The Application Of Granular Computing In Clustering Analysis

Granular computing(GrC), named information granulation processing, is a newnotion about information processing and a normal form in computing intelligence. Itcovers almost with all research areas related to granularity involvingmethod, theories and technologies.It is a new hot issue rising in artificialintelligence. It imitates how people think.As we know, people can observe andanalysis problem from quiet different granularity,and can easily transfer from onegranularity world to another.It has been extremely widely applied in KDD domainetc.Data Mining is defined as a process that extracts connotative, unknown anduseful information and knowledge from practical data which issubstantial,incomplete,noise,ambiguous and stochastic.clustering analysis is aextremely active research domain and is one of the main methods about datamining.It is an unsupervised learning method:no pre-definition class, clustering isthe process of grouping a set of physical or abstract objects into classes of similarobjects.A cluster is a collection of data objects that are similar to one another withinthe same cluster and are dissimilar to the objects in other clusters.it has a widelyapplication in text clustering,financial analysis,data evaluation,gene research andmarket diagnosis and so on.While clustering analysis and granular computing are consistent in nature, howto apply granular computing to clustering analysis is still initial phase and acomplete, systematic and uncontroversial structure of GrC is still has not beenformed.This paper analyses GrC and clustering's profile,discuss principle ofgranularity in clustering and a generic framework of clustering based ongranularity.Based on the framework, we propose a text clustering algorithmGCBGD(granularity clustering based on grid density).The experiment results showthat it has high performance and feasible. Finally, we analyses fuzzy clustering,discuss some classical algorithms of fuzzy clustering and hierarchy structure based on quotient space theory and information granularity. Further more, the applicationof fuzzy clustering in text clustering is shown.