基于决策树算法的可用性评估平台的实现  

Realization of Usability Evaluation Platform Based on Determine Tree Algorithm

作　　者：常晓红[1] 苏菲[2] 

机构地区：[1]中国科学院研究生院 [2]中国科学院软件研究所,北京100080

出　　处：《计算机仿真》2006年第9期53-57,共5页Computer Simulation

摘　　要：可用性是体现产品质量和市场竞争力的重要因素,可用性评估可以考察软件的效率、有效性、用户满意度,该文研究开发的可用性评估平台为软件开发过程的可用性评估提供专家知识支持。为了使可用性评估平台根据已有知识向用户提供具有价值的建议,使用改进后的启发式归纳分类算法:ID3决策树算法。根据对已有知识的信息熵计算,建立多叉决策树,然后利用交互过程中用户选择的节点信息得到用户的知识分类,向用户提供知识建议。系统在交互过程中,通过层层深入的交互方式,不断引导用户进行思考,做出选择,细化知识内容,排除冗余信息影响,明确具体属性值,实现主动的智能交互。通过改进的ID3决策树算法配合分层交互的方法建立的可用性评估平台系统能够有效地帮助用户完成软件的可用性评估工作,受到用户的认可。Usability is an important factor to the quality of products and market competition. Usability evaluation reflects efficiency,validity and user＇s satisfaction at software. The system of usability evaluation platform offers the knowledge for users to consummate the program of software engineering. The system improves on ID3 algorithm that is one of the most common and effective heuristic inductive clarification algorithm in determine tree algorithm, in order to support the knowledge to users. The system used it to learn the knowledge from the information that come from interaction, accounted the information entropy of knowledge to set up the determine tree, gets the type of users, acquired userg knowledge clarification, and supplied users with suggestive knowledge structure. In the interaction, the method of deepening and refining leads users to think and choose, so the system can clear the valuable attribute value. In this way, we can realize the intelligent interaction. To apply the ID3 determine tree algorithm, usability evaluation platform can assist users to effectually accomplish the work of usability evaluation. It was proved that the system is welcomed by users who want to evaluate software.

关 键 词：可用性评估 信息熵 决策树 智能交互 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]