Cluster Analysis

Finding groups in data by computing distances or similarities between elements. In ReverseEngineering, typically used for remodularization, ObjectIdentification, and ArchitectureExtraction.

---

Overviews are provided by:

• V. Tzerpos and RicHolt. ACDC: An Algorithm for Comprehension-Driven Clustering. WCRE 2000, pp. 258--267. IEEE Computer Society.
• A. Lakhotia. A unified framework for expressing software subsystem classification techniques. JournalOfSystemsAndSoftware (March 1997), 211-231.
• T. Wiggerts. Using clustering algorithms in LegacySystems remodularization. In 4th Working Conference on Reverse Engineering WCRE (1997), IEEE Computer Society, pp. 33-43.

---

Tool support:

• Bunch, by SpirosMancoridis and BrianMitchell. See http://serg.mcs.drexel.edu/bunch/
• Statistical packages, such SPSS and S-PLUS.

---

The maths of cluster analysis is described by L. Kaufman and P. J. Rousseeuw. Finding Groups in Data: An Introduction to ClusterAnalysis. John Wiley, 1990.

Very readable, highly recommended.

---

A comparison with mathematical ConceptAnalysis is provided in IdentifyingObjectsUsingClusterAndConceptAnalysis.

---

CategoryReverseEngineering | Contributions by ArieVanDeursen