The process of fixing bug is bug triage that aims to properly assign a developer to a new bug. Software companies pay out most of their expenses in dealing with these bugs. To reduce time and cost of bug triaging, an automated approach is developed to predict a developer with relevant experience to solve the new coming report. In proposed approach data reduction is done on bug data set which will reduce the scale of the data as well as increase the quality of the data. Instance selection and feature selection is also used simultaneously with historical bug data. Previously, text classification techniques are applied to conduct bug triage. The problem here is to get quality bug data sets as they are of very huge in size. In the proposed system, the problems of reducing the size and to improve the quality of bug data are addressed.First, pre-processing is done to the remove unimportant attributes and to identify missing terms. Then instance selection is combined with feature selection by using Dimensionality reduction technique to simultaneously reduce data size on the bug dimension and the word dimension. By using PSO algorithm, the reduction order is determined using fitness value. It is used to produce quality bug data set. The results show that the proposed system can effectively reduce the data size and improve the accuracy of bug triage. The proposed system provides an approach to leveraging techniques on data processing to form reduced and high eminence bug data in software improvement and maintenance.