Visualizing Time-based Weighted Coupling Using Particle Swarm Optimization to Aid Program Comprehension

Rully Agus Hendrawan, Katsuhisa Maruyama

Research output: Contribution to journalConference articlepeer-review


By knowing software coupling, developers can get better view of the software quality and improve their productivity in development and maintenance. This paper presents a method to visualize coupling network that are often very complex, using heuristic approach based on particle swarming optimization. Each node is placed randomly and assigned with initial speed. Node that are coupled together will be attracted each other and trying to get closer until they reach a particular distance. This distance is determined from the coupling value of two nodes. A closely related nodes will move closer until reaching a short distance. On each iteration, node position is dynamically updated based on attraction and repulsive force around them. Thus, gradually forming a near best solution of logical coupling graph. The coupling values are measured by mining the association rule from changes history. A software development project sometimes can be very active, updates happen within minutes. Sometimes it becomes slow with weekly or even monthly updates. Time-based weighted analysis method was used to accommodate these time sensitive situations. A co-change in a short duration will be weighted more than co-changes that happen in longer duration.

Original languageEnglish
Pages (from-to)597-604
Number of pages8
JournalProcedia Computer Science
Publication statusPublished - 2015
Event3rd Information Systems International Conference, 2015 - Shenzhen, China
Duration: 16 Apr 201518 Apr 2015


  • association mining
  • logical coupling
  • particle swarm optimization
  • program comprehension
  • source code visualization


Dive into the research topics of 'Visualizing Time-based Weighted Coupling Using Particle Swarm Optimization to Aid Program Comprehension'. Together they form a unique fingerprint.

Cite this