Research of Barrage Analysis Based on Hypernetwork

Barrage analysis is a new way to explore hotpots of movie. Users' instant emotions, reactions to the exact frame users are watching that living comments offer make it unique advantages over traditional movie reviews. For the large and complex social labeling network in reality, the hyper-graph and hyper-network theory can provide a comprehensive research perspective to better describe the multi-dimensional relationships of data. Based on the barrage data of film “Busan trip”, this article takes “user” as a node, and “comment-content” as a hyper-edge to build a hyper-graph model. The simulation results demonstrate that based on ten similarity indicators, we can establish the user similarity matrix. The results is significant as in the matrix, the Greedy algorithm is applied to the user's community structure detection to obtain the corresponding comment labels.