文章基本信息

标题：A Named Entity Recognition System Applied to Arabic Text in the Medical Domain
本地全文：下载
作者：Saad Alanazi ; Bernadette Sharp ; Clare Stanier 等
期刊名称：International Journal of Computer Science Issues
印刷版ISSN：1694-0784
电子版ISSN：1694-0814
出版年度：2015
卷号：12
期号：3
出版社：IJCSI Press
摘要：At the sixth Message Understanding Conference (MUC-6) in 1995, Named Entity Recognition (NER) was recognised as an essential sub field of information extraction and as an important contribution to natural language processing. The goal of NER is to extract specific predefined list of entities, which can include proper names, numerical expression and temporal expression. This paper introduces NAMERAMA which is a novel NER system based on Bayesian Belief Network (BBN). It extracts disease names, symptoms, treatment methods, and diagnosis methods from modern Arabic text in the medical domain. The results of the developed system shows that BBN performance is promising with 71.05% overall F-measure. The highest F-measure score was achieved in recognising disease names with 98.10% while the lowest was in recognising symptoms with 41.66%.
关键词：Named Entity Recognition; Bayesian Belief Network; Natural language processing; Machine learning.