# Advantages And Disadvantages Of Fuzzy Clustering

3280 Words14 Pages
OPTIMAL SENTENCE CLUSTERING USING HIERARCHICAL FUZZY RELATIONAL CLUSTERING INTEGRATED WITH ARTIFICIAL BEE COLONY ALGORITHM

Miss.Prasanthini.R1 Dr.Santhi.V2
PG Student Associate professor
PSG College of Technology Department of Computer Science
Department of Computer Science PSG College of Technology
Coimbatore Coimbatore
E-mail id: prasanthini962@gmail.com E-mail id: sannthi79@gmail.com

Abstract

Clustering is the process of aggregating or grouping data items. In many text processing activities the role of data clustering is inevitable. Sentence clustering plays a vital role in text mining and text processing activities. In real world, same idea can be conveyed using different sentences and for this reason
The prototypes or mixtures of Gaussians based conventional Fuzzy clustering approaches do not represent sentence similarity measures in a common metric space as well as it requires the data to be Euclidean. This becomes a disadvantage and for this reason relational fuzzy clustering algorithm is used, which takes relational data as input. But the documents usually represent information in a hierarchical structure. The fuzzy relational algorithm forms simple or flat partition clustering which gives us a single set of clusters. Hence there is no particular organization or structure within them. But there are cases where one cluster may be a sub cluster of other cluster. (At the time of clustering images, flowers should form the super cluster with roses and marigolds as the sub cluster) So for this reason hierarchical clustering is proposed, which is depicted by a tree diagram or…show more content…
Basic flow diagram

Clustering famous quotations

The famous quotation data set is used in this paper. The context provided in this data set is rich and challenging since it contains a lot of semantic content and are often couched in poetic use of language.

Hierarchical clustering
In every clustering technique the data items are classified based on some properties. Hierarchical clustering uses distance matrix as the clustering criteria to build the hierarchical structure. A hierarchical tree is a nested set of partitions represented by a tree diagram or dendrogram. Its working is based on the union between two closest clusters or to generate sub cluster of a cluster.

Hierarchical Fuzzy Relational Clustering (HFRC)
HFRCA algorithm is a recent renowned algorithm for sentence clustering and is capable of identifying sub clusters. The algorithm proceeds with the similarity measure calculation between the sentences. After which the PageRank is calculated, using which the sentences are clustered.

Artificial Bee colony