从语义相关性角度分析超链归纳主题搜索(HITS)算法,发现其产生主题漂移的原因在于页面被投影到错误的语义基上,因此引入局部密集因子LDF(Local Density Factor)的概念。为了解决Web内容的重叠性,基于切平面的概念提出了一种新的主题提...从语义相关性角度分析超链归纳主题搜索(HITS)算法,发现其产生主题漂移的原因在于页面被投影到错误的语义基上,因此引入局部密集因子LDF(Local Density Factor)的概念。为了解决Web内容的重叠性,基于切平面的概念提出了一种新的主题提取算法(CPTDA)。CPTDA不但可以发现用户最感兴趣的主题页面集合,还可以发现与查询相关的其他页面集合。在10个查询上的实验结果表明,与HITS算法相比,CPTDA算法不仅可以减少30%-52%的主题漂移率,而且可以发现与查询相关的多个主题。展开更多
A new kind of algorithm for semidefinite programming is presented by using theGolden Section Method, which sufficiently utilizes structure of the set of feasiblesolutions. Because of the simplicity and less operations...A new kind of algorithm for semidefinite programming is presented by using theGolden Section Method, which sufficiently utilizes structure of the set of feasiblesolutions. Because of the simplicity and less operations, the algorithm can be usedto solve large scale problems. As a successful application, a numerical example ofMax-cut problem is given.展开更多
文摘从语义相关性角度分析超链归纳主题搜索(HITS)算法,发现其产生主题漂移的原因在于页面被投影到错误的语义基上,因此引入局部密集因子LDF(Local Density Factor)的概念。为了解决Web内容的重叠性,基于切平面的概念提出了一种新的主题提取算法(CPTDA)。CPTDA不但可以发现用户最感兴趣的主题页面集合,还可以发现与查询相关的其他页面集合。在10个查询上的实验结果表明,与HITS算法相比,CPTDA算法不仅可以减少30%-52%的主题漂移率,而且可以发现与查询相关的多个主题。
文摘A new kind of algorithm for semidefinite programming is presented by using theGolden Section Method, which sufficiently utilizes structure of the set of feasiblesolutions. Because of the simplicity and less operations, the algorithm can be usedto solve large scale problems. As a successful application, a numerical example ofMax-cut problem is given.
