R-BERT

https://github.com/monologg/R-BERT

"Matching the Blanks: Distributional Similarity for Relation Learning"

https://github.com/mcschmitz/BERT-Relation-Extraction

https://medium.com/e-bot7-tech/matching-the-blanks-78e8063794b5

Bert-RE

https://github.com/plkmo/BERT-Relation-Extraction

NCE(Noise constrastive estimation)

http://incredible.ai/nlp/2017/11/25/Word2Vec-Noise-Constrastive-Estimation/

Distant supervision for relation extraction without labeled data

https://web.stanford.edu/~jurafsky/mintz.pdf

Freebase 사용해서 distant supervision 진행

→ 한국어 정리 페이지

https://roomylee.github.io/distant-supervision-relation-extraction/

My Direction

  1. ‘Knowledge Graph – A Powerful Data Science Technique to Mine Information from Text (with Python code)’을 먼저 적용
  2. label을 얻은 뒤, SSAN을 적용

My Considerations

  1. POS, NER의 결과를 어떻게 이용할 수 있을까?
  2. 그래프를 그린다면 관련이 클수록 node의 크기가 커져야 할 거 같음. 이건 어떻게 구현?