이 프로젝트는...
각 레파지토리에 대한 설명
# | repository | link | description |
---|---|---|---|
1 | preprocessing | https://github.com/perfume-reconmendation/preprocessing | 크롤링한 리뷰 데이터 전처리 |
2 | bert_classification | https://github.com/perfume-reconmendation/bert_classification | BERT 방법론을 이용해 향수 리뷰를 보고 향수 종류를aquatic,tobacco,floral,fruity(0~3)로 분류 |
3 | word-embedding_hyun | https://github.com/perfume-reconmendation/word-embedding_hyun | word2vec 방법론을 이용해 유저의 쿼리와 향수 리뷰들의 유사도를 구하고 추천 |
4 | doc2vec | https://github.com/perfume-reconmendation/doc2vec | doc2vec, bert 방법론을 이용해 유저의 쿼리와 향수 리뷰들의 유사도를 구하고 추천 |
5 | flask_server | https://github.com/perfume-reconmendation/flask_server | flask서버에 각자 만든 모델을 합침 |
6 | frontend | https://github.com/perfume-reconmendation/frontend | frontend 구축 |
7 | cnn_classification | https://github.com/perfume-reconmendation/cnn_classification | CNN 방법론을 이용해 향수 리뷰를 보고 향수 종류를aquatic,tobacco,floral,fruity(0~3)로 분류 |
8 | sentence_similarity_bert_kyuuuw | https://github.com/perfume-reconmendation/sentence_similarity_bert_kyuuuw | BERT 방법론을 이용해 pre-trained 된 모델을 통해 임베딩하고 문장 유사도 계산 |