BUMP: A Benchmark of Unfaithful Minimal Pairs for Meta-Evaluation of Faithfulness Metrics

TL;DR

BUMP introduces a benchmark dataset of minimally different, human-written summary pairs with single errors to evaluate faithfulness metrics in summarization, addressing limitations of existing benchmarks.

Contribution

The paper presents BUMP, a novel dataset for diagnosing faithfulness metrics' consistency, effectiveness on human texts, and sensitivity to specific errors in summarization.

Findings

BUMP summaries are harder to discriminate and less probable under SOTA models.

It reveals that the most discriminative metrics are often less consistent.

BUMP enables evaluation of metrics on individual error types.

Abstract

The proliferation of automatic faithfulness metrics for summarization has produced a need for benchmarks to evaluate them. While existing benchmarks measure the correlation with human judgements of faithfulness on model-generated summaries, they are insufficient for diagnosing whether metrics are: 1) consistent, i.e., indicate lower faithfulness as errors are introduced into a summary, 2) effective on human-written texts, and 3) sensitive to different error types (as summaries can contain multiple errors). To address these needs, we present a benchmark of unfaithful minimal pairs (BUMP), a dataset of 889 human-written, minimally different summary pairs, where a single error is introduced to a summary from the CNN/DailyMail dataset to produce an unfaithful summary. We find BUMP complements existing benchmarks in a number of ways: 1) the summaries in BUMP are harder to discriminate and less probable under SOTA summarization models, 2) unlike non-pair-based datasets, BUMP can be used to measure the consistency of metrics, and reveals that the most discriminative metrics tend not to be the most consistent, and 3) unlike datasets containing generated summaries with multiple errors, BUMP enables the measurement of metrics' performance on individual error types.