A Chinese entity extraction method based on maximum entropy principle

丁丰,袁保宗
DOI: https://doi.org/10.3321/j.issn:1001-8360.2001.05.007
2001-01-01
Abstract:In this paper, the maximum entropy principle is used to extract Chinese entity from free text. Nowadays information extraction plays important role in information processing, and entity extraction is the essential research area in information extraction. At First, the concept of entity, general entity extraction method and metric of performance evaluation are introduced. After that, based on maximum entropy principle, a stochastic model for entity extraction is established. The experimental results show that the proposed approach can improve performance of entity extraction.
What problem does this paper attempt to address?