Antoine Chammas Undergraduate Dissertation 2017/18

Building an Information Seeking Dialogue Agent

Supervised by R.Gaizauskas

Abstract

Given recent advances in speech recognition, there is an increasing interest in voice-driven information provision systems. Siri, Cortana, Alexa and Google's Assistant are all examples of such systems. Their architecture usually consists of speech recognition and speech generation components as initial and final components and then in between components to analyze the user input, manage the dialogue between the user and the system, carry out any information seeking task required to answer a user query and formulate a natural language response to the user.

The aim of this project is to design, implement, and evaluate an information seeking dialogue agent that aims to carry out a fluent dialogue with a user to answer questions using a large, multi-domain structured or semi-structured knowledge base as an information source. The agent will be text-based, the system should be able to carry out naturalistic dialogues on topics covered by the information sources such as "What is X?- Where is it? - Who built it?" etc.