Research on User Behavior of Smart Broadcasting Based on Big Data Analysis
This paper discusses the multi-dimensional prediction method of smart radio and television user behavior analysis,and proposes a user behavior prediction model combining Collaborative Filtering(CF)and deep learning.The model achieves accurate prediction of user behavior through four stages:user preference mining,temporal behavior prediction,social impact modeling and multi-dimensional feature fusion.The empirical study shows that the method is superior to the traditional model in accuracy,accuracy,recall,F1 value and ranking quality,which provides a new idea for intelligent broadcasting and television operation decision.
smart broadcastingbig data analysisuser behavior prediction
NETL
NSTL
万方数据
