Towards a Sentiment-Aware Conversational Agent

TL;DR

This paper introduces a sentiment-aware conversational agent that predicts appropriate sentiments and generates contextually and sentimentually suitable replies, improving dialogue quality and sentiment expression through an end-to-end approach.

Contribution

It presents a novel end-to-end model combining sentiment prediction and conditioned text generation for more emotionally aware dialogue systems.

Findings

Automatic and human evaluations show improved sentiment accuracy.

Guided text generation enhances reply quality.

Sentiment evaluation models effectively assess agent responses.

Abstract

In this paper, we propose an end-to-end sentiment-aware conversational agent based on two models: a reply sentiment prediction model, which leverages the context of the dialogue to predict an appropriate sentiment for the agent to express in its reply; and a text generation model, which is conditioned on the predicted sentiment and the context of the dialogue, to produce a reply that is both context and sentiment appropriate. Additionally, we propose to use a sentiment classification model to evaluate the sentiment expressed by the agent during the development of the model. This allows us to evaluate the agent in an automatic way. Both automatic and human evaluation results show that explicitly guiding the text generation model with a pre-defined set of sentences leads to clear improvements, both regarding the expressed sentiment and the quality of the generated text.