"Shivkal: A 1,500-Pair English\u2013Marathi Question Answering Dataset for Historical Knowledge Retrieval"
Abstract
"This dataset presents a collection of 1,500 question\u2013answer pairs developed for research in English\u2013Marathi cross-lingual question answering and information retrieval. The dataset focuses on historical knowledge associated with the Maratha period and is intended to support computational research involving low-resource and morphologically rich languages.Each instance consists of an English question, a corresponding reference answer, and a thematic category. The dataset is designed to facilitate evaluation of cross-lingual retrieval, question answering, multilingual representation models, semantic retrieval, reranking, and retrieval-augmented generation (RAG) systems. It provides a compact benchmark for investigating how NLP and language-model-based systems handle historical knowledge when the information domain is associated with Marathi-language resources and the user query is expressed in English.The dataset can be used for academic benchmarking, model evaluation, cross-lingual NLP research, and development of retrieval and question-answering systems for Indian languages. It is particularly relevant to research addressing the challenges of limited resources, linguistic variation, and cross-language knowledge access."
Read the paper
Related studies
- Knowledge Management Perspective of Generative Artificial Intelligence
- Navigating Generative Artificial Intelligence Promises and Perils for Knowledge and Creative Work
- Characterizing the evolving research on enterprise content management
- Large language models and multimodal foundation models for precision oncology
- From Generative AI to Extended Reality: Multidisciplinary Perspectives on the Challenges, Opportunities, and Future of Educational Computing
- Generative Artificial Intelligence: A New Engine for Advancing Environmental Science and Engineering
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.