Recognition And Prediction Of Flow Experience Based On Neural Network

Nowadays,the flow experience is attracting increasing attention of software and game designers.In order to identify and analyze the real-time flow experience without any interruption,we developed the BP neural network models to predict the flow experience by the physiological states and body postures during each game round.we recruited 70 participants to play mobile games,and recorded their physiological state,body posture,as well as the subjective ratings on the flow experience towards games.And we use this state as feed data for BP neural network models.Furthermore,the hierarchical linear model(HLM)were developed,trying to explain the relationship between the input and output variables.The results showed that(1)SDNN,LF positively predicted the flow experience;(2)head distance,head hand distance,chest distance negatively predicted the flow experience;(3)The prediction accuracy of head hand distance was the highest(80.525%)in body postures indexes.The prediction accuracy of HRV was the highest(84.211%)in physiological indexes.With a combination of body postures and physiological indexes,the prediction accuracy would reach to the highest(86.891%).The present findings provide a practical and effective approach as to how to recognize the flow experience level in mobile games processes.In this study,the neural network model is trained by using the combination of performance states and posture states.This indicator can identify the current level of flow experience without disturbing the subjects,and can upgrade the model continuously according to the feedback data of the subjects.Compared with the traditional measurement method,this method has the advantages of cheaper measurement tools and more convenient measurement process.It can be applied to the user experience research of mobile games,and provide reference for the design and development of mobile games.