基于协同过滤算法的视频智能推荐系统

摘要

According to the properties of intelligent TV terminal, we proposed a collaborative filtering recommendation algorithm based on time period. According to the difference of the crowd of different time to watch TV, 24 hours were divided into several periods, and implemented a video recommendation algorithm for smart TV combining with the video attribute weights and the recommendation algorithm based on item, for precisely positioning for family members in the time to watch TV for dynamic recommended.%根据智能电视终端的性质, 本文中提出基于时间段的协同过滤推荐算法.根据不同时间段观看电视的人群不同, 将24小时划分为多个时间段, 并且结合视频属性权值和基于物品的推荐算法实现一种适合智能电视的视频推荐算法, 精准地定位该时段收看电视的家庭成员并进行动态的推荐.