Abstract: Aiming at problems of traditional association analysis methods on webpages,this paper presents a new method of webpages association analysis based on named entities and relations.The new method uses named entities and relations,instead of traditional words,to express text feature.First,the new method adopts modify strategy based on vector similarity comparison to extract named entities on webpages.Second,it analyzes Maxfpminer algorithm and improves it.At last,the new method uses the improved algorithm to complete webpages association analysis.Results of experiment show that validity and readability of the new method's result excel to traditional methods.