基于meta-analysis和等级全息建模的可信软件开发风险识别研究  

Risk Identification Research for Trustworthy Software Development Using Meta-Analysis and Hierarchical Holographic Modeling

作　　者：宋浩[1,2] 李建平[2] 蔡晨[2] 米传民[2] 李铭禄[3] 

机构地区：[1]山东财经大学数学与数量经济学院,山东济南250014 [2]中国科学院科技政策与管理科学研究所,北京100190 [3]国家自然科学基金委员会,北京100085

出　　处：《武汉大学学报（理学版）》2012年第3期260-268,共9页Journal of Wuhan University:Natural Science Edition

基　　金：国家自然科学基金重大研究计划重点项目(90718042);中国博士后基金(20090450588)

摘　　要：对软件开发过程中的风险进行有效识别,是提升软件可信性的重要方法和手段.本文以面向可信的软件开发过程风险辨识为研究对象,采用meta-analysis方法,结合软件项目开发风险相关的文献,采取高频词提取的方法,得出了前20大风险,并对需求、用户、供应商等前10大风险进行了重点阐述.给出了可信软件风险识别的等级全息框架,以产品、项目管理、项目结果、属性等10个视角全息辨识可信软件开发过程中可能面临的风险,与等级可变的不同风险情景识别过程相结合,为可信软件的风险识别构建了系统的解决思路和框架.To identify all the risks in software development is an important strategy to improve the trustworthiness softwares. This paper discussed the development of risk identification for trustworthy software. Relative literatures about software project risk were collected using meta-analysis and top 20 risks were extracted from risk descriptions according to the frequency sequence. The top 10 risks were described, including requirement risk, user risk, supplier risk etc. Hierarchical holographic modeling containing 10 subcategories like product, project management, project result, attribute etc was presented to identify all risks in trustworthy software development. Combined with the altera ble sub framework for different risk scenes, we presented a systemic solution framework for risk identification in trust- worthy software development.

关 键 词：可信软件 风险识别 等级全息建模 META-ANALYSIS 风险管理 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论]