초록

용어보기논문에서 용어와 풀이말을 자동 추출한 결과로, 시범 서비스 중입니다. AI-Helper

Abstract ▼ AI-Helper

Named Entity Recognition is lo recognize meaningful named entities in texts and classify them into predefined classes. This job is the starting point for the relation extraction and the event extraction. Even though the accuracy of NER is high enough to stop researching for it, the results are mostl...

Named Entity Recognition is lo recognize meaningful named entities in texts and classify them into predefined classes. This job is the starting point for the relation extraction and the event extraction. Even though the accuracy of NER is high enough to stop researching for it, the results are mostly for researches with small amount of texts, not with real enormous texts. Thus, there are still many challenges to research. The accuracy of NER based on the machine learning is much affected by the accuracy of the learning model for the job. Likewise, the accuracy of the learning model depends considerably on the quality of training data. ln this paper, issues in annotation for training data are addressed and some guides for the issues are presented to improve the quality. ln addition, a NER model from modifying the idea of the decision tree is designed, which is expected to carry out NER with better accuracy and convenience to use and get feedback on the result of NER.

주제어