Objectives

From the Course Objectives there are several key points. First, this unit should allow reading more about clustering research in IS &R. In particular, all of the key ideas of clustering are given, and a summary of the experimental studies to-date is given. Second, students should be able to select, implement, design and develop clustering algorithms to be used in IS &R systems. This should be possible since some students will code an algorithm, others will work through the clustering process manually, and the remainder will use some software package that supports clustering. Finally, students should be able to discuss and explain the main issues related to clustering and its use in digital libraries and other related information services.

This Unit has the following objectives, for students to:

1. be able to test to see if circumstances exist (document collection has certain characteristics, or particular functionality is desired from the retrieval system) in which clustering may be beneficial;

2. be able to determine clustering by hand from raw data;

3. be able to program at least one clustering algorithm.