Detection of Code-Smells by Using Particle Swarm Optimization Technique (PSO) ****selected paper from International Conference On Computing (NECICC-2K16)

Hemalatha, Anandarao, Radhikaraju, Ramesh
2016 South Asian Journal of Engineering and Technology   unpublished
Code-smell is a hint that depicts something is going wrong somewhere in the program. If these smells are neglected then they give rise to errors, anomalies and even failure of the software project. The base for code-smells are from design-smells that are found in design phase of the software development life cycle. Some of the design-smells like BLOB, functional decomposition and spaghetti smells are detected by using evolutionary algorithms called genetic algorithms and genetic programming
more » ... e algorithms are executed by using search-based approach for detecting the smells. This approach searches each and every line of the program for finding smells. By this approach, there is a lack of optimization in the programs. To overcome this inability, paper addresses one of the popular optimization techniques called Particle Swarm Optimization (PSO). By using these technique one of the code-smells called lazy classes are identified and detected. PSO searches the smells very intelligently based on the severity of the smells that are found. These optimizer reduces the search entanglement of the smells. Thereby, efficiency of the programs increases. The experimental results shows the accuracy found between design and code smells in the graph format.
fatcat:g2w5couxurf5zdi4ibmtrhsmx4