Analysis of Human and AI-generated Text Classification

DOI

10.2991/978-94-6239-664-7_72How to use a DOI?

Keywords

AI; Human; Artificial Intelligence; Text; Generative AI. Prompting; ChatGPT; AI in Education; Natural Language Processing. Neural Network; Generated; AI-generated content; NLP

Abstract

Artificial Intelligence plays a vital role in generating Text that is used in our daily lives. However, the rapidly increasing usefulness of AI technology is a concern, according to need. This includes biases in the response of the generative AI. There are lots of generative AI that contribute to generating text, documents, articles, etc. However, the generative model is also concerned about the authentication of text. This paper surveys to collect human and AI responses to a question. The same question is set for AI tools to generate responses. This study evaluates and examines the linguistically and semantically important features of human and AI-generated text by using machine learning and deep learning classification. To achieve an optimal performance with the most suitable machine learning model, we also need a neural network model, which is based on deep learning. Then the response of this model classifies the text as human or AI-generated. The human-generated text will be much better than the AI-generated text. In the present digital world, AI captures 70% of information from humans, so it provides text like humans. In this paper, the Contribution of this novel work development of a methodology that can identify AI-generated text from human-authored text using different machine learning models. It has become significantly difficult to tell AI-generated text apart from human writing, and the existing detectors are restricted to minor datasets and superficial features. The study introduces a combined ML–DL approach that merges linguistic characteristics with VGG-16 semantic embeddings for the categorization of AI and human text. With the training on 487k samples, the model can reach 0.999 accuracy, thus proving to be very robust and able to generalize well. The findings validate the hybrid feature approach as being effective for the reliable detection of AI-generated texts.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Md. Rony
AU  - Jubair Khan Shahos
AU  - Adnan Mahmud Fuad
AU  - Sadman Sadik Khan
AU  - Asiful Islam
AU  - Md. Minhajul Islam
PY  - 2026
DA  - 2026/06/08
TI  - Analysis of Human and AI-generated Text Classification
BT  - Proceedings of the International Conference on Intelligent Data Analysis and Applications (IDAA 2025)
PB  - Atlantis Press
SP  - 1052
EP  - 1065
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-664-7_72
DO  - 10.2991/978-94-6239-664-7_72
ID  - Rony2026
ER  -