VoiceQuerySystem: A Voice-driven Database Querying System Using Natural Language Questions

Abstract

With recent development in natural language processing (NLP) and automatic speech recognition (ASR), voice-based interfaces have become a necessity for applications such as chatbots, search engines, and databases. In this demonstration, we introduce VoiceQuerySystem, a voice-based database querying system that enables users to conduct data operations with natural language questions (NLQs). Different from existing voice-based interfaces such as SpeakQL or EchoQuery, which restricts the voice input to be an exact SQL or follow a pre-defined template, VoiceQuerySystem attempts to achieve data manipulation via common NLQs, and thus does not require the user's technical background in SQL language. The underlying techniques in VoiceQuerySystem is a new task named Speech-to-SQL, which aims to understand the semantic in speech and then translate it into SQL queries. We explore two proposed approaches-the cascaded one and the end-to-end (E2E) one towards speech-to-SQL translation. The cascaded method first converts the user's voice-based NLQs into text by a self-developed ASR module, and then conducts downstream SQL generation via a text-to-SQL model (i.e., IRNet). In contrast, the E2E method is a novel neural architecture named SpeechSQLNet designed by us, which converts the speech signals into SQL queries directly without the middle medium as text. Extensive experiments and demonstrations validate the rationale of the speech-to-SQL task and the effectiveness of the proposed SpeechSQLNet model. To the best of our knowledge, this is the first system that provides a voice-based querying functionality on DBMS from common NLQs.