Role of Dependency Distance in Text Simplification: A Human vs ChatGPT Simplification Comparison

TL;DR

This paper compares human and ChatGPT text simplification, revealing that both reduce dependency distance, with humans achieving the greatest reduction, across sentences of varying grammatical difficulty.

Contribution

It provides a comparative analysis of dependency distance in human versus ChatGPT text simplification across different sentence complexities.

Findings

Humans produce the lowest dependency distances in simplified sentences.

ChatGPT reduces dependency distance more than original sentences but less than humans.

Dependency distance correlates with grammatical difficulty.

Abstract

This study investigates human and ChatGPT text simplification and its relationship to dependency distance. A set of 220 sentences, with increasing grammatical difficulty as measured in a prior user study, were simplified by a human expert and using ChatGPT. We found that the three sentence sets all differed in mean dependency distances: the highest in the original sentence set, followed by ChatGPT simplified sentences, and the human simplified sentences showed the lowest mean dependency distance.