Characterizing Physiological and Behavioral Responses Toward Human and AI-generated True and Fake News

Master thesis (2024)

Authors

L.A. Wu Electrical Engineering, Mathematics and Computer Science

Contributors

Pablo Cesar (mentor)
Abdallah El Ali Centrum Wiskunde & Informatica (CWI) (graduation committee member)
Ujwal Gadiraju Web Information Systems - EEMCS (mentor)
Faculty
Electrical Engineering, Mathematics and Computer Science, Electrical Engineering, Mathematics and Computer Science
More Info
expand_more

Published Date

12-07-2024

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Faculty

Electrical Engineering, Mathematics and Computer Science

Abstract

The spread of misinformation on social media has become a prevalent issue, and emerging AI technology further accerlates the generation of misinformation. In this study, we investigate how humans perceive AI-generated and human-written news differently and whether they can distinguish between the two. We conducted an experiment that asked participants to evaluate a news dataset consisting of 16 articles of different authenticity (True or Fake) and origin (Human or AI-generated). Physiological signals, including gaze and heart rate data were captured during the study for analysis. The goal was to predict how humans perceive human- and AI-generated news differently based on the collected physiological data. Various data analysis techniques were used to better understand physiological responses and news perceptions. The feasibility of predicting the origin of news, whether it is human- or AI-generated, and whether it is true or fake news based on the user data was assessed. Additionally, we explored how users' general personality and behavioral traits may relate to their ability to classify the news correctly.

Files

ThesisReportFinal_Lian_Wu_5354... (pdf)
(pdf | 4.64 Mb)
Unknown license