A Model For Predicting Influential Users In Social
A Model For Predicting Influential Users In Social
复制标题
预测社交中有影响力用户的模型
DOI：
--
复制
发表时间：
2014
期刊：
影响因子：
--
通讯作者：
D. Prashanth
中科院分区：
文献类型：
--
作者：
Network Sriganga;Ragini Krishna;D. Prashanth
在最近的过去和设想的全球商业场景的未来，社交网络平台，如：Face Book，Twitter，Google plus+，LinkedIn，Orkut等，被设想为获得信息，观点，喜好和不喜欢，个人资料匹配等。这些输入是非常重要的参数，以便为企业和组织设计、设计和提供许多营销和CRM战略。在计算理论中，社会网络可以用树或图的定义结构来表示。节点网络可以是静态的，也可以是动态的，因为它通过添加或删除节点和边而在一段时间内演变。社会网络中有影响力成员的研究是社会网络分析中的一个重要研究问题。为了找到最有影响力的人，必须确定最中心的节点。中心性是对最具影响力的节点的度量，它是根据中心性度量来度量的。不同的研究人员对中心性度量或中心性度量的变体给出了不同的定义，如：度中心性、贴近度中心性、图中心性、度中心性、动态中心性、α-中心性、特征向量中心性、页面排名、Katz状态得分等。在该工作中，根据考虑不同可调参数值的动态中心度分数，将注意力集中在动态网络上。然后，基于动态中心度得分，可以宣布社交网络中最具影响力的个人。中心性；动态中心性；图论；社会网络分析。
In the recent past and in the envisaged future of global business scenario, social networking platform such as : Face book, Twitter, Google plus +, LinkedIn, Orkut etc are visioned to get information, opinions, likes and dislikes, profile matching etc. These inputs are very essential parameters in order to design, devise and deliver many of the marketing and CRM strategies for corporates and organizations. In computational theory a social network can be represented in terms of defined structure either by tree or a graph. The network of nodes could be either static in nature or dynamic in nature as it evolves over period of time by adding or deleting nodes and edges. The study of influential members in a social network is an important research question in social network analysis. In order to find the most influential person, the most central node has to be identified. Centrality is the measure of most influential node, which is measured in terms of centrality metrics. There have been various definitions given by different researchers for centrality metric or variants of centrality metric, such as : degree centrality, closeness centrality, graph centrality, between-ness centrality, dynamic centrality, α-centrality, Eigen vector centrality, page rank, Katz Status score etc. It has been observed that most of the existing methods for measuring centrality metrics are suitable for static networks and the existing methods of computation of centrality either underestimate or overestimate centrality of some nodes. In this work concentration is laid on dynamic network in terms of dynamic centrality scores considering different values of tunable parameter. Then based on dynamic centrality score the most influential individual in a social network can be declared. Keywords— Centrality, Dynamic Centrality, Graph Theory, Social Network Analysis.