A Novel Fuzzy Based Approach For Clustering Sentence-Level Text
Fuzzy clustering algorithms allow patterns to belong to all clusters with differing degrees of membership. In
domains such as sentence clustering this is important. It is a novel fuzzy clustering algorithm that operates on relational input
data in which the data is in the form of a square matrix of pairwise similarities between data objects. A detailed study about
the Fuzzy Clustering Algorithm has been discussed in this paper. The fuzzy clustering algorithm uses a graph representation
of the data and operates in an Expectation Maximization framework in which the graph centrality of an object in the graph is
interpreted as likelihood. The task demonstrates that sentence clustering the algorithm is capable of identifying overlapping
clusters of semantically related sentences and it is for a great potential use in a variety of text mining tasks. Here we also
include results of applying the algorithm to benchmark data sets in several other domains.
Index Terms- Expectation Maximization, Fuzzy relational clustering, Graph Centrality, Natural Language Processing,
Similarity Measure, Sentence Level Clustering.