基于模糊集理论的知识融合方法研究

在知识管理领域中, 根据知识源的特点, 通过融合处理多源知识, 形成新的知识层, 提高知识的内涵、层次和置信度, 并提升实现系统任务目标的能力。在基本知识层、方法层和思想层三级融合层次的基础上, 研究多源知识融合处理, 得出融合结构与框架。将信息融合处理技术移植到知识融合处理中, 形成基于模糊集理论的知识融合算法, 得出处理流程和基于Petri网的融合模型。利用基于模糊集理论的知识融合算法, 综合各预警模型的观察结果, 讨论企业失败预警问题。实例和仿真结果表明:知识融合判别方法降低了企业失败预警判别的不确定性, 知识融合在知识管理领域应用是有效和可行的。

Multiple Source Knowledge Fusion Technique Based on Fuzzy Sets Theory

In the field of knowledge management, new knowledge level can be formed via multiple source knowledge fusion according to the characteristics of knowledge source, while the connotation, level and confidence of knowledge, as well as the ability to fulfill the system mission can be enhanced. The structure and framework of knowledge fusion is obtained from research on multiple source knowledge fusion at the foundation of the three fusion levels of basic knowledge level, method level and thought level. By introducing information fusion processing into knowledge fusion, a knowledge fusion algorithm based on fuzzy sets theory is formed. Besides, the processing flow and a fusion model based on Petri net are advanced. By utilizing the knowledge fusion algorithm based on fuzzy sets theory, the corporate failure prediction problem is discussed by synthesizing the observation results of different prediction models. It is proved by real examples and simulation results that the uncertainty of corporate failure prediction is reduced by knowledge fusion determination methods and the application of knowledge fusion in the field of knowledge management is effective and feasible.