Merge branch 'main' of https://github.com/epfl-ada/ada-2024-project-a…

…daventurers2024

results.ipynb

@@ -738,7 +738,7 @@
    "source": [
     "To understand which narrative elements might contribute to poor movie reception, we first analyzed individual tropes associated with the lowest average ratings. We then examined how these problematic tropes tend to appear together in films, identifying patterns in their co-occurrence and exploring potential combinations that might predict poor audience reception.\n",
     "\n",
-    "The box plot displays the rating distributions for the ten tropes associated with the lowest average ratings in our dataset. Most tropes show average ratings between 4 and 6, with considerable variation as indicated by the long whiskers. \"XtremeKoolLetterz\" stands out with both the lowest median rating and an outlier point near rating 2, suggesting that intentional misspellings in titles particularly displease audiences."
+    "The box plot displays the rating distributions for the ten tropes associated with the lowest average ratings in our dataset. Most tropes show average ratings between 4 and 6, with considerable variation as indicated by the long whiskers."
    ]
   },
   {
@@ -800,7 +800,7 @@
    "source": [
     "To understand what trope combinations might lead to poor audience reception, we isolated the ten clusters with the lowest average ratings from our dataset. The t-SNE visualization of these clusters reveals a notable pattern: most poorly-rated movies concentrate in a dense cluster, with some outliers scattered across the periphery. This concentration suggests that many poorly-rated movies share similar trope combinations, potentially indicating specific narrative patterns that consistently fail.\n",
     "\n",
-    "Unlike the full dataset visualization, which showed more uniform distribution, these worst-rated clusters display significant clumping, particularly evident in clusters 13 and 20 (shown in red and purple). The presence of isolated points at the edges of the plot suggests that while most poorly-rated movies follow similar trope patterns, there are also some films that receive bad ratings despite having unique or unusual trope combinations. This insight could be valuable for filmmakers, as it highlights both common problematic trope patterns to avoid."
+    "Unlike the full dataset visualization, which showed more uniform distribution, these worst-rated clusters display significant clumping. The presence of isolated points at the edges of the plot suggests that while most poorly-rated movies follow similar trope patterns, there are also some films that receive bad ratings despite having unique or unusual trope combinations. This insight could be valuable for filmmakers, as it highlights both common problematic trope patterns to avoid."
    ]
   },
   {