MaskEval: Weighted MLM-Based Evaluation for Text Summarization and Simplification

TL;DR

MaskEval is a flexible, reference-less evaluation metric for text summarization and simplification that uses weighted masked language modeling to assess multiple quality aspects.

Contribution

It introduces a novel MLM-based evaluation method with an attention-like weighting mechanism, adaptable to various quality dimensions without requiring references.

Findings

High correlation with human judgments in English tasks

Effective transfer between summarization and simplification evaluations

Demonstrates adaptability across multiple quality dimensions

Abstract

In text summarization and simplification, system outputs must be evaluated along multiple dimensions such as relevance, factual consistency, fluency, and grammaticality, and a wide range of possible outputs could be of high quality. These properties make the development of an adaptable, reference-less evaluation metric both necessary and challenging. We introduce MaskEval, a reference-less metric for text summarization and simplification that operates by performing masked language modeling (MLM) on the concatenation of the candidate and the source texts. It features an attention-like weighting mechanism to modulate the relative importance of each MLM step, which crucially allows it to be adapted to evaluate different quality dimensions. We demonstrate its effectiveness on English summarization and simplification in terms of correlations with human judgments, and explore transfer scenarios between the two tasks.