Research On Task Assignment For Joint Optimization Of Collaborative Cost And Execution Quality

Task assignment is one of the most important issues in social networks.Generally,the quality of task execution is mainly determined by the skill level of the participants.However,regardless of the relationship among members,only the skill level is considered when assigning tasks,it may lead to low execution quality.Therefore,task assignment considering both skill level and collaboration cost can guarantee the quality of task execution.Based on the different objectives,tasks are divided into collaborative tasks and competitive tasks.Collaborative tasks require the cooperation of participants to achieve a common goal,such as the launch of Chang'e IV,while competitive tasks require parties to fight or cooperate,and leverage multiple factors to achieve the fairness of the results.Obviously,the assignment of collaborative tasks should consider the skill level of participants and minimize the collaborative cost among participants,While competitive tasks should not only consider the skill level of participants,but also explore the influence of participants,maximize the collaborative cost among participants,and ensure the fairness of the results with high collaborative cost.In this dissertation,we conduct research on task assignment,which emphasizes collaboration cost and task execution quality,in social networks.The main work includes:1)We propose an execution quality maximization assignment algorithm for collaborative tasks.Based on concept lattice theory,which can discover cliques in large-scale online social networks,taking the close connection among clique members and the inherent requirements of collaborative tasks into account,the collaborative task assignment is modeled as an optimization problem.Due to the NP-hard nature of the optimization problem,a heuristic algorithm is proposed to solve the problem.The time complexity of the algorithm is t.2)We propose an joint optimization of member influence and collaborative cost for Competitive tasks.Taking the representativeness of candidate members,skill level and collaborative cost into account,combined with cliques generated by the concept lattice theory,we select the most influential members of each clique,and trade off the cooperationcost among candidates to ensure the fairness of task execution results.The competitive task assignment problem is modeled as an optimization problem.The optimization problem is proved as a NP-hard problem.Thus,a heuristic algorithm for joint optimization of member influence and collaborative cost is introduce.The time complexity of the algorithm is.3)The extensive experimental results on real datasets have verified the superiority of our proposed algorithms over benchmark algorithms in terms of intra-group collaboration cost,minimum intra-group collaboration cost,membership size and running time.