Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
February 11, 2026PLoS ONEOpen Access

Enhanced extractive text summarization framework for low-resourced Urdu language

View Full Paper
Ask AI
Bookmark
Share

Authors

SNShahzad NazirBaba Guru Nanak UniversityMAMuhammad AsifUniversity of OkaraSAShahbaz AhmadSuperior University

Discussion

Loading...

Member takes

Implication

This research develops and evaluates text summarization models for the low-resourced Urdu language, improving information extraction.

Key Points

  • The aim is to develop effective text summarization models for the low-resourced Urdu language while preserving essential semantics.
  • Created a large-scale dataset of text documents and human-annotated summaries.
  • Extracted fifteen features for each sentence in the supervised approach.
  • Performed feature reduction to minimize computational complexity.
  • Employed multiple machine learning and deep learning models for summarization.
  • Utilized four different models in the unsupervised approach.
  • Achieved approximately 7% improvement in ROUGE scores for supervised models.
  • Achieved around 12% improvement in ROUGE scores for unsupervised models.
  • Demonstrated that the proposed approaches outperform existing methods for Urdu text summarization.

Cite This Study

Nazir et al. (2026) studied this question.

synapsesocial.com/papers/698c1cb3267fb587c655f42dhttps://doi.org/10.1371/journal.pone.0341596
View Full Paper
Ask AI
Bookmark
Share