Publications/Natural Language Processing & LLMs for Low-Resource Languages
Journal · 2025
Urdu Paraphrased Text Reuse and Plagiarism Detection using Pre-trained LLMs and Deep Hybrid Neural Networks
Hafiz Rizwan Iqbal, Muhammad Sharjeel, Jawad Shafi, Usama Mehmood, Saeed-Ul Hassan, Agha Ali Raza
Multimedia Tools and Applications, vol. 84(35), 2025
Summary
Commercial plagiarism checkers often miss copied text that has been reworded, and very little research addresses this for Urdu. We propose D-TRaPPD, a deep neural architecture that feeds representations from pre-trained language models into convolutional and LSTM layers to detect paraphrased reuse in whole documents. On two gold-standard Urdu corpora, the sentence-embedding variant achieves F1 scores of 91.77 on real cases and 95.15 on simulated cases of paraphrased reuse and plagiarism, outperforming earlier methods.
Cite this paper
Hafiz Rizwan Iqbal, Muhammad Sharjeel, Jawad Shafi, Usama Mehmood, Saeed-Ul Hassan, Agha Ali Raza. Urdu Paraphrased Text Reuse and Plagiarism Detection using Pre-trained LLMs and Deep Hybrid Neural Networks. Multimedia Tools and Applications, vol. 84(35), 2025.
@article{iqbal2025urdub,
author = {Hafiz Rizwan Iqbal and Muhammad Sharjeel and Jawad Shafi and Usama Mehmood and Saeed-Ul Hassan and Agha Ali Raza},
title = {{Urdu Paraphrased Text Reuse and Plagiarism Detection using Pre-trained LLMs and Deep Hybrid Neural Networks}},
journal = {Multimedia Tools and Applications},
volume = {84},
number = {35},
year = {2025},
url = {https://aghaaliraza.com/publications/2025--MTA--Urdu_Paraphrased_Text_Reuse_and_Plagiarism_Detection_using_Pre-trained_Large_Language_Models_and_Deep_Hybrid_Neural_Networks.pdf}
}
- IRTIQA: Evolving in English, Evaluated in Urdu — Does Agent Self-Evolution Transfer Across Languages?Findings of EACL 2027, Athens, Greece
- UrduCultureQA: A Culturally Grounded Benchmark for Urdu Question Answering and Alignment EvaluationFindings of EACL 2027, Athens, Greece
- Pak3H: Evaluating the Cost of Cultural Mismatch in LLM Alignment with a Human-Contextualized Urdu BenchmarkEMNLP 2026 (Main), Budapest, Hungary
- Evaluating Robustness of LLM-Based Ambient Scribes for SOAP Note GenerationCHIL 2026 (Conference on Health, Inference, and Learning), PMLR vol. 333
- Kahaani: A Multimodal Co-Creative Storytelling SystemEACL 2026 Student Research Workshop, Rabat, Morocco