生物语义相似性在数据仓库中的应用与实现  

THE APPLICATION OF SEMANTIC SIMILARITY IN BIOINFORMATICS DATA WAREHOUSE

作　　者：夏燕[1] 张忠平[1] 朱扬勇[1] 

机构地区：[1]复旦大学计算机与信息技术系,上海200433

出　　处：《计算机应用与软件》2005年第11期5-7,共3页Computer Applications and Software

基　　金：国家‘863’高技术研究发展计划(编号:2002AA231011);上海市重大科技项目(编号:02DJ14013)的支持

摘　　要：通过有效手段从整合的异构生物信息资源中获取高质量的生物数据,能够为生物信息的分析和挖掘提供强有力的支持。本文基于B ioDW———一个整合的生物信息数据仓库系统,以Gene Ontology作为语义模型,应用信息的语义相似度计算方法,获取不同基因产物生物特征的相似度;考虑更多的生物结构信息,使相似性的计算更加符合基因产物自身的生物特征。基于信息的语义相似度计算方法与单纯的相似度计算相比,具有扩展性、准确率高等优点,为生物信息的进一步研究奠定了坚实的基础。The analysis and mining of bioinformation can be strongly supported by means of acquiring high quality bio-data efficiently and conveniently from integrated heterogeneous bioinformation resource. This article apply an information-based semantic similarity measure to compute similarity between biology characteristics of gene products, using Gene Ontology as semantic model, based on BioDW an integrated bio-data warehouse. We take more bio-structure information into account in order to make it much more accordant with computation of semantic similarity in biology characteristics of gene products. The computation approach of information-based semantic similarity has more advantages than original algorithm, such as easily extending, high precision and so on ,which establishes stable basis for further research of bioinformation.

关 键 词：生物语义相似性 数据仓库 生物信息资源 数据挖掘 生物特征 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]