欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
为了解决推荐算法中用户标签稀疏、推荐准确度不高的问题,提出了一种基于用户标签的微博推荐算法。利用TextRank排序方法提取用户发布微博中的关键词,并对该关键词进行扩展,将其作为表示用户兴趣的标签;再根据微博的效应函数和生命周期形成待推荐的微博列表,计算用户标签及其同义词在待推荐微博列表中出现的次数,将出现次数较多的TOP-k条微博推荐给用户。通过实验验证,该算法能够有效地解决用户标签的稀疏性问题,并能提高推荐算法的准确性。
英文摘要：
In order to solve the problem of data sparsity of user＇ s tag in the recommendation and the low accuracy, this paper put forward a algorithm recommended by microblog which based on user＇ s tag. Firstly,it used TextRank sorting method to extract the keywords in the mierohlog that user released, and extended this keywords as a tag which represented the user＇ s interest. Secondly, it formed the list to be recommended according to microblog＇ s effect function and the life cycle. Finally,it calculated the count of the user＇ s tag and its synonyms that appeared in the list to be recommended, and recommended the top-k microblog which occurred most frequently to users. Through the experiment, this method can effectively solve the problem of data sparsity of user＇ s tag and improves the accuracy of the recommendation.
同期刊论文项目
同项目期刊论文
期刊信息
位置：