Pair2vec: Compositional word-pair embeddings for cross-sentence inference

Computer Science

• Mutual Information 100%
• Background Knowledge 100%
• Common Sense 100%
• Natural-Language Understanding 100%
• Inference Model 100%
• Word Embedding 100%
• Embedding Pair 100%

Keyphrases

• Word Pairs 100%
• Cross-sentence 100%
• Inference Problem 33%
• Background Knowledge 33%
• Word Representation 33%
• Two-word 33%
• Inference Model 33%
• Word Embedding 33%
• Pointwise mutual Information 33%
• Compositional Function 33%
• SQuAD 2.0 33%
• SQuAD Dataset 33%
• Attention Layer 33%
• Pairwise Embedding 33%
• BiDAF 33%