Software Clustering

ACDC Algorithm

mark (mark@undisclosed.example.com) — 2010-05-08T15:20:52+00:00

ACDC Algorithm History Developed by Vassilios Tzerpos. Algorithm Intent ACDC produces flat or nested decompositions of a software system. The algorithm assigns meaningful names to the clusters it creates. Factbase Properties The factbase includes only dependencies between

Evaluation of Clustering Algorithms

bil (bil@undisclosed.example.com) — 2010-05-06T14:15:20+00:00

Evaluation of Clustering Algorithms Software clustering researchers have developed several evaluation methods for software clustering algorithms. This research is important because * Most software clustering work is evaluated based on case studies. It is important that the evaluation technique is not subjective.

Agglomerative Algorithms

bil (bil@undisclosed.example.com) — 2010-05-06T14:14:19+00:00

Agglomerative Algorithms History Algorithm Intent Factbase Properties The factbase is a generic dependency data table where each row describes one entity to be clustered. Each column contains the value for a specific attribute. Clustering Objectives Process Description

LIMBO Algorithm

bil (bil@undisclosed.example.com) — 2010-05-06T14:12:48+00:00

LIMBO Algorithm History The algorithm was developed by Periklis Andritsos and Vassilios Tzerpos. Algorithm Intent To generate decompositions that exhibit the least information lost when entities are represented by their clusters. Factbase Properties

Bunch Algorithm

bil (bil@undisclosed.example.com) — 2010-05-06T14:08:36+00:00

Bunch Algorithm History Algorithm Intent Bunch produces flat decompositions of a software system that exhibit high cohesion and low coupling. Factbase Properties The factbase is a module dependency graph (MDG). Clustering Objectives The Bunch algorithm constructs software decompositions with maximum cohesion and minimum coupling.

Decompositions

bil (bil@undisclosed.example.com) — 2010-05-06T13:58:37+00:00

Decompositions The output of a software clustering algorithm is either flat decompositions or nested decompositions of software systems. Flat A flat decomposition does not contain nested clusters. There is one level of clusters and one level of entities. An example of flat decomposition is shown below.

Terminology

bil (bil@undisclosed.example.com) — 2010-05-06T13:57:35+00:00

Terminology Entity An entity is defined as anything recognized as being capable of an independent existence and which can be uniquely identified. In software clustering, an entity may be a class, a source file, a function call statement, or a software module.

Structure Evaluation

bil (bil@undisclosed.example.com) — 2010-05-06T13:54:38+00:00

Structure Evaluation Installation Instructions * Configure jret library * Download file * Ensure that the mojo package is in your CLASSPATH. * Run application using the following command java structure_evaluation.Flat or java structure_evaluation.Nested

jret.tar.gz

mark (mark@undisclosed.example.com) — 2010-03-21T20:50:27+00:00

refsim.tar.gz

mark (mark@undisclosed.example.com) — 2010-03-21T19:46:55+00:00

sample.tar.gz

mark (mark@undisclosed.example.com) — 2010-03-21T19:46:49+00:00

upmojo.jar

mark (mark@undisclosed.example.com) — 2010-03-21T18:15:20+00:00

structure_evaluation.jar

mark (mark@undisclosed.example.com) — 2010-03-21T18:15:04+00:00

src_samples.jar

mark (mark@undisclosed.example.com) — 2010-03-21T18:14:51+00:00

limsim.jar

mark (mark@undisclosed.example.com) — 2010-03-21T18:14:44+00:00

jret.jar

mark (mark@undisclosed.example.com) — 2010-03-21T18:14:34+00:00

congruity.jar

mark (mark@undisclosed.example.com) — 2010-03-21T18:14:16+00:00

conruity.tar.gz

mark (mark@undisclosed.example.com) — 2009-06-19T15:09:00+00:00