Named Entity Recognition (NER) is a crucial task in Natural Language Processing (NLP) that identifies and recognises named entities in large structured or unstructured texts. These named entities can be names of persons, locations, organisations, dates etc. We developed a POS-tag rule-based system that performs NER for Hindi. Rules were created using the POS tokens and gazetteers were used for identifying named entities. We also focused on creating gazetteers for person names, locations, and organisations in the Hindi language. The system is capable of identifying and classifying 5 named entities: person, location, organisation, date, and time. Our system was evaluated on the benchmark Naamapadam dataset for three entities-person, location, and organisation. The F1-score was 0.42 and the precision was 0.90.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Rule-Based Named Entity Recognition for Hindi

  • Devanshi Suri,
  • Nimisha Malviya,
  • Nisheeth Joshi

摘要

Named Entity Recognition (NER) is a crucial task in Natural Language Processing (NLP) that identifies and recognises named entities in large structured or unstructured texts. These named entities can be names of persons, locations, organisations, dates etc. We developed a POS-tag rule-based system that performs NER for Hindi. Rules were created using the POS tokens and gazetteers were used for identifying named entities. We also focused on creating gazetteers for person names, locations, and organisations in the Hindi language. The system is capable of identifying and classifying 5 named entities: person, location, organisation, date, and time. Our system was evaluated on the benchmark Naamapadam dataset for three entities-person, location, and organisation. The F1-score was 0.42 and the precision was 0.90.