Do Text Simplification Systems Preserve Meaning? A Human Evaluation via Reading Comprehension

TL;DR

This paper introduces a human evaluation framework using reading comprehension questions to assess whether text simplification systems preserve meaning, revealing that even top systems often fail to maintain answerability for all questions.

Contribution

It presents a novel human-centered evaluation method for meaning preservation in text simplification and compares it with existing automatic metrics and QA systems.

Findings

Supervised pre-trained systems perform best on reading comprehension tasks.

Even the top systems leave at least 14% of questions unanswerable.

Current automatic metrics and QA systems only partially align with human judgments.

Abstract

Automatic text simplification (TS) aims to automate the process of rewriting text to make it easier for people to read. A pre-requisite for TS to be useful is that it should convey information that is consistent with the meaning of the original text. However, current TS evaluation protocols assess system outputs for simplicity and meaning preservation without regard for the document context in which output sentences occur and for how people understand them. In this work, we introduce a human evaluation framework to assess whether simplified texts preserve meaning using reading comprehension questions. With this framework, we conduct a thorough human evaluation of texts by humans and by nine automatic systems. Supervised systems that leverage pre-training knowledge achieve the highest scores on the reading comprehension (RC) tasks amongst the automatic controllable TS systems. However, even the best-performing supervised system struggles with at least 14% of the questions, marking them as "unanswerable'' based on simplified content. We further investigate how existing TS evaluation metrics and automatic question-answering systems approximate the human judgments we obtained.