Implementing subtopic recall and plummeting shadowing in Suffix Tree clustering

Mining data to find patterns has been studied for some time and current work on automatically grouping documents has roots in this research. We propose a dexterous partitioning strategy for Web search result presentation. We identify a canon to which a clustering must stick on to perk up the user´s search experience, avoid the redundant effect of query aspect recurrence, which is called shadowing. We present measures to quantify the shadowing effect, and we introduce an algorithm SSTC Subtopic Suffix Tree Clustering that optimizes the identified principles. The key idea of this is a dynamic reorganization of a clustering, similar to a faceted navigation system, We have also concentrated on the elimination of excessive clustering, i.e., either the number of cluster labels or the number of documents per cluster should not exceed the size of the result list Evaluations are done using the AMBIENT corpus and demonstrate the latent of our approach by a comparison with two well-known clustering search result algorithms STC and Lingo.