Named Entity Recognition using Conditional Random Field for Kannada Language
Bhuvaneshwari C Melinamath
Bhuvaneshwari C Melinamath, Department of Computer Science & Engineering, SVERI’s, College of Engineering, Pandharpur (Maharashtra), India.
Manuscript received on 07 September 2019 | Revised Manuscript received on 16 September 2019 | Manuscript Published on 26 October 2019 | PP: 413-416 | Volume-8 Issue-11S2 September 2019 | Retrieval Number: K106609811S219/2019©BEIESP | DOI: 10.35940/ijitee.K1066.09811S219
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open-access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Named Entity Recognition (NER) is a significant errand in Natural Language Processing (NLP) applications like Information Extraction, Question Answering and so on. In this paper, factual way to deal with perceive Kannada named substances like individual name, area name, association name, number, estimation and time is proposed. We have achieved higher accuracy in CRF approach than the in HMM approach. The accuracy of classification is more accurate in CRF approach due to flexibility of adding more features unlike joint probability alone as in HMM. In HMM it is not practical to represent multiple overlapping features and long term dependencies. CRF ++ Tool Kit is used for experimentation. The consequences of acknowledgment are empowering and the approach has the exactness around 86%.
Keywords: CRF, HMM, NER, NLP.
Scope of the Article: Pattern Recognition