GPlagNet: A Deep Semantic Neural Framework for Gujarati Document Plagiarism Detection
Journal of Engineering · 01.10.2026 · 32 (10) · 52-74
DOI: 10.31026/j.eng.2026.10.03 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Engineering, online erschienen am 01.10.2026. Verfasst von Bhumi Shah und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
The problem of plagiarism detection in low resource languages is still an important one because of the scarcity of linguistic resources and the lack of ability to detect semantic similarity in traditional lexical matching methods. Gujarati being one of the major Indian languages, does not have any strong plagiarism detection system which can detect paraphrased or similar documents in the same context. In this research the authors introduce GPlagNet, a deep semantic neural framework for Gujarati document plagiarism detection which integrates language-specific preprocessing with transformer-based representation learning. The proposed framework integrates language-specific normalization, tokenization, stop-word removal, stemming, and lemmatization before generating contextual document representations using the ALBERT transformer. The bi-encoder architecture produces fixed-length document representations, which are compared using cosine similarity to compute semantic similarity scores and rank candidate documents using a Top-K retrieval strategy. The proposed framework produced a score of 1.00 for an identical document and semantic similarity scores of 0.8054 and 0.8606 for lexically different but contextually related documents, indicating that the model captures semantic relatedness beyond lexical overlap. The results show that GPlagNet was able to identify many semantically related documents under the paraphrasing conditions represented in the evaluation dataset, indicating its potential as a promising approach for academic integrity and document similarity assessment in low-resource languages.
Zitieren
APA
Shah, B., & Ameta, G. K. (2026). GPlagNet: A Deep Semantic Neural Framework for Gujarati Document Plagiarism Detection. Journal of Engineering, 32(10), 52-74. https://doi.org/10.31026/j.eng.2026.10.03
Vancouver
Shah B, Ameta GK. GPlagNet: A Deep Semantic Neural Framework for Gujarati Document Plagiarism Detection. Journal of Engineering. 2026;32(10):52-74. doi:10.31026/j.eng.2026.10.03
BibTeX
@article{Shah2026,
author = {Shah, Bhumi and Ameta, Gaurav Kumar},
title = {GPlagNet: A Deep Semantic Neural Framework for Gujarati Document Plagiarism Detection},
journal = {Journal of Engineering},
year = {2026},
volume = {32},
number = {10},
pages = {52--74},
doi = {10.31026/j.eng.2026.10.03}
}
EUVIMED