Research

Automatic Extractive Text Summarization for Ho Language

IEEE Xplore · 2023

Automatic extractive summarization system for Ho, an endangered Munda language of the Austro-Asiatic family with limited digital resources. Develops preprocessing pipelines and summarization methods tailored to the morphological characteristics of Ho, contributing tools and baselines for an extremely low-resource language.

Paper ↗

NLP Text Summarization Low-Resource Languages Endangered Languages

Amharic ATS: Graph-Based vs. Statistical Extractive Summarization

IEEE Xplore · 2023

Comparative study of graph-based and statistical approaches to automatic extractive text summarization for Amharic, a low-resource Afro-Asiatic language. Evaluates TextRank and LexRank graph methods against TF-IDF and frequency-based statistical baselines, contributing benchmark infrastructure for Amharic NLP.

Paper ↗

NLP Text Summarization Low-Resource Languages Graph Methods