基于大数据分析的智慧广电用户行为研究
Research on User Behavior of Smart Broadcasting Based on Big Data Analysis
魏君对1
作者信息
- 1. 瑞安市华数广电网络有限公司,浙江 温州 325200
- 折叠
摘要
探讨智慧广电用户行为分析的多维预测方法,提出一种融合协同过滤(Collaborative Filtering,CF)与深度学习等技术的用户行为预测模型.该模型通过用户偏好挖掘、时序行为预测、社交影响建模及多维特征融合4 个阶段,实现了对用户行为的精准预测.实证研究表明,该方法在准确率、精确率、召回率、F1 值及排序质量等多个评估指标上均优于传统模型,为智慧广电运营决策提供了新思路.
Abstract
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.
关键词
智慧广电/大数据分析/用户行为预测Key words
smart broadcasting/big data analysis/user behavior prediction引用本文复制引用
出版年
2024
NETL
NSTL
万方数据