Research On Team And Event Organization Algorithm Based On Social Network

The team and event organization problem is a classic combinatorial optimization problem,which has been extensively studied in the field of operations research.How-ever,with the rapid development of network communication and the popularity of var-ious social platforms,the team and event organization problem in the background of social network has once again attracted the attention of many researchers.Due to the combination with the social network,the problem of team and event organization in the social network is different from the traditional version,so that it can no longer be solved simply by the approximation algorithm of classical problems such as set coverage or knapsack problem etc.Therefore,based on the full investigation and integration of real-life scenarios,for the first time,we officially defined social network-based team or-ganization revenue communication ratio maximization problem,social network-based event organization satisfaction maximization problem and social network-based travel planning utility maximization problem.Subsequently,we prove that these three prob-lems are all NP-hard problems,and there is no polynomial time approximation scheme for social network-based team organization revenue communication ratio maximization problem and social network-based event organization satisfaction maximization prob-lem.In order to solve the social network-based team organization revenue communi-cation ratio maximization problem,we designed two algorithms:expert heuristic and project heuristic.For the social network-based event organization satisfaction maxi-mization problem,we designed the satisfaction gain heuristic algorithm and the forward test optimization heuristic algorithm.For the social network-based travel planning util-ity maximization problem,we designed the utility heuristic algorithm and the similarity heuristic algorithm.Finally,we conducted a large number of experiments based on simulation and real data sets.The experimental results show that the algorithms we de-signed are better than the benchmark algorithm,and the experimental results on the real data set are similar with the experimental results on the simulation data set under the same conditions,and the effectiveness of our proposed algorithm is affirmed from the other hand.In particular,the project heuristic algorithm can achieve better results than the expert heuristic algorithm,but the corresponding cost is longer running time;the forward-test optimization heuristic algorithm can achieve not only better results than the satisfaction gain heuristic algorithm,but also the required runtime is shorter;and compared with the utility heuristic algorithm,the similarity heuristic algorithm is more suitable for the occasion of a large number of users.