arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1607.05809cs.CLcs.AI
收藏

基于标记问答对的神经语境会话学习

Neural Contextual Conversation Learning with Labeled Question-Answering Pairs

RSVP Technologies Inc.(RSVP科技有限公司);University of Waterloo(滑铁卢大学)

Kun Xiong, Anqi Cui, Zefeng Zhang, Ming Li

摘要:

神经会话模型倾向于在不同的上下文中产生通用或安全的响应,例如,回答\textit{"Of course"}的叙述性陈述或回答\textit{"I don't know"}的问题。在本文中,我们提出了一种端到端的方法来避免神经生成模型中的此类问题。标准的序列到序列(seq 2seq)模型中引入了额外的记忆机制,以便在生成句子时可以考虑上下文。三个seq 2seq模型,分别从隐藏输入,隐藏输入/输出和门控上下文注意结构中记忆一个固定大小的上下文向量,已经在汉语标记问答对数据集上进行了训练和测试。在困惑测试中,具有情境注意的模型优于其他模型,包括最先进的seq 2seq模型。新颖的语境模型产生了多样化和强大的反应,并能够适当地进行广泛的话题对话。

英文摘要:

Neural conversational models tend to produce generic or safe responses in different contexts, e.g., reply \textit{"Of course"} to narrative statements or \textit{"I don't know"} to questions. In this paper, we propose an end-to-end approach to avoid such problem in neural generative models. Additional memory mechanisms have been introduced to standard sequence-to-sequence (seq2seq) models, so that context can be considered while generating sentences. Three seq2seq models, which memorize a fix-sized contextual vector from hidden input, hidden input/output and a gated contextual attention structure respectively, have been trained and tested on a dataset of labeled question-answering pairs in Chinese. The model with contextual attention outperforms others including the state-of-the-art seq2seq models on perplexity test. The novel contextual model generates diverse and robust responses, and is able to carry out conversations on a wide range of topics appropriately.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑