游戏设计中基于DBN的用户体验评估模型研究

目的为了更直观、有效地评估游戏产品的用户体验(User Experience,UX),消除单一评估标准的不确定性。方法从传统的MDA游戏设计的角度出发,引入用户的生理特征测量,构建基于动态贝叶斯网络(Dynamic Bayesian Network,DBN)的用户体验评估模型。该模型通过MDAUX框架提取用户体验影响因子,作为贝叶斯网络的输入层节点,通过生理特征测量方法提取用户的脑电和眼动状态,作为贝叶斯网络输出层节点,以一阶隐马尔可夫模型(Hidden Markov Model,HMM)表示两个相邻时间片上用户体验元素的影响关系,从而动态地展示用户体验状态。结果通过生理特征测量实验验证该模型的可行性,通过建立知识平台实践了模型的应用。结论结合生理特征测量的用户体验评估模型可有效反映用户体验状态。

User Experience Evaluation Model Based on DBN in Game Design

The work aims to evaluate the user experience of the game more intuitively and effectively, so as to eliminate the uncertainty in single evaluation standard. From the perspective of traditional MDA game design, the biometric feature measurement of user was introduced to construct a user experience evaluation model based on Dynamic Bayesian Network. User experience elements were selected from the MDAUX framework as the input layer nodes of the Bayesian network. EEG and eyelid movement state of the user were selected as the output layer nodes of the Bayesian network through the biometric feature measurement. The first-order Hidden Markov Model represented the influence relationship between the user experience on two adjacent time slices, thereby realizing a dynamic evaluation of the user experience. The feasibility of the model was verified by biometric feature measurement and a knowledge platform was established to practice the application. The user experience model based on biometric feature measurement can effectively reflect the status of the user experience.