Skip to content

Latest commit

 

History

History
43 lines (34 loc) · 1.96 KB

literatures.md

File metadata and controls

43 lines (34 loc) · 1.96 KB

Literatures

NER

Domain alignment

Context-Specific Word Embedding

Tutorials

NER Dataset

Classification Dataset

  • 20-Newsgroups
  • Amazon Data
    • Usually use reduced set(Books, DVDs, Electronicsand Kitchen)(Blitzer et al.)
  • SentDat
    • (video game(V), electronics(E), software(S) and hotel (H))

Multilingual Dataset

  • NTCIR
    • EN & CN
  • Sanders/GermEval
    • EN & German, sentiment classification, pos/neu/neg, multilingual-multiclass

pretrained BERT