論文:Modeling Semantic Relationship in Multi-turn Conversations with Hierarchical Latent Variables
Lei Shen,Yang Feng,Haolan Zhan
模型框架如下圖:
zc is used to control the whole background in which the conversation takes place, zp is for the consistency of topic between
query and response pair, zq and zr try to model the content difference in each of them, respectively
先驗和後驗分佈用了正態分佈,其中的期望和方差用神經網絡計算出來。
實驗效果:
效果提升一般。