Analysis of co-occurrence relationship between named entity in Web page

In order to analyze the closeness of named entities in massive web pages, the word co-occurrence algorithm FDC(frequency, term distance, co-collection ratio) is employed to evaluate the co-occurrence relationships between the named entities by their co-occurrence frequency, relative position and the ratio of co-occurrence among a document. And by employing the proper value of named entities´ co-occurrence frequency and the relative distances between the two named entities, the FDC algorithm is improved. Experiments show that the improved FDC algorithm has better performance.