Genetic Approach based Bug Triage for Sequencing the Instance and Features

In software industry analyzing bug by various tester and developer is a costly approach. So collecting these bug reports and triage is done manually which consume time with high rate of error. Here proposed work has focus on this triage of the bug reports by reducing the dataset size. In order to reduce cost of bug triage proper sequencing of the instance and feature selection is done. Here instance and feature selection are clustered by using list of words, keywords and bug id as fitness function parameters. Two stage learning genetic algorithm named as teacher learning based optimization was used for clustering. As genetic algorithms are unsupervised learning approach, so new set bug report triage is adopt by the proposed work. Experiment is done on real dataset of bug reports. Result shows that proposed work is better on precision value by 38.5% while execution time was reduce by 29.2% as compared with existing procedures.