A Knowledge Graph Approach for Analyzing Player Social Media Reviews

In recent years, as the gaming industry has developed, both the number of games and players have continuously increased. However, in recent years, numerous conflicts have arisen between game developers and players, leading to significant consequences for both parties. These incidents stem from a lack of effective communication between the two sides. Meanwhile, player communities have amassed a wealth of authentic reviews posted by actual players, reflecting their most genuine opinions about the games. To address this issue, the present study conducts sentiment analysis on reviews from player communities and proposes the use of graph neural networks (GNN) along with a knowledge graph constructed from gaming wikis to uncover the deep-seated reasons behind the various emotions expressed by players. Additionally, a pre-trained large model is employed to better understand player feedback, thereby enabling game developers to establish more effective communication with their player base.