The Research On User Preference Modelling Based On Product Usage Data

With the development of sensor technology and network communication technology,the concept of interactive design has expanded from the interface design of computer application to the design of interactive products that support people's daily work and life.It has been widely used in industrial design,product design,information design and other fields.Interactive design can record users' behavior and product status when products are used by computer chips,sensors,wireless networks and other technologies.With the development of interactive design,more and more usage data can be collected in the products operation,including human-computer interaction data,product operation status data,user feedback data,etc.These data can objectively reflect the performance and status of product,and the usage context of users.User preference modeling is used to establish the relationship between product selection set and different product attributes.User preference analysis can obtain user's choice behavior at different product attributes level,which is widely used in product development,market segmentation,brand competition analysis and price strategy,etc.However,the data sources for user preference modeling mainly consist of questionnaires and online review data at present,monitoring data such as human-computer interaction data and product operation status is not included,if these data can be applied for user preference modeling,not only can user preference information be obtained timely and accurately,but can be transformed into information and knowledge related to design parameters to support product design and improvement,thus effectively improving product performance and user satisfaction.In order to solve the above problems,a user preference modeling method based on usage data proposed in this paper,which is used to establish the relationship between usage data from various sources and user preference.User preference modeling mainly includes two stages,the first stage is user experience modeling based on usage data.User preference are influenced by many factors,for product design,especially for interactive product design,researchers more focus on the impact of user experience on user preference.In this paper,convolutional neural network is used to construct the relationship between the operation data and user experience evaluation,and further extract the operation data features for different user experience dimensions as the measures of specific user experience.In the second stage,the idea of artificial neural network is used to improve the structural equation model,an improved structural equation model is proposed to establish the relationship between user experience and user preference.Finally,in order to prove the effectiveness of the proposed method,the user preference modeling of smartphones is taken as an example to show the whole process.The results show that the proposed method can effectively establish the mapping relationship between usage data and user preference.Through the user preference model,not only the changes of user preference and user experience during products utilization can be obtained timely and accurately,but also the main product performance parameters affecting user preference can be analyzed,which can provide decision support information for product design and improvement.