Multilingual Deception Detection of GPT-generated Hotel Reviews

This repository contains the dataset and code for our paper.

Data

All data is available at all_data. Source label 0 represents real hotel reviews and label 1 represents fake/ LLM-generated hotel reviews.

Features

Topic Modeling features can be accessed interactively in topic_analysis

Models

GPT-4 generation

All generation code is available at LLM_generation.

Deception Detection models

XLM-Roberta, Random Forest and Naive Bayes models, together with interpretable features are available at Deception Detection Models.

Citation

@inproceedings{ignat-etal-2025-maide,
    title = "{MA}i{DE}-up: Multilingual Deception Detection of {AI}-generated Hotel Reviews",
    author = "Ignat, Oana  and
      Xu, Xiaomeng  and
      Mihalcea, Rada",
    editor = "Chiruzzo, Luis  and
      Ritter, Alan  and
      Wang, Lu",
    booktitle = "Findings of the Association for Computational Linguistics: NAACL 2025",
    month = apr,
    year = "2025",
    address = "Albuquerque, New Mexico",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2025.findings-naacl.88/",
    pages = "1636--1653",
    ISBN = "979-8-89176-195-7"
}