Quantifying the Risks of Tool-assisted Rephrasing to Linguistic Diversity

TL;DR

This paper investigates how widespread use of rephrasing tools and language models might impact linguistic diversity by measuring semantic and vocabulary changes across various domains.

Contribution

It introduces a method to quantify the risk of language reduction caused by rephrasing tools on large-scale human-generated text.

Findings

Rephrasing tools cause measurable semantic shifts.

Vocabulary diversity decreases with tool usage.

Risks vary across different text domains.

Abstract

Writing assistants and large language models see widespread use in the creation of text content. While their effectiveness for individual users has been evaluated in the literature, little is known about their proclivity to change language or reduce its richness when adopted by a large user base. In this paper, we take a first step towards quantifying this risk by measuring the semantic and vocabulary change enacted by the use of rephrasing tools on a multi-domain corpus of human-generated text.