论文部分内容阅读
针对虚拟社区的知识管理问题,提出一种知识地图管理模型。利用信息检索和数据挖掘技术创建和维护知识地图,实现虚拟社区知识管理的动态可增量更新,避免了随知识的增长而需要重新构建整个知识地图的缺陷。此外,用精准率和召回率来评价模型创建功能,并首次提出纯粹性、多样性、特异性和SAI四项指标用于评价模型的维护功能。基于SARP和ADICC两种文档训练集测试验证,各项指标均能达到合理范围。
Aiming at the problem of knowledge management in virtual community, a knowledge map management model is proposed. The use of information retrieval and data mining technology to create and maintain a knowledge map, to achieve dynamic incremental updates of knowledge management of virtual communities, to avoid the need to rebuild the entire knowledge map with the knowledge growth defects. In addition, the accuracy of the model and the recall rate were used to evaluate the model creation function. For the first time, the four indexes of pureness, diversity, specificity and SAI were proposed to evaluate the model maintenance function. Based on the training set of SARP and ADICC training set test and verification, all the indicators can reach a reasonable range.