Training Conversational Agents with Generative Conversational Networks

TL;DR

This paper introduces Generative Conversational Networks to automatically generate training data for social conversational agents, enabling effective training with significantly less seed data and improving conversational AI capabilities.

Contribution

The work presents a novel approach using Generative Conversational Networks to reduce data requirements for training social conversational agents.

Findings

Achieves near-baseline performance with only 10% seed data

Demonstrates effectiveness on TopicalChat dataset

Validates with both automatic metrics and human evaluation

Abstract

Rich, open-domain textual data available on the web resulted in great advancements for language processing. However, while that data may be suitable for language processing tasks, they are mostly non-conversational, lacking many phenomena that appear in human interactions and this is one of the reasons why we still have many unsolved challenges in conversational AI. In this work, we attempt to address this by using Generative Conversational Networks to automatically generate data and train social conversational agents. We evaluate our approach on TopicalChat with automatic metrics and human evaluators, showing that with 10% of seed data it performs close to the baseline that uses 100% of the data.