基于聚类算法的视频推荐系统的算法研究

针对目前视频推荐系统功能少、推荐数据不全面,以及采集用户数据不全推荐能力弱,用户体验差等问题,提出了一种基于聚类算法的视频推荐系统算法。通过自主开发的聚类算法中的AP-聚类分层分析所获取到的不同的用户信息,以及不同的用户的视频的关注度信息,形成不同的用户视频信息聚类库,结合用户视频信息库中的历史数据,进行标签设置,自主进行计算权重,利用开发的聚类层进行用户对视频信息的喜好进行排序,生成喜好序列表,最后通过聚类分层模型计算出不同用户在聚类信息库中的喜好视频信息,进行标签和推荐。该算法成功的对不同用户的视频信息进行了标签和推荐,且推荐针对性较强,命中率较高,具有一定的学术研究价值和实际推广意义。

Study on Video Recommendation System Based on Clustering Algorithm

At present,video recommendation system has few functions,incomplete recommendation data,incomplete recommendation ability of collecting user data and poor user experience,etc.,a video recommendation system algorithm based on clustering algorithm is proposed.Through independent development of clustering algorithm to AP-hierarchical clustering analysis of the access to different users of information,and the popularity of video information of different users,it forms different user video information clustering library,based on historical data in the user’s video database,tag set,it calculates the weight independently,and it uses development of clustering layer user preferences to sort of video information,generate preference sequence table,finally through hierarchical clustering model calculates the different users in clustering the be fond of video information in the database,labels and recommendations.This algorithm successfully labels and recommends the video information of different users,and the recommendation is highly targeted with high hit rate,which has certain academic research value and practical promotion significance.