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：
为了提高个性化用户兴趣建模的准确率,对用户建模过程进行了优化。在计算文档相似度时,综合考虑特征词的语义关系以及在文档中的分布情况,引入加权语义网,提高了文档相似度计算精度;在计算兴趣度权值时,引入有效信息的概念及量化方法,以解决用户兴趣类权值计算过于主观的问题,并提出具体权值算法,提高了权值计算的准确性。实验结果表明,改进的方法在用户兴趣聚类和兴趣类别权值计算的准确率上都较以往方法有较大提高。
英文摘要：
In order to improve the efficiency of personalized user profile modeling,this paper optimized the process of user profile modeling. While computing the document similarity,semantic relations and considered the distribution of feature words integrated,bronght weighted semantic net in to improve the accuracy of document similarity computing. While computing interests category weight,to settle the problem that user interests category weight calculating was too subjective,introduced the concept and quantization of effective information,besides,increased accuracy of interests category weight calculating by introducing specific algorithm of weight. The experiment results indicate that the improved method has greatly heightened the accuracy in the respect of interests clustering and interests category weight calculating compared to previous method.
同期刊论文项目
同项目期刊论文
期刊信息
位置：