软件网络中关键函数节点的识别方法

Identification method of key function node in software network

在考虑函数节点的传播能力以及节点的影响力基础上,提出一种识别软件网络中关键节点的算法。首先,定义了节点的传播能力,将节点的叶子节点的传播能力定为一个单位,层层向上递归,根据由上层节点向下传播的规律可以求得每个节点的传播能力。其次,针对软件网络中的节点的依赖函数集和调用函数集对当前节点的作用定义节点的影响力。之后,通过本文提出的算法综合考虑节点的传播力和影响力挖掘网络中的关键节点。该算法不仅考虑节点的局部特性以及整体结构,同时还考虑了网络自身的特性,因此对比其他方法而言更加准确。最后,通过在数据集上执行算法以分析验证上述算法有效性。

On the basis of considering the propagation capability and the influence of function nodes,an algorithm to identify key function node in software network are proposed and named IKN.First,the propagation capability of nodes is defined;the propagation capacity of leaf nodes is set as one unit,recursive upward layer by layer,according to the rule of downward propagation from the upper nodes,the propagation capacity of each node can be obtained.Secondly,the influence of the node dependent on effect which acts on the current node by set of dependent functions and set of call functions is defined.And then,through IKN algorithm,the key nodes in the network are excavated by considering the propagation capability and influence of nodes.The algorithm not only considers the local characteristics of the nodes and the overall structure,but also the characteristics of the network itself,so IKN algorithm is more accurate than other algorithm.At last,this paper verifies the effectiveness of the IKN algorithm by performing the algorithm on the dataset.