Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
June 13, 2026Scientific ReportsOpen Access

Transformer-based coreference resolution modeling for Amharic text

View Full Paper
Ask AI
Bookmark
Share

Authors

LALingerew Bantie Asmare

Discussion

Loading...

Member takes

Overview

Randomized trial models coreference resolution in Amharic, suggesting improved NLP capabilities.

Key Points

  • This study aims to develop an effective transformer-based model for coreference resolution in Amharic, a low-resource language.
  • Proposed a transformer-based approach using multilingual BERT (mBERT) for coreference resolution.
  • Created a novel annotated Amharic corpus with 312 documents and 18,763 mentions for training and evaluation.
  • Utilized standard coreference evaluation metrics such as MUC and B3 to assess model performance.
  • Achieved F1-scores of 80%, 85.71%, 90.9%, 88.86%, and 81.7% across various evaluation metrics.
  • Demonstrated superior performance over existing state-of-the-art Amharic coreference systems.
  • Showed that contextual embeddings effectively capture semantic relationships for Amharic coreference resolution.

Cite This Study

Lingerew Bantie Asmare (2026) studied this question.

synapsesocial.com/papers/6a2cf57cfaef96ed7f057895https://doi.org/10.1038/s41598-026-46130-8
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Amharic neural coreference resolution with multi-head attention and named entity recognition2026
  2. 2Optimized Text Embedding Models and Benchmarks for Amharic Passage Retrieval2025
  3. 3News Classification in Low‐Resource Languages: Insights From Transformer and Baseline Models2026 · 1 citations
  4. 4Bidirectional Transformer-Based Neural Machine Translation for Amharic and Tigrinya: Bridging Morphological Complexity and Data Scarcity2025
  5. 5Coreference Resolution Based on High-Dimensional Multi-Scale Information2024