n",
        "dbo = Namespace('http://dbpedia.org/ontology/')\n",
        "dbr = Namespace('http://dbpedia.org/resource/')\n",
        "\n",
        "# Bind the prefixes to the graph\n",
        "rdf_graph.bind('dbp', 'http://dbpedia.org/property/')\n",
        "rdf_graph.bind('dbo', 'http://dbpedia.org/ontology/')\n",
        "rdf_graph.bind('dbr', 'http://dbpedia.org/resource/')\n",
        "\n",
        "# initialize the SPARQLWrapper endpoint\n",
        "sparql = SPARQLWrapper('https://dbpedia.org/sparql')"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "mwlWO_4xdWlB",
      "metadata": {
        "id": "mwlWO_4xdWlB"
      },
      "source": [
        "For each triple `(album, df.artist, artist)` I will try to metch the album and the artist in the following way\n",
        "\n",
        "**ALBUM:**\n",
        "\n",
        "*   `replace(\"http://example.org/DataFrame/\", \"http://dbpedia.org/resource/\")`, simply changing the URI and asking to DBpedia if the album exists\n",
        "*   `replace(\"http://example.org/DataFrame/\", \"http://dbpedia.org/resource/\") + '_(Album)`, simply changing the URI and adding the string \"_(Album)\" and asking to DBpedia if the album exists\n",
        "*   `SELECT ?album WHERE {{ ?album rdfs:label {album_label} }}`, searching every entity that have a label that is the same to our album name (with some simple string changes like `replace(\"_\", \" \")` )\n",
        "\n",
        "**ARTIST:**\n",
        "\n",
        "*   `replace(\"http://example.org/DataFrame/\", \"http://dbpedia.org/resource/\")`, simply changing the URI and asking to DBpedia if the artist exists\n",
        "*   `SELECT ?artist WHERE {{ ?album rdfs:label {artist_label} }}`, searching every entity that have a label that is the same to our artist name (with some simple string changes like `replace(\"_\", \" \")` )\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "Uc3GVxSPqVWq",
      "metadata": {
        "id": "Uc3GVxSPqVWq"
      },
      "outputs": [],
      "source": [
        "# Retrieve every s, o linked by df.artist, that are album and artist\n",
        "for album, artist in rdf_graph.subject_objects(df.artist):\n",
        "\n",
        "    # This will be useful later, to see what URIs use to link album with artist\n",
        "    flag_1, flag_2, flag_3, flag_4, flag_5 = False, False, False, False, False\n",
        "\n",
        "    # I replace my prefix with the one corresponding to DBpedia Resources | FLAG_1\n",
        "    dbr_album_1 = album.replace(\"http://example.org/DataFrame/\", \"http://dbpedia.org/resource/\")\n",
        "\n",
        "    sparql.setQuery(f\"ASK {{<{dbr_album_1}> a dbo:Album }}\")\n",
        "    sparql.setReturnFormat('json')\n",
        "    results = sparql.query().convert()\n",
        "    if results['boolean'] is True:\n",
        "        flag_1 = True\n",
        "        album_1 = dbr_album_1.replace(\"http://dbpedia.org/resource/\", \"\")\n",
        "        rdf_graph.add((dbr[album_1], OWL.sameAs, album))\n",
        "        rdf_graph.add((dbr[album_1], RDF.type, dbo.Album))\n",
        "\n",
        "    ############################################################################\n",
        "    # ending with '_(album)' | FLAG_2\n",
        "    dbr_album_2 = dbr_album_1 + '_(Album)'\n",
        "\n",
        "    sparql.setQuery(f\"ASK {{<{dbr_album_2}> a dbo:Album }}\")\n",
        "    sparql.setReturnFormat('json')\n",
        "    results = sparql.query().convert()\n",
        "    if results['boolean'] is True:\n",
        "        flag_2 = True\n",
        "        album_2 = dbr_album_2.replace(\"http://dbpedia.org/resource/\", \"\")\n",
        "        rdf_graph.add((dbr[album_2], OWL.sameAs, album))\n",
        "        rdf_graph.add((dbr[album_2], RDF.type, dbo.Album))\n",
        "\n",
        "    ############################################################################\n",
        "    # album linked with rdfs:label and a string | FLAG_3\n",
        "    album_label = album.replace(\"http://example.org/DataFrame/\", \"\")\n",
        "    album_label = album_label.replace(\"_\", \" \")\n",
        "    album_label = str(album_label)\n",
        "\n",
        "    sparql.setQuery(f'SELECT ?album WHERE {{ ?album rdfs:label \"{album_label}\" }}')\n",
        "    sparql.setReturnFormat(\"csv\")\n",
        "    results = sparql.query()\n",
        "    results = results.convert()\n",
        "    results = results.decode().split()\n",
        "    # remove the first result \"album\", which represent the bindings of the results\n",
        "    results.pop(0)\n",
        "    if len(results) > 0:\n",
        "        flag_3 = True\n",
        "        dbr_album_3_results = [] # useful later if I get more than one result\n",
        "        for album in results:\n",
        "            if 'http://dbpedia.org/resource/' in album:\n",
        "                try:\n",
        "                    dbr_album = album.replace(\"http://dbpedia.org/resource/\", \"\")\n",
        "                    dbr_album_3_results.append(dbr_album)\n",
        "                except:\n",
        "                    continue\n",
        "        for dbr_album_3 in dbr_album_3_results:\n",
        "            try:\n",
        "                rdf_graph.add((dbr[dbr_album_3], OWL.sameAs, album))\n",
        "                rdf_graph.add((dbr[dbr_album_3], RDF.type, dbo.Album))\n",
        "            except:\n",
        "                continue\n",
        "\n",
        "    ############################################################################\n",
        "    ############################################################################\n",
        "    ############################################################################\n",
        "    # Artist\n",
        "    dbr_artist = artist.replace(\"http://example.org/DataFrame/\", \"http://dbpedia.org/resource/\")\n",
        "\n",
        "    # dbo:Artist or dbo:MusicalArtist or dbo:Group | FLAG_4\n",
        "    sparql.setQuery(f\"ASK {{ {{<{dbr_artist}> a dbo:Artist }} UNION {{<{dbr_artist}> a dbo:MusicalArtist }} UNION {{<{dbr_artist}> a dbo:Group }} }}\")\n",
        "    sparql.setReturnFormat('json')\n",
        "    results = sparql.query().convert()\n",
        "    if results['boolean'] is True:\n",
        "        flag_3 = True\n",
        "        artist_name_1 = dbr_artist.replace(\"http://dbpedia.org/resource/\", \"\")\n",
        "        rdf_graph.add((dbr[artist_name_1], OWL.sameAs, artist))\n",
        "        rdf_graph.add((dbr[artist_name_1], RDF.type, dbo.Artist))\n",
        "\n",
        "    ############################################################################\n",
        "    # artist linked with rdfs:label and a string | FLAG_5\n",
        "    artist_label = artist.replace(\"http://example.org/DataFrame/\", \"\")\n",
        "    artist_label = artist_label.replace(\"_\", \" \")\n",
        "    artist_label = str(artist_label)\n",
        "\n",
        "    sparql.setQuery(f'SELECT ?artist WHERE {{ ?artist rdfs:label \"{artist_label}\" }}')\n",
        "    sparql.setReturnFormat(\"csv\")\n",
        "    results = sparql.query()\n",
        "    results = results.convert()\n",
        "    results = results.decode().split()\n",
        "    # remove the first result \"artist\", which represent the bindings of the results\n",
        "    results.pop(0)\n",
        "    if len(results) > 0:\n",
        "        flag_5 = True\n",
        "        dbr_artist_2_results = [] # useful later if I get more than one result\n",
        "        for artist in results:\n",
        "            if 'http://dbpedia.org/resource/' in artist :\n",
        "                try:\n",
        "                    dbr_artist = artist.replace(\"http://dbpedia.org/resource/\", \"\")\n",
        "                    dbr_artist_2_results.append(dbr_artist)\n",
        "                except:\n",
        "                    continue\n",
        "        for dbr_artist_2 in dbr_artist_2_results:\n",
        "            try:\n",
        "                rdf_graph.add((dbr[dbr_artist_2], OWL.sameAs, artist))\n",
        "                rdf_graph.add((dbr[dbr_artist_2], RDF.type, dbo.Artist))\n",
        "            except:\n",
        "                continue\n",
        "\n",
        "    ############################################################################\n",
        "    # add the relation album-artist (dbo:artist)\n",
        "    # I use the flags to see what is the correct option like 'name' or 'name_(album)'\n",
        "    if flag_1 is True and flag_4 is True:\n",
        "        rdf_graph.add((dbr[album_1], dbo.artist, dbr[artist_name_1]))\n",
        "    if flag_2 is True and flag_4 is True:\n",
        "        rdf_graph.add((dbr[album_2], dbo.artist, dbr[artist_name_1]))\n",
        "    if flag_3 is True and flag_4 is True:\n",
        "        for album_3 in dbr_album_3_results:\n",
        "            rdf_graph.add((dbr[album_3], dbo.artist, dbr[artist_name_1]))\n",
        "    if flag_1 is True and flag_5 is True:\n",
        "        for artist_name_2 in dbr_artist_2_results:\n",
        "            rdf_graph.add((dbr[album_1], dbo.artist, dbr[artist_name_2]))\n",
        "    if flag_2 is True and flag_5 is True:\n",
        "        for artist_name_2 in dbr_artist_2_results:\n",
        "            rdf_graph.add((dbr[album_2], dbo.artist, dbr[artist_name_2]))\n",
        "    if flag_3 is True and flag_5 is True:\n",
        "        for album_3 in dbr_album_3_results:\n",
        "            for artist_name_2 in dbr_artist_2_results:\n",
        "                rdf_graph.add((dbr[album_3], dbo.artist, dbr[artist_name_2]))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "GBUpAYxObWIl",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GBUpAYxObWIl",
        "outputId": "0d821ec4-c108-471f-c277-bb869fb2e482"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<Graph identifier=N3948b1d4a19d48eeb33615b6dd375aa0 (<class 'rdflib.graph.Graph'>)>"
            ]
          },
          "execution_count": 34,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# SAVE ON DISK\n",
        "rdf_graph.serialize(destination=\"./data/graph_02.ttl\")"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "64Ldj8SE8FVd",
      "metadata": {
        "id": "64Ldj8SE8FVd"
      },
      "source": [
        "---\n",
        "**Considerations:**\n",
        "\n",
        "Before of running this cell (in /data/graph_01.ttl) we have 79,884 lines.\n",
        "\n",
        "After this operations (in /data/graph_02.ttl) we have 92,271 lines.\n",
        "\n",
        "Our merged df (/data/merged_dataframe.csv)  have 5205 rows.\n",
        "\n",
        "My operations in this cells should add (if we suppose that everything is working perfectly, every URI is perfectly metched with the one in DBpedia) 5 triples for each pair `album, artist in rdf_graph.subject_objects(df.artist)`, that in total should be approximately 26,250 triples.\n",
        "\n",
        "So, more or less, we can say that 2500 pairs album, artist are been matched correctly, kind of half of our entities."
      ]
    },
    {
      "cell_type": "markdown",
      "id": "tR4-Hog7erDS",
      "metadata": {
        "id": "tR4-Hog7erDS"
      },
      "source": [
        "### 2.2 ) Enrich my informations with DBpedia"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "xn2RfRpueqxo",
      "metadata": {
        "id": "xn2RfRpueqxo"
      },
      "source": [
        "For searching the others possible URIs, that could be used in DBpedia, starting from the properties: `dbp:hometown`, `dbo:birthPlace`, `dbo:country`, `dbo:formerBandMember`, I used mainly **DBpedia lookup endpoint**.\n",
        "\n",
        "I have done query of the kind\n",
        "\n",
        "      ```\n",
        "      SELECT DISTINCT ?p\n",
        "      WHERE {\n",
        "        ?s a dbo:Band .\n",
        "        ?s dbo:formerBandMembers ?o.\n",
        "        ?s ?p ?o .\n",
        "      } LIMIT 10\n",
        "      ```\n",
        "\n",
        "So that I can easily see what other predicates can be used in the same situation of the one in witch I'm interested, using always proper domain.\n",
        "\n",
        "Because a predicate can have different meanings depending on the subject (like `dbo:hometown`), if I do like this (like domain and rage restrictions on the predicate) I can filter much better what predicate is used in a particular way\n",
        "\n",
        "And for kind of more specific/precise predicate I searched for some known entity, and seen on DBpedia by means of what predicates was linked with other entities"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "mtjisPTWgSeT",
      "metadata": {
        "id": "mtjisPTWgSeT"
      },
      "source": [
        "---\n",
        "**Band members**\n",
        "      \n",
        "      < band  ?p ?bandMember >\n",
        "      \n",
        "where ?p is: `dbo:bandMember`, `dbp:currentMembers`, `dbo:formerBandMember`, `dbp:pastMembers`\n",
        "\n",
        "---\n",
        "\n",
        "**City and country of birth of members**\n",
        "\n",
        "      < member  ?p1 ?cityOfBirth >\n",
        "      < ?cityOfBirth ?p2 ?countryOfBirth >\n",
        "\n",
        "where ?p1 is:  `dbo:hometown`, `dbp:hometown`, `dbo:birthPlace`, `dbp:birthPlace`, `dbp:origin`\n",
        "\n",
        "and ?p2 is: `dbo:country`, `dbp:country`, `dbo:subdivision / dbo:country` (the last for cities in the united states)\n",
        "\n",
        "---\n",
        "\n",
        "**Founding city and country of the band**\n",
        "\n",
        "      < band  ?p1 ?foundingCity >\n",
        "      < ?foundingCity ?p2 ?foundingCountry >\n",
        "\n",
        "where ?p1 is: `dbo:hometown`, `dbp:hometown`, `dbo:birthPlace`, `dbp:birthPlace`, `dbp:origin`\n",
        "\n",
        "and ?p2 is: `dbo:country`, `dbp:country`, `dbo:subdivision / dbo:country` (the last one for cities in the united states)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "_f454r8qmV5m",
      "metadata": {
        "id": "_f454r8qmV5m"
      },
      "source": [
        "I tried to search more options possible when i do the query on the DBpedia endpoint, but then in the graph I added only one option for the property that we have searched, to keep the graph more light as possible.\n",
        "\n",
        "Here I have a simple rapresentation, for example the band will be linked to the founding country with `dbp:hometown / dbo:country`, I will create a new URI that will be directly linked to what we want, like `df:foundingCountry` when we will build the ontology in the point 4."
      ]
    },
    {
      "cell_type": "markdown",
      "id": "nlhk0lbUVG2v",
      "metadata": {
        "id": "nlhk0lbUVG2v"
      },
      "source": [
        "The way that I found as the simplest for updating my graph is to run a `CONSTRUCT` query with the sparqlwrapper, save the result and then make the union between the resulting graph and my current graph."
      ]
    },
    {
      "cell_type": "markdown",
      "id": "5wIvE5-25kJt",
      "metadata": {
        "id": "5wIvE5-25kJt"
      },
      "source": [
        "Since I have a lot of data, to improve slightly the performances of the program, what I can do is to store every band that I analyze in a set and, before doing my queries, I check if the current band that I'm analyzing is already in this set. Since is useless checking members, birthPlace... and so on, for the same entity multiple times."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "ezUdPJNdWUUK",
      "metadata": {
        "id": "ezUdPJNdWUUK"
      },
      "outputs": [],
      "source": [
        "# Load the graph of the previous stage, without having to redo everything\n",
        "rdf_graph = Graph().parse(\"./data/graph_02.ttl\")\n",
        "\n",
        "df = Namespace('http://example.org/DataFrame/')\n",
        "dbp = Namespace('http://dbpedia.org/property/')\n",
        "dbo = Namespace('http://dbpedia.org/ontology/')\n",
        "dbr = Namespace('http://dbpedia.org/resource/')\n",
        "rdf_graph.bind('df', 'http://example.org/DataFrame/')\n",
        "rdf_graph.bind('dbp', 'http://dbpedia.org/property/')\n",
        "rdf_graph.bind('dbo', 'http://dbpedia.org/ontology/')\n",
        "rdf_graph.bind('dbr', 'http://dbpedia.org/resource/')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "sff-HhYMIjiJ",
      "metadata": {
        "id": "sff-HhYMIjiJ"
      },
      "outputs": [],
      "source": [
        "already_analized_bands = set()\n",
        "\n",
        "# Retrieve every s, o linked by df.artist, that are album and artist\n",
        "for band in rdf_graph.subjects(RDF.type, dbo.Artist):\n",
        "    if band not in already_analized_bands:\n",
        "        # Band members and former band members\n",
        "        query = f\"\"\"\n",
        "        CONSTRUCT {{\n",
        "            <{band}> a dbo:Band;\n",
        "                     dbo:bandMember ?bandMember;\n",
        "                     dbo:formerBandMember ?formerBandMember.\n",
        "        }}\n",
        "\n",
        "        WHERE {{\n",
        "            <{band}> a dbo:Band.\n",
        "\n",
        "            {{ <{band}> dbo:bandMember ?bandMember }}\n",
        "            UNION\n",
        "            {{ <{band}> dbp:currentMembers ?bandMember }}\n",
        "\n",
        "            OPTIONAL {{\n",
        "                {{ <{band}> dbo:formerBandMember ?formerBandMember }}\n",
        "                UNION\n",
        "                {{ <{band}> dbp:pastMembers ?formerBandMember }}\n",
        "            }}\n",
        "        }}\n",
        "        \"\"\"\n",
        "\n",
        "        sparql.setQuery(query)\n",
        "        sparql.setReturnFormat(\"xml\")\n",
        "        results = sparql.query().convert()\n",
        "        rdf_graph += results\n",
        "\n",
        "        ########################################################################\n",
        "        # founding city and country for the bands\n",
        "        query = f\"\"\"\n",
        "        CONSTRUCT {{\n",
        "            <{band}> dbp:hometown ?foundingCity.\n",
        "            ?foundingCity dbo:country ?foundingCountry.\n",
        "        }}\n",
        "\n",
        "        WHERE {{\n",
        "            <{band}> a dbo:Band.\n",
        "            ?foundingCity a dbo:City.\n",
        "\n",
        "            {{ <{band}> dbo:hometown ?foundingCity }}\n",
        "            UNION\n",
        "            {{ <{band}> dbp:hometown ?foundingCity }}\n",
        "            UNION\n",
        "            {{ <{band}> dbo:birthPlace ?foundingCity }}\n",
        "            UNION\n",
        "            {{ <{band}> dbp:birthPlace ?foundingCity }}\n",
        "            UNION\n",
        "            {{ <{band}> dbp:origin ?foundingCity }}\n",
        "\n",
        "            OPTIONAL {{\n",
        "                ?foundingCountry a dbo:Country\n",
        "                {{ ?foundingCity dbo:country ?foundingCountry }}\n",
        "                UNION\n",
        "                {{ ?foundingCity dbp:country ?foundingCountry }}\n",
        "                UNION\n",
        "                {{ ?foundingCity dbo:subdivision / dbo:country ?foundingCountry }}\n",
        "            }}\n",
        "        }}\n",
        "        \"\"\"\n",
        "        sparql.setQuery(query)\n",
        "        sparql.setReturnFormat(\"xml\")\n",
        "        results = sparql.query().convert()\n",
        "        rdf_graph += results\n",
        "\n",
        "        # add the current band to the set of the analyzed bands\n",
        "        already_analized_bands.add(band)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "Hy_I6Ot7n3V2",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Hy_I6Ot7n3V2",
        "outputId": "f92505b0-f003-4ad6-a4f4-aa27d5185f07"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<Graph identifier=Nf0ca5e3c78fe42acaf5189d5b992e8e3 (<class 'rdflib.graph.Graph'>)>"
            ]
          },
          "execution_count": 38,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# SAVE ON DISK\n",
        "rdf_graph.serialize(destination=\"./data/graph_03.ttl\")"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "qfDSaLG7owj8",
      "metadata": {
        "id": "qfDSaLG7owj8"
      },
      "source": [
        "Now I do kind of the same, but for the members, so I do a for loop using `rdf_graph.objects(None, dbo.bandMember)`, that gives an iteration on every object of this kind, that are the members of the bands. I are helping us to do a simpler iteration, finding directly the members, I am not searching anymore via the bands, but I use the informations on my graph."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "h3tSsASpDRjZ",
      "metadata": {
        "id": "h3tSsASpDRjZ"
      },
      "outputs": [],
      "source": [
        "# Load the graph of the previous stage, without having to redo everything\n",
        "rdf_graph = Graph().parse(\"./data/graph_03.ttl\")\n",
        "\n",
        "df = Namespace('http://example.org/DataFrame/')\n",
        "dbp = Namespace('http://dbpedia.org/property/')\n",
        "dbo = Namespace('http://dbpedia.org/ontology/')\n",
        "dbr = Namespace('http://dbpedia.org/resource/')\n",
        "rdf_graph.bind('df', 'http://example.org/DataFrame/')\n",
        "rdf_graph.bind('dbp', 'http://dbpedia.org/property/')\n",
        "rdf_graph.bind('dbo', 'http://dbpedia.org/ontology/')\n",
        "rdf_graph.bind('dbr', 'http://dbpedia.org/resource/')\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "MP41pmE8_M6l",
      "metadata": {
        "id": "MP41pmE8_M6l"
      },
      "outputs": [],
      "source": [
        "already_analized_members = set()\n",
        "\n",
        "# member is every entity that iis object of a triple that have as predicate dbo.bandMember\n",
        "# (based on how I have build the graph in the previous step)\n",
        "for member in rdf_graph.objects(None, dbo.bandMember):\n",
        "    if member not in already_analized_members:\n",
        "        # I don't check for literals or something else\n",
        "        if \"http://dbpedia.org\" in member:\n",
        "        # city and country of birth for the members\n",
        "            query = f\"\"\"\n",
        "            CONSTRUCT {{\n",
        "                <{member}> dbo:birthPlace ?cityOfBirth.\n",
        "                ?cityOfBirth dbo:country ?countryOfBirth.\n",
        "            }}\n",
        "\n",
        "            WHERE {{\n",
        "                <{member}> a dbo:Person.\n",
        "                ?cityOfBirth a dbo:City.\n",
        "\n",
        "                {{ <{member}> dbo:hometown ?cityOfBirth }}\n",
        "                UNION\n",
        "                {{ <{member}> dbp:hometown ?cityOfBirth }}\n",
        "                UNION\n",
        "                {{ <{member}> dbo:birthPlace ?cityOfBirth }}\n",
        "                UNION\n",
        "                {{ <{member}> dbp:birthPlace ?cityOfBirth }}\n",
        "                UNION\n",
        "                {{ <{member}> dbp:origin ?cityOfBirth }}\n",
        "\n",
        "                OPTIONAL {{\n",
        "                    ?countryOfBirth a dbo:Country\n",
        "\n",
        "                    {{ ?cityOfBirth dbo:country ?countryOfBirth }}\n",
        "                    UNION\n",
        "                    {{ ?cityOfBirth dbp:country ?countryOfBirth }}\n",
        "                    UNION\n",
        "                    {{ ?cityOfBirth dbo:subdivision / dbo:country ?countryOfBirth }}\n",
        "                }}\n",
        "            }}\n",
        "            \"\"\"\n",
        "\n",
        "            sparql.setQuery(query)\n",
        "            sparql.setReturnFormat(\"xml\")\n",
        "            results = sparql.query().convert()\n",
        "            rdf_graph += results\n",
        "\n",
        "            # add the current member to the set of the analyzed members\n",
        "            already_analized_members.add(member)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "vbkGExq_VoH8",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vbkGExq_VoH8",
        "outputId": "79accafe-403e-4b5a-fa84-e92d01d76adb"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<Graph identifier=N9bfe2aafc2ee439897a674d258c318ef (<class 'rdflib.graph.Graph'>)>"
            ]
          },
          "execution_count": 41,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# SAVE ON DISK\n",
        "rdf_graph.serialize(destination=\"./data/graph_04.ttl\")"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "6c66fbc6",
      "metadata": {
        "id": "6c66fbc6"
      },
      "source": [
        "## **`3. Gathering Information from unstructured data`**\n"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "kHL-FevA-u9F",
      "metadata": {
        "id": "kHL-FevA-u9F"
      },
      "source": [
        "What I will do in steps:\n",
        "1. get the article from Wikipedia and extract the summary, using [Wikipedia-api](https://pypi.org/project/Wikipedia-API/)\n",
        "      ```\n",
        "    wiki_wiki = wikipediaapi.Wikipedia('en')\n",
        "    page_py = wiki_wiki.page('Kurt Cobain')\n",
        "\n",
        "    if page_py.exists():\n",
        "        summary = page_py.summary[:200]\n",
        "        print(\"Page - Summary: %s\" % summary)\n",
        "    else:\n",
        "        print(\"Page does not exist.\")\n",
        "      ```\n",
        "\n",
        "    \n",
        "2. extract the band member role (guitarist, bassist, keyboardist, singer, drummer, etc) using simple pattern matchig with my set, if some artists have multiple roles, I add them all,\n",
        "3. add this information to the KG."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "87b18550",
      "metadata": {
        "id": "87b18550"
      },
      "outputs": [],
      "source": [
        "# Load the graph of the previous stage, without having to redo everything\n",
        "rdf_graph = Graph().parse(\"./data/graph_04.ttl\")\n",
        "\n",
        "df = Namespace('http://example.org/DataFrame/')\n",
        "dbp = Namespace('http://dbpedia.org/property/')\n",
        "dbo = Namespace('http://dbpedia.org/ontology/')\n",
        "dbr = Namespace('http://dbpedia.org/resource/')\n",
        "rdf_graph.bind('df', 'http://example.org/DataFrame/')\n",
        "rdf_graph.bind('dbp', 'http://dbpedia.org/property/')\n",
        "rdf_graph.bind('dbo', 'http://dbpedia.org/ontology/')\n",
        "rdf_graph.bind('dbr', 'http://dbpedia.org/resource/')\n"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "lVo498Oc1Dmw",
      "metadata": {
        "id": "lVo498Oc1Dmw"
      },
      "source": [
        "Here I have generated a simple set for pattern matching, and the role that I will use."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "cOeBE6EYn-7z",
      "metadata": {
        "id": "cOeBE6EYn-7z"
      },
      "outputs": [],
      "source": [
        "# We create a new namespace for the inferences that we will add\n",
        "wiki = Namespace('http://example.org/WikipediaApi/')\n",
        "rdf_graph.bind('wiki', 'http://example.org/WikipediaApi/')\n",
        "\n",
        "wiki['role']\n",
        "\n",
        "# set with possible roles\n",
        "roles_set = {\n",
        "    \"Lead Vocalist\",\"Singer\", \"Backing Vocalist\", \"Vocalist\", \"Lead Guitarist\",\n",
        "     \"Guitarist\", \"Rhythm Guitarist\", \"Bassist\", \"Drummer\", \"Keyboardist\",\n",
        "     \"Pianist\", \"Synthesizer Player\", \"Saxophonist\", \"Trumpeter\", \"Trombonist\",\n",
        "     \"Violinist\", \"Cellist\", \"Flutist\", \"Clarinetist\", \"Percussionist\",\n",
        "     \"Director\", \"Sampler\", \"Programmer\", \"Beatboxer\", \"Backup\",\n",
        "     \"Session Musician\", \"Songwriter\", \"Composer\", \"Sound Engineer\",\n",
        "     \"Musical Director\", \"DJ\", \"Conductor\", \"Producer\", \"Roadie\", \"Technician\"\n",
        "}\n"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "Hdb9V1AA1S0K",
      "metadata": {
        "id": "Hdb9V1AA1S0K"
      },
      "source": [
        "Now we iterate over each band member, and with some simple string operation we transform the URI to something that can be easily searched on Wikipedia. Then finally, after we have extracted the summary, we do an operation of simple pattern matching with the set of roles.\n",
        "\n",
        "We will add name in the graph in the following way:\n",
        "\n",
        "    < dbr_member wiki:role wiki:role_name >"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "Flwo_xk61SZY",
      "metadata": {
        "id": "Flwo_xk61SZY"
      },
      "outputs": [],
      "source": [
        "already_analized_members = set()\n",
        "\n",
        "# SETTING FOR WIKIPEDIAAPI\n",
        "# Specify the user agent\n",
        "user_agent = \"krr2Project/1.0 (alessandro.ghiotto01@universitadipavia.it) Python/3.10.12 WikipediaAPI/0.6.0\"\n",
        "\n",
        "# Create Wikipedia object with the specified user agent\n",
        "wiki_wiki = wikipediaapi.Wikipedia(\n",
        "    language='en',\n",
        "    extract_format=wikipediaapi.ExtractFormat.WIKI,\n",
        "    user_agent=user_agent\n",
        ")\n",
        "\n",
        "# iterate over the members of the bands\n",
        "for member in rdf_graph.objects(None, dbo.bandMember):\n",
        "    if member not in already_analized_members:\n",
        "        # str_member is the string that I will use to search in Wikipedia\n",
        "        str_member = member.replace('http://dbpedia.org/resource/', '')\n",
        "        str_member = str_member.replace('_', ' ')\n",
        "        str_member = str(str_member)\n",
        "\n",
        "        try:\n",
        "            page_py = wiki_wiki.page(str_member)\n",
        "            if page_py.exists():\n",
        "                # If I find the page, nice!\n",
        "                # check the summary and if the roles-keywords appear in it\n",
        "                summary = page_py.summary[:200]\n",
        "\n",
        "                # I check if one (or more) of my roles are present in the summary\n",
        "                # And each one that appears, we append it to a list\n",
        "                found_roles = []\n",
        "                for element in roles_set:\n",
        "                    if element in summary:\n",
        "                        found_roles.append(element)\n",
        "\n",
        "                for role_name in found_roles:\n",
        "                    # Here we add the triples\n",
        "                    rdf_graph.add((member, wiki['role'], wiki[role_name]))\n",
        "        except:\n",
        "            # Handle any API response errors (Do nothing and proceed to the next iteration)\n",
        "            pass\n",
        "\n",
        "        # add the current member to the set of the analyzed members\n",
        "        already_analized_members.add(member)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "7HC-bF0g8mtt",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7HC-bF0g8mtt",
        "outputId": "946062f7-976e-4da1-a6a9-af7a228080b1"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<Graph identifier=Nd80d70ee1cf44ee291e1e8bac1cbfcbe (<class 'rdflib.graph.Graph'>)>"
            ]
          },
          "execution_count": 45,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# SAVE ON DISK\n",
        "rdf_graph.serialize(destination=\"./data/graph_05.ttl\")"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "03677112-ebe7-4ec8-9bfa-16380b6fcb60",
      "metadata": {
        "id": "03677112-ebe7-4ec8-9bfa-16380b6fcb60"
      },
      "source": [
        "## **`4. Include a small ontology and a  genres taxonomy`**\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "CCoSB3AVyaJz",
      "metadata": {
        "id": "CCoSB3AVyaJz"
      },
      "source": [
        "### 4.1 ) Domain and range restictions"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "R7pkOcULyTQD",
      "metadata": {
        "id": "R7pkOcULyTQD"
      },
      "source": [
        "Here we Add the domain and range for each property you use. This will be needed later when I will apply the reasoner for inferring the types of the entities on my KG.\n",
        "\n",
        "Since now I'm working only my current graph (`rdf_graph`), I can use the RDFLib, which comes with an implementation of the [SPARQL 1.1 Query](https://www.w3.org/TR/sparql11-query/) and [SPARQL 1.1 Update](https://www.w3.org/TR/sparql11-update/) query languages.\n",
        "This is useful since I need to run SPARQL queries on my RDFLib's `Graph` objects.\n",
        "So now I can run SPARQL Update (insert, delete) queries that can be run using the method `update` rather than `query`; that is a bit more convinient than saving results of a `construct` query as graph, and then make the union with my graph."
      ]
    },
    {
      "cell_type": "markdown",
      "id": "uoEI_99D4bU-",
      "metadata": {
        "id": "uoEI_99D4bU-"
      },
      "source": [
        "      # ------------------------------------------------- #\n",
        "      |                      SCHEMA                       |\n",
        "      # ------------------------------------------------- #\n",
        "      ```\n",
        "      df:artist rdfs:domain df:Album ; rdfs:range df:Artist .\n",
        "      df:genre rdfs:domain df:Album ; rdfs:range df:Genre .\n",
        "      df:subgenre rdfs:domain df:Album ; rdfs:range df:Subgenre .\n",
        "      dbo:artist rdfs:domain dbo:Album ; rdfs:range dbo:Artist .\n",
        "      dbo:bandMember rdfs:domain dbo:Band ; rdfs:range dbo:BandMember .\n",
        "      dbp:hometown rdfs:domain dbo:Band ; rdfs:range dbo:City .\n",
        "      dbo:birthPlace rdfs:domain dbo:Person ; rdfs:range dbo:City .\n",
        "      dbo:country rdfs:domain dbo:City  ; rdfs:range dbo:Country .\n",
        "      wiki:role rdfs:domain dbo:BandMember ; rdfs:range wiki:MusicRole.\n",
        "      ```\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "DpjmkuVa1H75",
      "metadata": {
        "id": "DpjmkuVa1H75"
      },
      "outputs": [],
      "source": [
        "# Load the graph of the previous stage, without having to redo everything\n",
        "rdf_graph = Graph().parse(\"./data/graph_05.ttl\")\n",
        "\n",
        "df = Namespace('http://example.org/DataFrame/')\n",
        "dbp = Namespace('http://dbpedia.org/property/')\n",
        "dbo = Namespace('http://dbpedia.org/ontology/')\n",
        "dbr = Namespace('http://dbpedia.org/resource/')\n",
        "wiki = Namespace('http://example.org/WikipediaApi/')\n",
        "rdf_graph.bind('df', 'http://example.org/DataFrame/')\n",
        "rdf_graph.bind('dbp', 'http://dbpedia.org/property/')\n",
        "rdf_graph.bind('dbo', 'http://dbpedia.org/ontology/')\n",
        "rdf_graph.bind('dbr', 'http://dbpedia.org/resource/')\n",
        "rdf_graph.bind('wiki', 'http://example.org/WikipediaApi/')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "eVyqpMFB9S5u",
      "metadata": {
        "id": "eVyqpMFB9S5u"
      },
      "outputs": [],
      "source": [
        "# SCHEMA\n",
        "schema = (\n",
        "    (df.artist, RDFS.domain, df.Album),         (df.artist, RDFS.range, df.Artist),\n",
        "    (df.genre, RDFS.domain, df.Album),          (df.genre, RDFS.range, df.Genre),\n",
        "    (df.subgenre, RDFS.domain, df.Album),       (df.subgenre, RDFS.range, df.Subgenre),\n",
        "    (dbo.artist, RDFS.domain, dbo.Album),       (dbo.artist, RDFS.range, dbo.Artist),\n",
        "    (dbo.bandMember, RDFS.domain, dbo.Band),    (dbo.bandMember, RDFS.range, dbo.BandMember),\n",
        "    (dbp.hometown, RDFS.domain, dbo.Band),      (dbp.hometown, RDFS.range, dbo.City),\n",
        "    (dbo.birthPlace, RDFS.domain, dbo.Person),  (dbo.birthPlace, RDFS.range, dbo.City),\n",
        "    (dbo.country, RDFS.domain, dbo.City),       (dbo.country, RDFS.range, dbo.Country),\n",
        "    (wiki.role, RDFS.domain, dbo.BandMember),   (wiki.role, RDFS.range, wiki.MusicRole),\n",
        ")\n",
        "\n",
        "for triple in schema:\n",
        "    rdf_graph.add(triple)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "8TXs9XxF9gqr",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8TXs9XxF9gqr",
        "outputId": "6814a934-379c-4d9c-989e-cc9434d2a1b6"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<Graph identifier=Nbd58458fdc194fbfb740c767c4162f36 (<class 'rdflib.graph.Graph'>)>"
            ]
          },
          "execution_count": 48,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# SAVE ON DISK\n",
        "rdf_graph.serialize(destination=\"./data/graph_06.ttl\")"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "qsgol_dFyj4d",
      "metadata": {
        "id": "qsgol_dFyj4d"
      },
      "source": [
        "### 4.2 ) Genres taxonomy"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "06f89ab2-6eb5-4d14-9366-59f4da8603b2",
      "metadata": {
        "id": "06f89ab2-6eb5-4d14-9366-59f4da8603b2"
      },
      "source": [
        "Extract a genres taxonomy from the dataset and I include it in my RDF KG. This will be useful later for inferring genres from subgenres.\n",
        "\n",
        "In my graph I have something of this kind:\n",
        "\n",
        "      df:Album_name df:genre df:genre_name;\n",
        "                    df:subgenre df:subgenre_name.\n",
        "\n",
        "My idea is to retrieve each genre and subgenre linked with every album that I have stored in my KG. Then for every possible pair genre-subgenre (linked to the same album) I add the triple `(genre, RDFS.subClassOf, subgenre)` if the string corresponding to the genre is contained in the string corresponding to the subgenre.\n",
        "\n",
        "      df:#1_Record df:genre df:Folk,\n",
        "                            df:Pop,\n",
        "                            df:Rock;\n",
        "                   df:subgenre df:Folk_Pop,\n",
        "                               df:Power_Pop.\n",
        "\n",
        "For example in this situation I get as genres `df:Folk`, `df:Pop`, `df:Rock`, and as subgenres `df:Folk_Pop` and `df:Power_Pop`.\n",
        "\n",
        "I want to add for example `df:Folk_Pop rdfs:subClassOf df:Folk`, but not `df:Folk_Pop rdfs:subClassOf df:Rock`.\n",
        "\n",
        "I use sets, so I don't get duplicates."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "ysDpZLaM2jHS",
      "metadata": {
        "id": "ysDpZLaM2jHS"
      },
      "outputs": [],
      "source": [
        "# Load the graph of the previous stage, without having to redo everything\n",
        "rdf_graph = Graph().parse(\"./data/graph_06.ttl\")\n",
        "\n",
        "df = Namespace('http://example.org/DataFrame/')\n",
        "dbp = Namespace('http://dbpedia.org/property/')\n",
        "dbo = Namespace('http://dbpedia.org/ontology/')\n",
        "dbr = Namespace('http://dbpedia.org/resource/')\n",
        "wiki = Namespace('http://example.org/WikipediaApi/')\n",
        "rdf_graph.bind('df', 'http://example.org/DataFrame/')\n",
        "rdf_graph.bind('dbp', 'http://dbpedia.org/property/')\n",
        "rdf_graph.bind('dbo', 'http://dbpedia.org/ontology/')\n",
        "rdf_graph.bind('dbr', 'http://dbpedia.org/resource/')\n",
        "rdf_graph.bind('wiki', 'http://example.org/WikipediaApi/')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "FQQYUCZSiZe4",
      "metadata": {
        "id": "FQQYUCZSiZe4"
      },
      "outputs": [],
      "source": [
        "# create two sets in witch I will store my genres and subgenres\n",
        "genres_set = set()\n",
        "subgenres_set = set()\n",
        "\n",
        "# I do an iteration over the Albums\n",
        "for album in rdf_graph.subjects(df.genre, None):\n",
        "\n",
        "    query = f\"\"\"\n",
        "        SELECT ?genre ?subgenre\n",
        "        WHERE {{\n",
        "            <{album}> df:genre ?genre ;\n",
        "                      df:subgenre ?subgenre.\n",
        "        }}\n",
        "    \"\"\"\n",
        "    results = rdf_graph.query(query)\n",
        "\n",
        "    for res in results:\n",
        "        genres_set.add(res['genre'])\n",
        "        subgenres_set.add(res['subgenre'])\n",
        "\n",
        "\n",
        "def has_only_one_word(input_string):\n",
        "    words = input_string.split()\n",
        "    return len(words) == 1\n",
        "\n",
        "# now I check each possible pair genre-subgenre\n",
        "for genre in genres_set:\n",
        "    genre_str = genre.replace('http://example.org/DataFrame/', '')\n",
        "    genre_str = genre_str.replace('_', ' ')\n",
        "\n",
        "    if has_only_one_word(genre_str):\n",
        "        for subgenre in subgenres_set:\n",
        "            subgenre_str = subgenre.replace('http://example.org/DataFrame/', '')\n",
        "            subgenre_str = subgenre_str.replace('_', ' ')\n",
        "\n",
        "            if genre_str in subgenre_str:\n",
        "                rdf_graph.add((subgenre, RDFS.subClassOf, genre))\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "HElfjEJhpcSF",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HElfjEJhpcSF",
        "outputId": "c37fc902-b476-4f5c-b2f8-59c97c72937a"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<Graph identifier=N0429d6fed8bd41aa9c9d014fbcca6a1c (<class 'rdflib.graph.Graph'>)>"
            ]
          },
          "execution_count": 51,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# SAVE ON DISK\n",
        "rdf_graph.serialize(destination=\"./data/graph_07.ttl\")"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "0GZoexQpjc_N",
      "metadata": {
        "id": "0GZoexQpjc_N"
      },
      "source": [
        "I add that whenever I have `album_name df:genre genre_name`, I add `album_name rdf:type genre_name`"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "cW5LtCPapXKX",
      "metadata": {
        "id": "cW5LtCPapXKX"
      },
      "outputs": [],
      "source": [
        "# Load the graph of the previous stage, without having to redo everything\n",
        "rdf_graph = Graph().parse(\"./data/graph_07.ttl\")\n",
        "\n",
        "df = Namespace('http://example.org/DataFrame/')\n",
        "dbp = Namespace('http://dbpedia.org/property/')\n",
        "dbo = Namespace('http://dbpedia.org/ontology/')\n",
        "dbr = Namespace('http://dbpedia.org/resource/')\n",
        "wiki = Namespace('http://example.org/WikipediaApi/')\n",
        "rdf_graph.bind('df', 'http://example.org/DataFrame/')\n",
        "rdf_graph.bind('dbp', 'http://dbpedia.org/property/')\n",
        "rdf_graph.bind('dbo', 'http://dbpedia.org/ontology/')\n",
        "rdf_graph.bind('dbr', 'http://dbpedia.org/resource/')\n",
        "rdf_graph.bind('wiki', 'http://example.org/WikipediaApi/')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "rYE2d6AAjdm1",
      "metadata": {
        "id": "rYE2d6AAjdm1"
      },
      "outputs": [],
      "source": [
        "# UPDATE QUERY\n",
        "\n",
        "rdf_graph.update(\"insert { ?album rdf:type ?genre } WHERE { {?album df:genre ?genre} UNION {?album df:subgenre ?genre} }\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "-0PRA5mLqDIK",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-0PRA5mLqDIK",
        "outputId": "0943cb49-372e-45e1-d8f9-40566bfbebd3"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<Graph identifier=N599ab1fae12145c1aa07d448b91f9f57 (<class 'rdflib.graph.Graph'>)>"
            ]
          },
          "execution_count": 54,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# SAVE ON DISK\n",
        "rdf_graph.serialize(destination=\"./data/graph_08.ttl\")"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "c84075ec",
      "metadata": {
        "id": "c84075ec"
      },
      "source": [
        "## **`5. Materialize inferences`**"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "pV2YVmShrtCN",
      "metadata": {
        "id": "pV2YVmShrtCN"
      },
      "source": [
        "Now I execute some SPARQL update query which, at each step simulate some kind of inference pattern, for example from RDFS, OWLR, and some new rule that I can add.\n",
        "\n",
        "\n",
        "*   `domain` and `range`\n",
        "*   `subClassOf` and `subPropertyOf`\n",
        "*   OWL features, like `owl:sameAs`...\n",
        "*   if I have that `[album] rdf:type [genre1]` and `[genre1] rdfs:subClassOf [genre2]`, add the triple `[album] rdf:type [genre2]`\n",
        "*   for bands:  dbp:hometown ⊑ `mi:foundingCity`\n",
        "*   for bands:  dbp:hometown / dbo:country ⊑ `mi:foundingCountry`\n",
        "*   for people:  dbo:birthPlace ⊑ `mi:cityOfBirth`\n",
        "*   for people:  dbo:birthPlace / dbo:country ⊑ `mi:countryOfBirth`\n",
        "*   `mi:bornIn`: transitive closure of dbp:hometown, dbo:birthPlace and dbo:country\n",
        "*   `mi:genre`: genre of the artist inferred through the genre of the albums\n",
        "*   ` mi:subgenre`: subgenre of the artist inferred through the subgenre of the album\n",
        "*   `mi:subgenre`: subproperty of `mi:genre`\n",
        "\n",
        "I create this new property like `mi:genre`, that are like subproperty of many other more specific property, so it will be easier to query later the graph."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "b30047dc",
      "metadata": {
        "id": "b30047dc"
      },
      "outputs": [],
      "source": [
        "# Load the graph of the previous stage, without having to redo everything\n",
        "rdf_graph = Graph().parse(\"./data/graph_08.ttl\")\n",
        "\n",
        "df = Namespace('http://example.org/DataFrame/')\n",
        "dbp = Namespace('http://dbpedia.org/property/')\n",
        "dbo = Namespace('http://dbpedia.org/ontology/')\n",
        "dbr = Namespace('http://dbpedia.org/resource/')\n",
        "wiki = Namespace('http://example.org/WikipediaApi/')\n",
        "rdf_graph.bind('df', 'http://example.org/DataFrame/')\n",
        "rdf_graph.bind('dbp', 'http://dbpedia.org/property/')\n",
        "rdf_graph.bind('dbo', 'http://dbpedia.org/ontology/')\n",
        "rdf_graph.bind('dbr', 'http://dbpedia.org/resource/')\n",
        "rdf_graph.bind('wiki', 'http://example.org/WikipediaApi/')\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "QyeT_Dibwdb8",
      "metadata": {
        "id": "QyeT_Dibwdb8"
      },
      "outputs": [],
      "source": [
        "# We add a new URI so that we can easily distinguish this new materialized inferences\n",
        "mi = Namespace('http://example.org/MaterializeInference/')\n",
        "rdf_graph.bind('mi', 'http://example.org/MaterializeInference/')"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "jwAGpNNCsJU3",
      "metadata": {
        "id": "jwAGpNNCsJU3"
      },
      "source": [
        "A simple way to add new inferences is using the RDFS and OWL2 vocabulary with some property of the relations and property of the classes, for example we will use: `rdfs:subClassOf` and `rdfs:subPropertyOf`\n",
        "      \n",
        "      # ------------------------------------------------- #\n",
        "      |                      SCHEMA                       |\n",
        "      # ------------------------------------------------- #\n",
        "      ```\n",
        "      dbp:hometown rdfs:subPropertyOf mi:foundingCity.\n",
        "      dbo:birthPlace rdfs:subPropertyOf mi:cityOfBirth.\n",
        "      dbp:hometown rdfs:subPropertyOf mi:bornIn.\n",
        "      dbo:birthPlace rdfs:subPropertyOf mi:bornIn.\n",
        "      dbo:country rdfs:subPropertyOf mi:bornIn.\n",
        "      mi:bornIn rdf:type owl:TransitiveProperty.\n",
        "      mi:subgenre rdfs:subPropertyOf mi:genre.\n",
        "      ```\n",
        "I can easily materialize the inferences that are derived from those property using the RDFS_OWL2 reasoner\n",
        "`owlrl.DeductiveClosure(owlrl.RDFS_OWLRL_Semantics).expand(rdf_graph)`"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "H9FEDmgFOiFA",
      "metadata": {
        "id": "H9FEDmgFOiFA"
      },
      "source": [
        "First I add this triples and then I apply the RDFS_OWLRL Reasoner"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "dECT4K_tOhCk",
      "metadata": {
        "id": "dECT4K_tOhCk"
      },
      "outputs": [],
      "source": [
        "schema = (\n",
        "    (dbp.hometown, RDFS.subPropertyOf, mi.foundingCity),\n",
        "    (dbo.birthPlace, RDFS.subPropertyOf, mi.cityOfBirth),\n",
        "    (dbp.hometown, RDFS.subPropertyOf, mi.bornIn),\n",
        "    (dbo.birthPlace, RDFS.subPropertyOf, mi.bornIn),\n",
        "    (dbo.country, RDFS.subPropertyOf, mi.bornIn),\n",
        "    (mi.bornIn, RDF.type, OWL.TransitiveProperty),\n",
        "    (mi.subgenre, RDFS.subPropertyOf, mi.genre)\n",
        ")\n",
        "\n",
        "for triple in schema:\n",
        "    rdf_graph.add(triple)\n",
        "\n",
        "# domain and range, subClassOf and subPropertyOf, OWL features\n",
        "owlrl.DeductiveClosure(owlrl.RDFS_OWLRL_Semantics).expand(rdf_graph)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "5McX5-iIyxP9",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5McX5-iIyxP9",
        "outputId": "8c8bc534-751f-4c01-befb-ad0d5064c503"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<Graph identifier=N0a918b8fdb9c4ed4a302d72692d05e57 (<class 'rdflib.graph.Graph'>)>"
            ]
          },
          "execution_count": 58,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# SAVE ON DISK\n",
        "rdf_graph.serialize(destination=\"./data/graph_09.ttl\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "ed33b5de",
      "metadata": {
        "id": "ed33b5de"
      },
      "outputs": [],
      "source": [
        "# Load the graph of the previous stage, without having to redo everything\n",
        "rdf_graph = Graph().parse(\"./data/graph_09.ttl\")\n",
        "\n",
        "df = Namespace('http://example.org/DataFrame/')\n",
        "dbp = Namespace('http://dbpedia.org/property/')\n",
        "dbo = Namespace('http://dbpedia.org/ontology/')\n",
        "dbr = Namespace('http://dbpedia.org/resource/')\n",
        "wiki = Namespace('http://example.org/WikipediaApi/')\n",
        "rdf_graph.bind('df', 'http://example.org/DataFrame/')\n",
        "rdf_graph.bind('dbp', 'http://dbpedia.org/property/')\n",
        "rdf_graph.bind('dbo', 'http://dbpedia.org/ontology/')\n",
        "rdf_graph.bind('dbr', 'http://dbpedia.org/resource/')\n",
        "rdf_graph.bind('wiki', 'http://example.org/WikipediaApi/')\n"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "jk89xSctoieZ",
      "metadata": {
        "id": "jk89xSctoieZ"
      },
      "source": [
        "I apply this entailment pattern"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "xMivieg3qaU5",
      "metadata": {
        "id": "xMivieg3qaU5"
      },
      "outputs": [],
      "source": [
        "# if I have that [album] rdf:type [genre1] and [genre1] rdfs:subClassOf [genre2], add the triple [album] rdf:type [genre2]\n",
        "query = \"\"\"\n",
        "    INSERT { ?album rdf:type ?genre2 }\n",
        "    WHERE {\n",
        "        ?album rdf:type ?genre1.\n",
        "        ?genre1 rdfs:subClassOf ?genre2.\n",
        "        ?genre1 a df:Genre.\n",
        "        ?genre2 a df:Subgenre.\n",
        "    }\n",
        "\"\"\"\n",
        "rdf_graph.update(query)\n",
        "\n",
        "\n",
        "# mi:foundingCountry :- dbp:hometown / dbo:country\n",
        "query = \"\"\"\n",
        "    INSERT { ?band mi:foundingCountry ?country }\n",
        "    WHERE { ?band dbp:hometown / dbo:country ?country }\n",
        "\"\"\"\n",
        "rdf_graph.update(query)\n",
        "\n",
        "\n",
        "# mi:countryOfBirth :- dbo:birthPlace / dbo:country\n",
        "query = \"\"\"\n",
        "    INSERT { ?member mi:countryOfBirth ?country }\n",
        "    WHERE { ?member dbo:birthPlace / dbo:country ?country }\n",
        "\"\"\"\n",
        "rdf_graph.update(query)\n",
        "\n",
        "\n",
        "# mi:genre, genre of the artist inferred through the genre of the album\n",
        "query = \"\"\"\n",
        "    INSERT { ?artist mi:genre ?genre }\n",
        "    WHERE { ?album df:genre ?genre; df:artist ?artist }\n",
        "\"\"\"\n",
        "rdf_graph.update(query)\n",
        "\n",
        "\n",
        "# mi:subgenre, subgenre of the artist inferred through the subgenre of the album\n",
        "query = \"\"\"\n",
        "    INSERT { ?artist mi:subgenre ?subgenre }\n",
        "    WHERE { ?album df:subgenre ?subgenre; df:artist ?artist }\n",
        "\"\"\"\n",
        "rdf_graph.update(query)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "G0x-3EUoZ68p",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "G0x-3EUoZ68p",
        "outputId": "609a7ac6-3d0a-4388-9843-ae4ce1ee6c90"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<Graph identifier=N9ebf5091361647f8817d4a8cd7991c9b (<class 'rdflib.graph.Graph'>)>"
            ]
          },
          "execution_count": 65,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# SAVE ON DISK\n",
        "rdf_graph.serialize(destination=\"./data/graph_10.ttl\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "KBmSqQNTjQdG",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KBmSqQNTjQdG",
        "outputId": "caf0a32d-a006-4f20-942c-fe0caf7c0a25"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<Graph identifier=N9ebf5091361647f8817d4a8cd7991c9b (<class 'rdflib.graph.Graph'>)>"
            ]
          },
          "execution_count": 66,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# rerun the owl2 rl + rdfs reasoner for the new inferences\n",
        "owlrl.DeductiveClosure(owlrl.RDFS_OWLRL_Semantics).expand(rdf_graph)\n",
        "\n",
        "# SAVE ON DISK\n",
        "rdf_graph.serialize(destination=\"./data/graph_11.ttl\")"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "pmY-CKYJC4D1",
      "metadata": {
        "id": "pmY-CKYJC4D1"
      },
      "source": [
        "## **`6. Query the Graph`**"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "954faa3c-172c-4bb8-8127-d90d2cb65906",
      "metadata": {
        "id": "954faa3c-172c-4bb8-8127-d90d2cb65906"
      },
      "source": [
        "Query my graph in a way that:\n",
        "- the usefulness of having added information from dbpedia is highlighted\n",
        "- inference is exploited\n",
        "\n",
        "\n",
        "| ID | Query | Intention/insight |\n",
        "|----|------|-------------------|\n",
        "| 1  | Which are the artist with many bands? | to spot session musicians, long-career artists, 'eclectic' artists |\n",
        "| 2  | Which is the city with more bands in absolute? | maybe to discover an influential city. |\n",
        "| 3  | Which is the city with more bands in absolute that plays [ Grunge ]( https://en.wikipedia.org/wiki/Grunge )? | maybe to discover a music scene in a particular city. |\n",
        "| 4  | Give me some band/artist from the same city of Jonny_Greenwood | could be a way to find similar musical artists. |\n",
        "| 5  | which are the bands with Brazilian and American members? | maybe because I love both the Brazilian and American sound and I hope there is something that matches both? |\n",
        "| 6  | what are the albums with the highest average rating according to Rate Your Music | check if my initial data in the dataframes are well maintained in the graph |\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "a9f37816",
      "metadata": {
        "id": "a9f37816"
      },
      "outputs": [],
      "source": [
        "# Load the graph of the previous stage, without having to redo everything\n",
        "rdf_graph = Graph().parse(\"./data/graph_11.ttl\")\n",
        "\n",
        "df = Namespace('http://example.org/DataFrame/')\n",
        "dbp = Namespace('http://dbpedia.org/property/')\n",
        "dbo = Namespace('http://dbpedia.org/ontology/')\n",
        "dbr = Namespace('http://dbpedia.org/resource/')\n",
        "wiki = Namespace('http://example.org/WikipediaApi/')\n",
        "mi = Namespace('http://example.org/MaterializeInference/')\n",
        "rdf_graph.bind('df', 'http://example.org/DataFrame/')\n",
        "rdf_graph.bind('dbp', 'http://dbpedia.org/property/')\n",
        "rdf_graph.bind('dbo', 'http://dbpedia.org/ontology/')\n",
        "rdf_graph.bind('dbr', 'http://dbpedia.org/resource/')\n",
        "rdf_graph.bind('wiki', 'http://example.org/WikipediaApi/')\n",
        "rdf_graph.bind('mi', 'http://example.org/MaterializeInference/')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "53cb4f00",
      "metadata": {
        "id": "53cb4f00",
        "outputId": "f6209c38-cb41-48bb-be84-ec6bb2bc02b1"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Artists with many bands:\n",
            "The artist http://dbpedia.org/resource/Joey_Vera is/was part of 6 bands\n",
            "The artist http://dbpedia.org/resource/Madlib is/was part of 4 bands\n",
            "The artist http://dbpedia.org/resource/Jason_Tait is/was part of 4 bands\n",
            "The artist http://dbpedia.org/resource/Neil_Finn is/was part of 4 bands\n",
            "The artist http://dbpedia.org/resource/Manfred_Mann_(musician) is/was part of 4 bands\n",
            "The artist http://dbpedia.org/resource/Jukka_Koskinen is/was part of 4 bands\n",
            "The artist http://dbpedia.org/resource/Kai_Hahto is/was part of 4 bands\n",
            "The artist http://dbpedia.org/resource/Mattias_Bärjed is/was part of 4 bands\n",
            "The artist http://dbpedia.org/resource/Tommy_Karevik is/was part of 4 bands\n",
            "The artist http://dbpedia.org/resource/Ishmael_Butler is/was part of 4 bands\n"
          ]
        }
      ],
      "source": [
        "# 1 Which are the artist with many bands?\n",
        "query = \"\"\"\n",
        "    SELECT ?artist (COUNT(?band) AS ?numBands)\n",
        "    WHERE {\n",
        "        ?band dbo:bandMember ?artist.\n",
        "    } GROUP BY ?artist ORDER BY DESC(?numBands)\n",
        "\"\"\"\n",
        "results = rdf_graph.query(query)\n",
        "\n",
        "print('Artists with many bands:')\n",
        "count = 0\n",
        "for res in results:\n",
        "    # Print some example of dbr URI\n",
        "    if \"http://dbpedia.org/resource\" in res['artist']:\n",
        "        print(f\"The artist {res['artist']} is/was part of {res['numBands']} bands\")\n",
        "        count += 1\n",
        "        if count == 10:\n",
        "            break"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "246a88b7",
      "metadata": {
        "id": "246a88b7",
        "outputId": "cfdb2fa1-4a16-406c-890d-391c114a5f4e"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "The city with the most bands in absolute is:\n",
            "http://dbpedia.org/resource/London, with 84 bands\n"
          ]
        }
      ],
      "source": [
        "# 2 Which is the city with more bands in absolute?\n",
        "query = \"\"\"\n",
        "    SELECT  ?city (COUNT(?band) AS ?numBands)\n",
        "    WHERE {\n",
        "    ?band mi:foundingCity ?city.\n",
        "    } GROUP BY ?city ORDER BY DESC(?numBands)\n",
        "\"\"\"\n",
        "results = rdf_graph.query(query)\n",
        "\n",
        "\n",
        "print('The city with the most bands in absolute is:')\n",
        "for res in results:\n",
        "    print(f\"{res['city']}, with {res['numBands']} bands\")\n",
        "    break\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "6e0606bd",
      "metadata": {
        "id": "6e0606bd",
        "outputId": "1dbaa4a3-f612-4d6f-d07a-7ac4a275cc39"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "The city with the most bands that plays Grunge in absolute is:\n",
            "http://dbpedia.org/resource/Seattle, with 8 bands\n"
          ]
        }
      ],
      "source": [
        "# 3 What is the city with the most bands playing the Grunge genre ?\n",
        "query = \"\"\"\n",
        "    SELECT  ?city (COUNT(?band) AS ?numBands)\n",
        "    WHERE {\n",
        "    ?band mi:foundingCity ?city;\n",
        "          mi:genre df:Grunge.\n",
        "    } GROUP BY ?city ORDER BY DESC(?numBands)\n",
        "\"\"\"\n",
        "results = rdf_graph.query(query)\n",
        "\n",
        "\n",
        "print('The city with the most bands that plays Grunge in absolute is:')\n",
        "for res in results:\n",
        "    print(f\"{res['city']}, with {res['numBands']} bands\")\n",
        "    break\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "5a9b2366",
      "metadata": {
        "id": "5a9b2366",
        "outputId": "fced59c8-42c5-45fd-fcf7-ab2fd9408be1"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Some band/artist from the same city of Jonny_Greenwood:\n",
            "http://dbpedia.org/resource/Colin_Greenwood\n",
            "http://dbpedia.org/resource/Ed_O'Brien\n",
            "http://dbpedia.org/resource/Gaz_Coombes\n",
            "http://dbpedia.org/resource/Rob_Coombes\n"
          ]
        }
      ],
      "source": [
        "# 4 Give me some band/artist from the same city of Jonny_Greenwood\n",
        "query = \"\"\"\n",
        "    SELECT  ?artist\n",
        "    WHERE {\n",
        "    ?artist  mi:cityOfBirth ?city.\n",
        "    dbr:Jonny_Greenwood  mi:cityOfBirth ?city.\n",
        "    FILTER (?artist != dbr:Jonny_Greenwood)\n",
        "    }\n",
        "\"\"\"\n",
        "results = rdf_graph.query(query)\n",
        "\n",
        "\n",
        "print('Some band/artist from the same city of Jonny_Greenwood:')\n",
        "for res in results:\n",
        "    # Print some example of dbr URI\n",
        "    if (\"http://dbpedia.org/resource\" in res['artist'] ):\n",
        "        print(res['artist'] )"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "b34b817f",
      "metadata": {
        "id": "b34b817f",
        "outputId": "e5a678c0-0a0a-4a46-86fa-216226b066b7"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Bands with Brazilian and American members:\n",
            "http://dbpedia.org/resource/The_Strokes\n",
            "http://dbpedia.org/resource/Megadeth\n",
            "http://dbpedia.org/resource/Sepultura\n"
          ]
        }
      ],
      "source": [
        "# 5 which are the bands with Brazilian and American members?\n",
        "query = \"\"\"\n",
        "    SELECT DISTINCT ?band\n",
        "    WHERE {\n",
        "    ?band dbo:bandMember ?member1, ?member2.\n",
        "    ?member1 mi:countryOfBirth dbr:Brazil.\n",
        "    ?member2 mi:countryOfBirth dbr:United_States.\n",
        "    }\n",
        "\"\"\"\n",
        "results = rdf_graph.query(query)\n",
        "\n",
        "print('Bands with Brazilian and American members:')\n",
        "count = 0\n",
        "for res in results:\n",
        "    # Print some example of dbr URI\n",
        "    if \"http://dbpedia.org/resource\" in res['band']:\n",
        "        print(res['band'] )\n",
        "        count += 1\n",
        "        if count == 10:\n",
        "            break\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "id": "4ab45fd1",
      "metadata": {
        "id": "4ab45fd1",
        "outputId": "1a7ed7fd-41e3-44fe-ea91-7c4fc90961e0"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "The albums with the highest average rating according to Rate Your Music are the following:\n",
            "http://dbpedia.org/resource/The_Black_Saint_and_the_Sinner_Lady\n",
            "http://dbpedia.org/resource/A_Love_Supreme\n",
            "http://dbpedia.org/resource/In_the_Court_of_the_Crimson_King\n",
            "http://dbpedia.org/resource/To_Pimp_a_Butterfly\n",
            "http://dbpedia.org/resource/Madvillainy\n",
            "http://dbpedia.org/resource/Abbey_Road\n",
            "http://dbpedia.org/resource/Remain_in_Light\n"
          ]
        }
      ],
      "source": [
        "# 6 what are the albums with the highest average rating according to \"Best albums of all time\" of Rate Your Music\n",
        "query = \"\"\"\n",
        "    SELECT DISTINCT ?album\n",
        "    WHERE {\n",
        "        ?album df:averageRating ?averageRating\n",
        "    } ORDER BY DESC(?averageRating) LIMIT 50\n",
        "\"\"\"\n",
        "results = rdf_graph.query(query)\n",
        "\n",
        "print('The albums with the highest average rating according to Rate Your Music are the following:')\n",
        "for res in results:\n",
        "    if \"http://dbpedia.org/resource\" in res['album']:\n",
        "        print(res['album'] )\n",
        "        count += 1\n",
        "        if count == 10:\n",
        "            break"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "w3RWLLOEHlvj",
      "metadata": {
        "id": "w3RWLLOEHlvj"
      },
      "source": [
        "**Thanks for the attention!**\n",
        "\n",
        "Alessandro Ghiotto"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "toc_visible": true
    },
    "ipub": {
      "titlepage": {
        "author": "Authors Name",
        "email": "authors@email.com",
        "institution": [
          "Institution1",
          "Institution2"
        ],
        "subtitle": "Sub-Title",
        "supervisors": [
          "First Supervisor",
          "Second Supervisor"
        ],
        "tagline": "A tagline for the report.",
        "title": "Main-Title"
      }
    },
    "kernelspec": {
      "display_name": "Python 3 (ipykernel)",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.9.7"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 5
}