Interactive Semantic Parsing for If-Then Recipes via Hierarchical Reinforcement Learning

Ziyu Yao, Xiujun Li, Jianfeng Gao, Brian M. Sadler, Huan Sun

AAAI 2019 pp. 2547-2554

doi:10.1609/AAAI.V33I01.33012547 /aaai/2019/yao2019aaai-interactive/

Abstract

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

PDF AAAI Semantic Scholar

Cite

Text

Yao et al. "Interactive Semantic Parsing for If-Then Recipes via Hierarchical Reinforcement Learning." AAAI Conference on Artificial Intelligence, 2019. doi:10.1609/AAAI.V33I01.33012547

Markdown

[Yao et al. "Interactive Semantic Parsing for If-Then Recipes via Hierarchical Reinforcement Learning." AAAI Conference on Artificial Intelligence, 2019.](https://mlanthology.org/aaai/2019/yao2019aaai-interactive/) doi:10.1609/AAAI.V33I01.33012547

BibTeX

@inproceedings{yao2019aaai-interactive,
  title     = {{Interactive Semantic Parsing for If-Then Recipes via Hierarchical Reinforcement Learning}},
  author    = {Yao, Ziyu and Li, Xiujun and Gao, Jianfeng and Sadler, Brian M. and Sun, Huan},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2019},
  pages     = {2547-2554},
  doi       = {10.1609/AAAI.V33I01.33012547},
  url       = {https://mlanthology.org/aaai/2019/yao2019aaai-interactive/}
}