Tackling Temporal Questions in Natural Language Interface to Databases

Abstract

Temporal aspect is one of the most challenging areas in Natural Language Interface to Databases (NLIDB). This paper tackles temporal aspect and examines how it is being carried in NLIDB. We discuss how it is being supported by the research community at both levels: popular annotated dataset (e.g. Spider) and current advanced models. We also present a new dataset and accompanied databases created for supporting temporal aspect in NLIDB. We experiment with Picard and ValueNet models to investigate how our new dataset helps these models understand, learn and improve their performance in temporal aspect.