The Research Of Team Formation In Social Networks Based On Intimacy

Traditional team formation has been studied broadly.However,with various social platforms and network communication being popular.The team formation in the social network has aroused the enthusiasm of many scholars again.The team formation in social networks is very different from traditional one,it can't be implemented by traditional methods such as set cover,task assignment,or maximum matching.In a team,personal ability is very important,because it matters whether a project can be completed successfully.At the same time,people are concerned about the team's work efficiency.The study found that collaboration among team members is the foundation of a efficient team.As a result,when forming a team,besides focusing on the individual's ability,mutual influence between members should be considered.The common method is to introduce a factor i.e.the cost of collaboration that measures the efficiency of collaboration between team members.Low cost of collaboration represents a strong collaboration ability.Therefore,the purpose of team formation in a social network is to get the lowest cost of collaboration in a team.This thesis focuses on the team building approach based on leaders and takes the intimacy relationship between the leader and team members as the study subject.The Leader-Finding algorithm for finding team leaders in a social network and the two team building algorithms(intimacy-priority and distance-priority)established on the basis of the leader-finding algorithm are proposed.At the same time,since the intimacy relationship between team members in the social network needs to be updated with the dynamic changes of the network,this thesis proposes a proximity-based distance algorithm based on the idea of routing,and establishes a distance list for each node in the network.And for the team formation algorithm provides the basis of calculating the cost of collaboration to ensure the team's timeliness.In the end,we performed experiments on the DBLP data sets by applying intimacy-priority and distance-priority algorithms proposed in this thesis from three aspects of the team collaboration cost,the number of members,and the running time of the algorithm.The comparison with the Greedy-set-cover and RaresFirst algorithms proves that our algorithms are more effective and reasonable.