To date, numerous proposed clustering methods have been published in a long series of papers and included in several outstanding books. This book, however, as the authors say, is the first monograph on the topic of relational data clustering that addresses both fundamental aspects and specific applications.

Part 1, “Models,” consists of seven chapters. Following an introductory chapter, the next three chapters briefly present the main features of heterogeneous and homogeneous relational clustering. The subsequent two chapters discuss a more general framework for data clustering when heterogeneous relations, homogeneous relations, and attributes are considered in clustering multiple-type-related data objects. The final chapter of this part discusses some new data clustering methods based on evolutionary approaches.

Part 2, “Algorithms,” focuses on the computational aspects of relational data clustering, and presents suitable algorithmic developments for the previously presented theoretical approaches. The next two chapters present the nonnegative block value decomposition (NBVD) co-clustering algorithm to factorize the relational data matrix into the row-coefficient, block value, and column-coefficient matrices, together with the proof of its correctness and a new algorithm for identifying the hidden structure of a k-partite heterogeneous relation graph. Next, the authors develop the hard, soft, and balanced variants of the collective learning genetic algorithm (CLGA) as computational approaches in clustering homogeneous relational data. The authors then address hard and soft variants of the mixed membership relational clustering (MMRC) algorithm for data clustering algorithms in the framework of a probabilistic model under a large number of exponential family distributions.

Part 3, “Applications,” consists of five chapters, which present several applications of the previously presented algorithms.

The list of references contains some of the most relevant titles of published papers and books related to the topic of relational data clustering.

The broad conceptual framework for relational data clustering, together with the mature analysis of the computational approaches and specific applications, make the book an excellent graduate-level textbook. It is also a reference of crucial importance for researchers and practitioners who deal with clustering topics in various theoretical and applicative areas.