Deep Learning-Based Player Behavior Modeling and Game Interaction System Optimization Research

This paper proposes a deep learning-based player behavior modeling and interaction system optimization method. By constructing a BiLSTM-Attention behavior recognition model, it realizes modeling and behavior classification of high-dimensional time-sequence operation data, and dynamically adjusts interaction strategies and response parameters in combination with the classification results to optimize the game feedback mechanism. The experiments are conducted on a large-scale player behavior dataset collected from the actual game environment, and the evaluation results show that the method outperforms the traditional model in terms of accuracy, click precision, response delay and user satisfaction, which verifies the effectiveness of the proposed scheme in improving the adaptability and smoothness of the interaction system, and it has a good prospect of application and popularization value.