Do Text Simplification Systems Preserve Meaning? A Human Evaluation via Reading Comprehension
This addresses the need for reliable meaning preservation in text simplification, particularly for users who rely on simplified content, though it is incremental as it focuses on evaluation rather than new simplification methods.
The paper tackled the problem of evaluating whether text simplification systems preserve meaning by introducing a human evaluation framework based on reading comprehension questions, finding that even the best-performing supervised system struggled with at least 14% of questions, marking them as unanswerable.
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.