Abstract: Text Processing is essential to organize data or to extract needful information from a heap of available Big Data. Sentence clustering is one of the processes used in Text mining task. Text document may contain Hierarchical structure which relate to more than one theme at a same time. Hence Hierarchical Fuzzy Clustering Algorithm can be used for clustering such text data. The paper presents a novel Hierarchical Fuzzy Relational Eigenvector Centrality-based Clustering (HFRECC) Algorithm which is extension of FRECCA Algorithm. It solves the problems like complexity, sensitivity and changeability of clusters and is useful for natural language document (NLP) and operates in Expectation-Maximization Framework and is capable to identify overlapping clusters. The algorithm uses graph representation of data and works on relational data provided viz., data in pairwise similarities among data objects.

Keywords: HFRECCA, FRECCA, Page-rank, Sentence clustering, Expectation-Maximization.