Agha Ali Raza

Publications/Natural Language Processing & LLMs for Low-Resource Languages

Journal · 2024

Urdu Paraphrase Detection: A Novel DNN-based Implementation using a Semi-Automatically Generated Corpus

Hafiz Rizwan Iqbal, Rashad Maqsood, Agha Ali Raza, Saeed-Ul Hassan

Natural Language Engineering (NLE), vol. 30(2), 2024 · IF 1.9

Summary

Detecting when two sentences say the same thing in different words underpins plagiarism detection, but Urdu resources for it were small and hand-built. We present a semi-automatic pipeline for generating paraphrase corpora and use it to build a corpus of 3,147 Urdu sentence pairs, manually tagged as paraphrased or not. We also propose two neural approaches; the stronger, D-TRAPPD, achieves F1 scores of 96.80 for paraphrase detection and 88.90 for text reuse and plagiarism detection, outperforming previous methods. The corpus, models, and code are freely available.

Cite this paper

Hafiz Rizwan Iqbal, Rashad Maqsood, Agha Ali Raza, Saeed-Ul Hassan. Urdu Paraphrase Detection: A Novel DNN-based Implementation using a Semi-Automatically Generated Corpus. Natural Language Engineering (NLE), vol. 30(2), 2024.

@article{iqbal2024urdu,
  author = {Hafiz Rizwan Iqbal and Rashad Maqsood and Agha Ali Raza and Saeed-Ul Hassan},
  title = {{Urdu Paraphrase Detection: A Novel DNN-based Implementation using a Semi-Automatically Generated Corpus}},
  journal = {Natural Language Engineering},
  volume = {30},
  number = {2},
  year = {2024},
  url = {https://aghaaliraza.com/publications/2023--NLE--Urdu_Paraphrase_Detection.pdf}
}

More on this theme

All 18 works on this theme →