The Research Of Clustring Analysis's Application In Web Text Mining

The subject first do research on the key technology of web text mining deeply,and discuss gathering web page,segementation,word frequency statistic and coputering the weight of the feature.Then,we analyse web text clustering algorithm.Because the web data is dimensional and directional data.,we know the old clustering algorithm face the large challenge.So,based on the old clustering algorithm,we combine maximum entropy principle to build maximu entropy function model.We apply it in the web text clustering.It can avoid local minima and get local minima.At last,we build web text clustering model ande apply it to cluster web text.The experiment prove that the new algorithm can get a good result.