Reasher And Realization Of Usability Evaluation Guide System Based On Cased-Based Reasoning

The usability evaluation has continued to mature and now it can be applied in many fields. With the increased application and maturity of technology, rapidity, effectiveness and rationality have generally become the necessary requirements of a good usability evaluation. The formulation of the plan largely depends on the designer's personal experience and knowledge. The inheritance and reuse of this knowledge and experience and getting knowledge and experience from typical cases can help the usability evaluation designers improve their working efficiency to a large degree.In this thesis the CBR technology is applied to the process of generating the usability evaluation cases and a usability evaluation guide system is brought forward based on the reasoning of cases. By studying and analyzing the present tool bases as well as the needs of the evaluation guide system, the thesis makes clear of the whole frame of the evaluation guide system and the working process. Moreover, the difficult and the key points of the evaluation guide system are identified based on the working process, that is the identification of the case retrieval feature sets and the study of the related case-based reasoning algorithm. These two essential parts has been discussed in the thesis in detail. In the part which studies the case retrieval feature sets, scores of typical cases of usability evaluation have been researched and analyzed based on the author's practical experience and the statistical data has been analyzed. In the end the author get the case retrieval feature sets of the usability evaluation guide system. When studying the case-based reasoning algorithm and the methods of its realization, the case retrieval is the core step of the system. This thesis analyzes several commonly used retrieval strategies and finally chooses the method of retrieval step by step, which uses the key words as the primary retrieval and the use KNN to calculate similarity. In the process of calculating similarity, weight distribution has a great influence on the whole calculation effects. This thesis chooses the Analytic Hierarchy Process to calculate the weight value rather than the experts grading method most in use in CBR system. This effectively avoids the influence of experts' subjective opinions on the efficiency of the system. Furthermore, the thesis designs some simple case retrieval functions of the evaluation guide system and studies its realization.After the studying of case retrieval feature sets and the case-based reasoning algorithm, the author has designed the interface of the main functions and complete the interaction flow of the main function. In the end, the thesis tests the core of the usability evaluation guide system, the case retrieval feature sets. And the results indicate that case-based reasoning usability evaluation guide system can provide customers the cases they need effectively and guide the users come up with a usability evaluation plan rapidly and efficiently. It provides a new way of learning for designers in user-centre design working.