On Evaluating and Comparing Open Domain Dialog Systems

TL;DR

This paper presents a comprehensive, multi-metric evaluation strategy for open domain dialog systems, aiming to reduce subjectivity and better correlate with human judgments, based on data from the Alexa Prize competition.

Contribution

It introduces a novel multi-metric evaluation framework that provides granular analysis and a unified scoring mechanism for conversational agents, leveraging large-scale real-world data.

Findings

Metrics correlate well with human judgment

Proposed evaluation reduces subjectivity in assessments

Framework applied successfully in Alexa Prize competition

Abstract

Conversational agents are exploding in popularity. However, much work remains in the area of non goal-oriented conversations, despite significant growth in research interest over recent years. To advance the state of the art in conversational AI, Amazon launched the Alexa Prize, a 2.5-million dollar university competition where sixteen selected university teams built conversational agents to deliver the best social conversational experience. Alexa Prize provided the academic community with the unique opportunity to perform research with a live system used by millions of users. The subjectivity associated with evaluating conversations is key element underlying the challenge of building non-goal oriented dialogue systems. In this paper, we propose a comprehensive evaluation strategy with multiple metrics designed to reduce subjectivity by selecting metrics which correlate well with human judgement. The proposed metrics provide granular analysis of the conversational agents, which is not captured in human ratings. We show that these metrics can be used as a reasonable proxy for human judgment. We provide a mechanism to unify the metrics for selecting the top performing agents, which has also been applied throughout the Alexa Prize competition. To our knowledge, to date it is the largest setting for evaluating agents with millions of conversations and hundreds of thousands of ratings from users. We believe that this work is a step towards an automatic evaluation process for conversational AIs.