Improving Response Selection in Multi-Turn Dialogue Systems by Incorporating Domain Knowledge

TL;DR

This paper introduces a neural network model that enhances multi-turn dialogue response selection by integrating domain knowledge through external descriptions, leading to improved performance over existing methods.

Contribution

The work presents a novel architecture combining context attention and domain knowledge encoding to advance response selection in multi-turn dialogues.

Findings

Outperforms state-of-the-art response selection models

Effectively incorporates domain-specific knowledge

Improves response relevance in multi-turn conversations

Abstract

Building systems that can communicate with humans is a core problem in Artificial Intelligence. This work proposes a novel neural network architecture for response selection in an end-to-end multi-turn conversational dialogue setting. The architecture applies context level attention and incorporates additional external knowledge provided by descriptions of domain-specific words. It uses a bi-directional Gated Recurrent Unit (GRU) for encoding context and responses and learns to attend over the context words given the latent response representation and vice versa.In addition, it incorporates external domain specific information using another GRU for encoding the domain keyword descriptions. This allows better representation of domain-specific keywords in responses and hence improves the overall performance. Experimental results show that our model outperforms all other state-of-the-art methods for response selection in multi-turn conversations.