User Interest Modeling With Facial Expressions

Facial expression recognition is the extraction and classification of human facial expressions by computer,so that the computer can obtain the emotional information of the human,so that the mental activity state of the human can be inferred,and the advanced intelligent interaction of human-machine is realized.Facial expression recognition technology is a challenging cross-cutting topic in the fields of psychology,image processing,physiology,pattern recognition and machine vision.It provides reference and reference for comprehensive and stereoscopic recording of network user behavior and preferences,providing more accurate personalized information recommendation and network user group positioning,and predicting network user information behavior.When a user reads a related document,facial expressions such as neutral,happy,and dazed,etc.,characterize different immersive states of the user.By synthesizing the user's general behavior,body language such as facial expressions,and the content of the document itself,we can build a comprehensive,three-dimensional user interest model.This paper firstly based on the computer-side Web reading system,through the user experiment,collects the facial expression and the text content corresponding to the facial expression generated by the user during the browsing of Sina Weibo content.The collected facial expression image is subjected to visual feature processing,and the facial expression can represent whether the user is interested in the related content.At the same time,text information processing is performed on the reading document corresponding to the expression to construct a user interest space.Finally,the relationship between facial expressions and user preferences is explored in conjunction with facial expressions and interest spaces in textual content.The research in this paper shows that the facial expression generated by the user in the social network can reflect the user's interest in the reading process,and the user's facial expression can be used to establish the user interest model,which can achievebetter results.This paper introduces facial expression based on image processing into user research,which will help to expand user behavior attributes,deepen user behavior modeling,and enrich traditional user behavior research.