基于一种改进的粒群算法的异构案例库语义检索策略

Semantic Search Strategy Of Heterogeneous Database Based On An Improved Particle Swarm Optimization

在现今,竞争情报具有很重要的作用,但企业不同使用的特征项不同,制约了案例之间相似度的比较。为此使用Ontology,以完成不同名称特征项的相似度计算,进行相似度的计算必需有好的权重分配的策略。在该项目中,为了达到好的效果,建了一套基于带极值扰动简单粒子群权重分配策略。根据实验结果,该策略能够有效的提高检索的准确率及查全率。

At present,competitive intelligence has an important role,but different companies used different features,and restricted the comparison of the similarity between the cases.so we adopt Ontology,to complete calculation of the similarity between different features,he calculation of similarity needs to be a good weight distribution strategy.In this project,in order to achieve good results,build on a weight distribution strategy based on tsPSO.According to the results,this strategy can effectively improve the accuracy of accuracy and recall.