Contributor
Debarghya Datta

DBpedia Hindi Chapter


Mentors
Ronit Banerjee, Ananya, Dr. Sanju Tiwari, Dr. Ronak
Organization
DBpedia
Technologies
python, virtuoso, sparql
Topics
natural language processing, knowledge graph, large language model, Relation Extraction, Low Resource Language
DBpedia, an ever-evolving repository of structured knowledge extracted from Wikipedia, has played a pivotal role in organizing information. Yet, its dependence solely on English Wikipedia limits its applicability in diverse linguistic environments. Our project seeks to rectify this by creating a DBpedia chapter sourced exclusively from Hindi Wikipedia. This endeavor not only ensures the continual enhancement and relevance of the Hindi ontology but also broadens its accessibility to a wider audience. Given the challenge of limited natural language processing (NLP) support for Hindi, a low-resource language, we will leverage the capabilities of Large Language Models (LLMs) to tackle multilingual text processing effectively. Through this effort, we aim to establish hi.dbpedia.org, a platform in Hindi, comprising a knowledge graph sourced from Hindi Wikipedia, a user-friendly web interface, and a SPARQL endpoint for seamless data querying.