基于非相关文献知识发现方法挖掘微生物的潜在致病性  

Microbial Pathogenicity Mining Based on Disjoint Literature-based Knowledge Discovery Methods

作　　者：冉升 钱领 高艳 

机构地区：[1]重庆西信天元数据资讯有限公司,重庆401127

出　　处：《预防医学情报杂志》2017年第3期291-295,共5页Journal of Preventive Medicine Information

基　　金：国家863课题"微生物数字资源知识管理系统构建及关键技术研究"资助项目(2014AA021503)

摘　　要：利用非相关文献知识发现模型,提出了挖掘微生物潜在致病性的知识发现方法。以ABC模型为基础,基因作为中间词,通过检索、排序、筛选、语义过滤等技术实现知识发现。本方法的特点在于中间词的获取中辅以适当的人工干预:对多亚基蛋白基因进行拓展,基因分组后分别获取目标文献。利用该方法进行了微生物潜在致病能力的知识发现实验,发现空肠弯曲菌(Campylobacter jejuni)可能引发尿脓毒症(urosepsis)。该方法可用于发现微生物潜在致病能力,指导医学研究和疾病防治。Based on the disjoint literature -based knowledge discovery model, we proposed a method to explore the microbial potential pathogenicity. Based on the ABC model, this method used the genes as intermediate word, to realize the knowledge discovery with searching, sorting, screening and semantic filtering. The most o- riginal feature of this method was the artificial intervention during the intermediate obtaining progress： expan- ding the multi - subunit protein genes and obtaining the target literatures after gene classification. We found the Campylobacter jejuni could cause urosepsis by this method. And it could be used to explore the potential patho- genicity of microorganisms and help medical research and disease control.

关 键 词：微生物 基因 非相关文献知识发现 

分 类 号：G250[文化科学—图书馆学]