AI-Powered Fake Review Detection for College Admission Using BERT and DeBERT

Savitha P., Christopher R., Dinesh M., Jeevabharathi A.

2025

Abstract

Online audits plays a crucial role in influencing college admission decisions. However, the presence of fake reviews, whether excessively positive or misleading negative can distort perceptions and misleading prospective students. This model proposes an AI-powered fake audit discovery framework outlined particularly for college affirmation audits. The framework leverages Common Dialect Preparing (NLP) procedures and directed learning models to distinguish beguiling substances. A web plugin is created to analyze surveys in real-time, with the dataset collected from different college audit stages and preprocessed utilizing methods like stop word expulsion, lemmatization, and estimation investigation. Progressed models counting BERT, DeBERTa, and XLNet are prepared to classify surveys as either honest to goodness or fake, with DeBERT a conveying the most noteworthy exactness. This framework progresses straight forwardness within the college choice preparation, enabling students and guardians to form educated choices based on bonafide surveys, whereas too helping educate in keeping up their notoriety by sifting out deceiving or false surveys.

Paper Citation

in Harvard Style

P. S., R. C., M. D. and A. J. (2025). AI-Powered Fake Review Detection for College Admission Using BERT and DeBERT. In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25; ISBN 978-989-758-777-1, SciTePress, pages 244-248. DOI: 10.5220/0013880800004919

in Bibtex Style

@conference{icrdicct`2525,
author={Savitha P. and Christopher R. and Dinesh M. and Jeevabharathi A.},
title={AI-Powered Fake Review Detection for College Admission Using BERT and DeBERT},
booktitle={Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25},
year={2025},
pages={244-248},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013880800004919},
isbn={978-989-758-777-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25
TI - AI-Powered Fake Review Detection for College Admission Using BERT and DeBERT
SN - 978-989-758-777-1
AU - P. S.
AU - R. C.
AU - M. D.
AU - A. J.
PY - 2025
SP - 244
EP - 248
DO - 10.5220/0013880800004919
PB - SciTePress