论文部分内容阅读
将范例推理技术应用于问卷设计知识重用过程中,提出了基于范例推理的问卷设计知识重用系统的应用框架,研究了问卷设计CBR系统中的一些关键技术:范例的表示、范例组织、范例检索算法、范例修正优化和重用、范例学习等.范例检索采用KNN法与遗传禁忌算法相结合的混合检索机制,通过建立GA适应度函数模型以及选择、禁忌交叉、禁忌变异算子等自动确定权重.我们组织和建立了初步的范例知识库,进行了相关实验.结果表明,融合遗传算法的范例检索方法可较大程度上缩短设计时间、提高问卷设计效率和质量,并有效地支持问卷设计的知识重用和知识管理创新.
In the process of reusing knowledge of case-based reasoning, we put forward the application framework of knowledge-based knowledge reuse system based on case-based reasoning, and studied some key technologies in the design of CBR system: demonstration of examples, organization of examples, and retrieval of examples , Sample revision optimization and reuse, example learning, etc. The example retrieval uses a hybrid search mechanism combining KNN method and genetic tabu search algorithm, and automatically determines weight by establishing GA fitness function model, selection, taboo crossover, taboo mutation operator, etc. We Organized and established a preliminary example of knowledge base, carried out the relevant experiments.The results show that fusion genetic algorithm based on the sample retrieval method to a greater extent, shorten the design time, improve the efficiency and quality of the questionnaire design, and effectively support the knowledge reuse of questionnaire design And knowledge management innovation.