基于项目协同过滤的电视产品营销推荐模型

互联网技术的快速发展和应用拓展使我们迎来了三网融合的时代,为传统广播电视媒介带来了发展机遇。节目数据的剧增一方面丰富了电视节目的内容,另一方面却为用户选择带来了困难,这就要求电视运营商建立合理的个性化推荐模型。该文采用基于物品的协同过滤的个性化推荐算法,通过分析用户观看收视信息数据、电视产品信息数据,同时考虑到目前传统的互联网资源推荐系统大都是针对个体推荐,在对家庭不同成员的推荐时可能会出现适得其反的情况,综合考虑整个家庭成员的点播集合,构成了家庭用户完整的历史观看记录,分析每个家庭成员的偏好,建立电视产品营销推荐模型,做出节目的个性化推荐。同时对不同节目的标签进行组成分析,以数据图的形式更加直观地展示在结果中,用以了解不同时期标签的热度与关注度,从而进一步得出影视作品的热度,对不同时期的推荐偏好做出指导性建议。

The rapid development and application of Internet technology has ushered in the era of triple play,which has brought opportunities for the development of traditional broadcast and television media.The dramatic increase of program data has on the one hand enriched the content of television programs and on the other hand brought difficulties to user selection.This requires television operators to establish a reasonable personalized recommendation model.This article adopts a personalized recommendation algorithm based on item-based collaborative filtering,and analyzes users’viewing information data and TV product information data,taking into account that the current traditional Internet resource recommendation systems are mostly for individual recommendations,and are recommended for different family members.There may be counterproductive situations in which the ondemand collection of the entire family member is taken into account,constitutes a complete historical viewing record of the family user,analyzes the preferences of each family member,establishes a television product marketing recommendation model,and makes a personalized recommendation of the program.At the same time,the composition of the labels of different programs is analyzed and displayed in the results in the form of data graphs more intuitively to understand the heat and attention of the labels in different periods,thereby further obtaining the popularity of film and television works and recommending preferences for different periods.Make guidelines.