Team Formation In A Social Network Integrating Minimal Communication Cost And Recommendation Based On Ranking

Team formation is a topic that has been studied for quite a while now especially in social networks. Many methods and approaches have been suggested and put forward with varying degrees of success and efficiency. In this research thesis we study the problem of finding more efficient aspect of this classic problem by basing our work on a subset of a larger social network of skilled individuals to perform the task and also using a novel recommendation approach within the pool of individuals as to which team will work together best. This is team formation with recommendation. The members of the recommended team should not only have the needed skills for the task or project, but can also work ef-fectively together as a team and based on some features, the members of the network should be able to recommend each other. Experiments conducted based the DBLP dataset show that this framework is dependable practically and gives useful and intuitive results. The critical question of features some more complex expert team formation systems will need to consider when making recommendations for team membership is attempted here. A survey of one of the features suggested in contemporary research is provided. An analysis of these characteristics using a real-world data set is conducted to determine how the sample feature is relevant in team ranking and choosing the best team.This thesis combines the power of expert recommendation systems (ER) with the cost reduction functions of the team formation problem.