Crossing Variational Autoencoders for Answer Retrieval
Wenhao Yu, Lingfei Wu, Qingkai Zeng, Tao Shu, Yu Deng, Meng Jiang Β· 2020
Answer retrieval is to find the most aligned answer from a large set of candidates given a question.Learning vector representations of questions/answers is the key factor.Questionanswer alignment and question/answer semantics are two important signals for learning the representations.Existing methods learned semantic representations with dual encoders or dual variational auto-encoders.The semantic information was learned from language models or question-to-question (answer-to-answer) generative processes.However, the alignment and semantics were too separate to capture the aligned semantics between question and answer.In this work, we propose to cross variational auto-encoders by generating questions with aligned answers and generating answers with aligned questions.Experiments show that our method outperforms the state-of-theart answer retrieval method on SQuAD.Question Answer Decoder π(π|π π ) π(π|π π ) π(π¦|π§ !, π§ " ) π(π¦|π§ !, π§ " ) Question Answer Question Answer Decoder Encoder Encoder Decoder Decoder Encoder π(π§ !|π) π(π§ " |π) Encoder Encoder (a) Dual-Encoders (Yang et al., 2019)Question Answer Decoder π(π|π π ) π(π|π π ) π(π¦|π§ !, π§ " ) π(π¦|π§ !, π§ " ) Question Answer Question Answer Decoder Encoder Encoder Decoder Decoder Encoder π(π§ !|π) π(π§ " |π) Encoder Encoder (b) Dual-VAEs (Shen et al., 2018) π§ !~π(π§ ! ) π§ " ~π(π§ " ) Question Answer π§ !~π(π§ ! ) π(π§ !|π) π§ " ~π(π§ " ) π(π§ " |π) Question Answer π(π¦|π§ !, π§ " ) π(π|π§ ! ) π(π|π§ " )