论文部分内容阅读
场景构造是机会发现过程中的关键活动之一.文中基于六度分离理论对事件的相关矩阵进行了p阶扩展,在对机会发现场景的结构进行形式化描述的基础上,提出机会发现活动中基于事件的p阶扩展相关矩阵,采用聚类分析的方法进行机会发现场景中事件簇构造,进而实现机会发现场景构造的思想与方法.对机会发现场景构造方法性能的评估进行探讨,明确了以效率系数作为机会发现场景构造方法的评估标准.实验表明,基于相关矩阵的机会发现场景构造具有较高的精度和效率系数,而基于3阶扩展矩阵的机会发现场景构造更适合在线机会发现的情形.
Scene construction is one of the key activities in the process of opportunity discovery.In this paper, based on six degrees of separation theory, the correlation matrix of events is extended by p-order. Based on the formal description of the structure of opportunity discovery scene, Based on the p-order extended correlation matrix of events, clustering analysis is used to construct event clusters in the opportunity discovery scenario, and then the idea and method of scene discovery in opportunistic discovery are discussed. The performance evaluation of the scenario discovery approach is discussed, The efficiency coefficient is used as the evaluation criterion for the method of constructing the scene of chance discovery.Experiments show that the chance discovery scene based on the correlation matrix has higher accuracy and efficiency coefficient, while the opportunity discovery scene based on the third-order extension matrix is more suitable for the situation of online chance discovery .