Learning to Generate Structured Queries from Natural Language with Indirect Supervision
This addresses the challenge of training data acquisition for natural language to SQL systems, making it more scalable by leveraging abundant online question-answer resources.
The paper tackles the problem of generating SQL queries from natural language by using indirect supervision from question-answer pairs instead of direct SQL annotations, which are harder to obtain. The proposed end-to-end neural model with reinforcement learning outperforms baseline models on datasets from movie and academic publication domains.
Generating structured query language (SQL) from natural language is an emerging research topic. This paper presents a new learning paradigm from indirect supervision of the answers to natural language questions, instead of SQL queries. This paradigm facilitates the acquisition of training data due to the abundant resources of question-answer pairs for various domains in the Internet, and expels the difficult SQL annotation job. An end-to-end neural model integrating with reinforcement learning is proposed to learn SQL generation policy within the answer-driven learning paradigm. The model is evaluated on datasets of different domains, including movie and academic publication. Experimental results show that our model outperforms the baseline models.