Interactive Semantic Parsing for If-Then Recipes via Hierarchical Reinforcement Learning
Interactive Semantic Parsing for If-Then Recipes via Hierarchical Reinforcement Learning
复制标题
DOI：
10.1609/aaai.v33i01.33012547
复制
发表时间：
2018-08
期刊：
影响因子：
--
通讯作者：
Ziyu Yao;Xiujun Li;Jianfeng Gao;Brian M. Sadler;Huan Sun
中科院分区：
文献类型：
--
作者：
Ziyu Yao;Xiujun Li;Jianfeng Gao;Brian M. Sadler;Huan Sun
在给定文本描述的情况下，大多数现有的语义解析器一次合成一个程序。然而，仅仅基于描述来生成正确的程序是相当具有挑战性的，而在现实中，描述往往是模棱两可或不完整的。在这篇文章中，我们研究了交互式语义分析，在这种分析中，代理可以通过多轮对话向用户提出澄清问题来解决歧义，这是一种重要的程序类型，称为“IF-THEN食谱”。我们开发了一种基于层次化强化学习(HRL)的代理，该代理能够在向用户提出最少问题的情况下显著提高句法分析性能。在模拟和人工评估下的结果表明，我们的代理的性能大大优于非交互式语义分析器和基于规则的代理。
Given a text description, most existing semantic parsers synthesize a program in one shot. However, it is quite challenging to produce a correct program solely based on the description, which in reality is often ambiguous or incomplete. In this paper, we investigate interactive semantic parsing, where the agent can ask the user clarification questions to resolve ambiguities via a multi-turn dialogue, on an important type of programs called “If-Then recipes.” We develop a hierarchical reinforcement learning (HRL) based agent that significantly improves the parsing performance with minimal questions to the user. Results under both simulation and human evaluation show that our agent substantially outperforms non-interactive semantic parsers and rule-based agents.1