Ontology Based Turkish Named Entity Recognition

Autor: Okan Ciftci, Fatih Soygazi
Rok vydání: 2020
Předmět:
Zdroj: 2020 Innovations in Intelligent Systems and Applications Conference (ASYU).
DOI: 10.1109/asyu50717.2020.9259897
Popis: Named entity recognition (NER) is an actual problem in natural language processing. NER is applied for extracting entities such as person, location or organization. In this paper, it is expressed how to integrate ontologies to NER problem as a solution proposal. We implemented the overall operation by initially gathering essential Turkish entities, fetching the linked information from the ontologies about entity names, extracting features with this linked information and classifying the entities according to these features.
Databáze: OpenAIRE