Web Structure Mining Based on Maximum Flow and Page Similar Value
Xiaojun Wu · Computer Technology and Development · 2011
Aiming to Web structure mining algorithm is easy for a topic drift and mutually strengthening relations among the hots of the problem,a method of hyperlink structure mining based on the maximum flow and the page similarity value is presented.On the basis of traditional HITS algorithm,this method introduced the page similarity value and adopted the Web communities identification based on the maximum flow to construct the models of feature vector space.And then the calculation eventually won the highest value of authority-set and hub-set by iterative method.Experimental results show that the method has better recall and precision,what's more it effectively inhibits the theme of Web structure mining algorithms drift,has some practical value.