This paper provides a hybrid approach to the Named Entity Recognition (NER) in Manipuri language. The hybrid approach so used is the combination of statistical approach (Conditional Random Field, CRF) and rule-based approach. The rule-based approach helps in defining various unique word features that are used in accurately classifying the Named Entities by the CRF classifier. With small corpus size, this hybrid approach proves to have Recall, Precision and F-score of 92.26%, 94.27% and 93.3% respectively. © 2013 Springer-Verlag.
CITATION STYLE
L, J., & Kaur, D. (2013). Named entity recognition in Manipuri: A hybrid approach. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8105 LNAI, pp. 104–110). https://doi.org/10.1007/978-3-642-40722-2_11
Mendeley helps you to discover research relevant for your work.