NewsSumm: The World's Largest Human-Annotated Multi-Document News Summarization Dataset for Indian English

NewsSumm is a comprehensive multi-document summarization dataset specifically curated for Indian news articles. This dataset is designed to facilitate research in Natural Language Processing (NLP), particularly in the domains of text summarization, multi-document synthesis, and information extraction.

Dataset Overview:The NewsSumm dataset contains a large collection of news articles from diverse Indian sources, along with their corresponding summaries. The dataset is structured to support both extractive and abstractive summarization tasks.

Key Features:- Large-scale collection of Indian news articles- Multiple documents per topic/event- High-quality reference summaries- Diverse topics covering politics, sports, business, technology, and social issues- Suitable for training and evaluating state-of-the-art summarization models

Applications:This dataset can be used for:- Multi-document abstractive summarization research- Single-document summarization- Information retrieval and extraction- Cross-document coreference resolution- Fact-checking and verification systems- NLP model benchmarking

Developed by Suvidha Foundation as part of our ongoing research in AI and NLP for social impact.

Contribution of SyamPrasad Guda

SyamPrasad Guda made significant and sustained contributions to the development of the NewsSumm dataset. He has been an integral part of this research project since 2022, playing a key role in shaping the dataset into a high-quality, world-class resource. His efforts were instrumental in data curation, annotation quality control, validation, and continuous refinement of the dataset. Through his technical expertise, attention to detail, and long-term commitment to the project, we were able to ensure consistency, accuracy, and reliability across the dataset. The quality of results achieved in NewsSumm would not have been possible without his valuable contributions and dedication.