Using a Logic Programming Framework to Control Database Query Dialogues in Natural Language

Abstract

We present a natural language question/answering system to interface the University of Évora databases that uses clarification dialogs in order to clarify user questions. It was developed in an integrated logic programming framework, based on constraint logic programming using the GnuProlog(-cx) language [2,11] and the ISCO framework [1]. The use of this LP framework allows the integration of Prolog-like inference mechanisms with classes and inheritance, constraint solving algorithms and provides the connection with relational databases, such as PostgreSQL. This system focus on the questions’ pragmatic analysis, to handle ambiguity, and on an efficient dialogue mechanism, which is able to place relevant questions to clarify the user intentions in a straightforward manner. Proper Nouns resolution and the pp-attachment problem are also handled. This paper briefly presents this innovative system focusing on its ability to correctly determine the user intention through its dialogue capability.