Data Clustering

Chapter 20

Semisupervised Clustering

Amrudin Agovic

Reliancy, LLCSaint Louis Park, [e-mail]

Arindam Banerjee

University of Minnesota at Twin Cities,Minneapolis, [e-mail]

20.1 Introduction

Semisupervised clustering (SSC) has become an important part of data mining. With an ever increasing volume of data in several problem domains, it is more important than ever to leverage known information and observed relationships among data points to guide clustering.

Clustering methods are broadly divided into two groups depending on the data representation they use: feature-based, where each data point has a representation in terms of a feature vector or a structured representation such as sequence, time series, or graphs, and