GO FIGURE: A Meta Evaluation of Factuality in Summarization

TL;DR

This paper introduces GO FIGURE, a comprehensive meta-evaluation framework for assessing factuality metrics in summarization, revealing strengths and limitations of current metrics across multiple tasks.

Contribution

It proposes five necessary conditions for evaluating factuality metrics and demonstrates the framework's effectiveness across various summarization datasets and metrics.

Findings

QA metrics generally outperform standard factuality metrics.

Performance of factuality metrics depends heavily on question generation methods.

The framework is extensible to multiple factuality and generation metrics.

Abstract

While neural language models can generate text with remarkable fluency and coherence, controlling for factual correctness in generation remains an open research question. This major discrepancy between the surface-level fluency and the content-level correctness of neural generation has motivated a new line of research that seeks automatic metrics for evaluating the factuality of machine text. In this paper, we introduce GO FIGURE, a meta-evaluation framework for evaluating factuality evaluation metrics. We propose five necessary and intuitive conditions to evaluate factuality metrics on diagnostic factuality data across three different summarization tasks. Our benchmark analysis on ten factuality metrics reveals that our meta-evaluation framework provides a robust and efficient evaluation that is extensible to multiple types of factual consistency and standard generation metrics, including QA metrics. It also reveals that while QA metrics generally improve over standard metrics that measure factuality across domains, performance is highly dependent on the way in which questions are generated.