HEDGES AND BOOSTERS IN AI AND HUMAN WRITING: A COMPARATIVE ANALYSIS

Shalevska, Elena (2024) HEDGES AND BOOSTERS IN AI AND HUMAN WRITING: A COMPARATIVE ANALYSIS. Knowledge – International Journal, 65 (5). ISSN 1857-923X

[thumbnail of HEDGES+AND+BOOSTERS+IN+AI+AND+HUMAN+WRITING+A+COMPARATIVE+ANALYSIS.pdf] Text
HEDGES+AND+BOOSTERS+IN+AI+AND+HUMAN+WRITING+A+COMPARATIVE+ANALYSIS.pdf - Published Version

Download (629kB)

Abstract

The present qualitative study aims to measure and describe the frequencies and context of use of various hedges and boosters in a self-created corpus of 100 ChatGPT 3.5-generated essays on 100 different topics. The results were then compared to 100 human-written sample essays. The analysis, primarily done via AntConc, shows high frequency use of some hedges (such as “may”) and no use of boosters (such as “clearly”, “definitely”, “certainly” etc.) in AI works, in comparison to human texts which tend to use a wider array of hedges and boosters, more frequently. This discrepancy highlights the distinct linguistic styles between AI-generated and human produced texts, suggesting potential areas for further development in AI language processing.

Item Type: Article
Subjects: Scientific Fields (Frascati) > Humanities > Languages and literature
Divisions: Faculty of Education
Depositing User: MA Elena Shalevska
Date Deposited: 27 Mar 2025 09:06
Last Modified: 27 Mar 2025 09:06
URI: https://eprints.uklo.edu.mk/id/eprint/10825

Actions (login required)

View Item View Item