结合协同过滤与隐语义模型的视频推荐策略

Video Recommendation Strategy Combining Collaborative Filtering and Implicit Semantic Model

观看视频已逐渐成为人们休闲娱乐方式之一,但视频网站却面临着如何从用户的需求出发更好地推荐视频资源的困惑.传统的协同过滤算法仅通过用户之间的相似度来建立对物品的兴趣关系,忽略了用户行为所带来的隐含信息以及物品之间的分类信息.因此本文在传统协同过滤算法的基础上融合了隐语义模型进行推荐,借助隐语义模型增加推荐结果的多样性,并借助协同过滤算法保证推荐的及时性,通过两种算法的融合能够有效地提高推荐的多样性并保证推荐的性能.

Video has gradually become an important way of leisure and entertainment for people today.But nowadays video websites are facing the problem of how to better recommend video resources based on the needs of users.Traditional collaborative filtering algorithms only use the similarity between users to establish interest in items,Ignoring hidden information brought by user behavior and the classification information between items.Therefore,this paper integrates an implicit semantic recommendation model based on the traditional collaborative filtering algorithm,increases the diversity of recommendation results with the help of implicit semantic models,and ensures the timeliness of recommendations with the help of collaborative filtering.The fusion of the two algorithms can effectively improve the diversity of recommendations and ensure the performance of recommendations.