生物医学实体关系抽取的研究

人类基因组计划使生物医学的研究取得了前所未有的成就,在研究结果中得到了大量的生物医学实体,如基因、蛋白质、器官、疾病和药物等,但这些生物实体之间存在什么样的关系仍不完全清楚.作为生物医学研究成果载体的文献呈指数增长,已成为科研人员获取知识的瓶颈.文本挖掘能够解决信息超载问题,故对生物医学实体关系的挖掘流程和评价指标进行介绍,对生物医学文本挖掘在研究生物实体关系抽取中采用的基于统计的方法、基于自然语言处理的方法和基于模式匹配的方法进行了阐述,对各种方法进行了综合比较,同时介绍了国内外相关研究.

Relation extraction from biomedical entities

Human genome project has ushered in unprecedented achievements in biomedical field and a large number of biomedical entities have been found such as genes,proteins,organs,diseases and drugs.However,the relation between these entities still remains unclear.As a document carrier of achievements in biomedicine,the exponentially increased literature has become the barrier of scientists in obtaining knowledge.Since text mining can overcome the overload of information,its processes in relation extraction from bio...