Now a day’s software companies spend nearly 45% of cost in dealing with software bugs. A lot of time and money is wasted in finding solution to the bugs. To overcome manual work problems in this project we are using text classification techniques for automatic bug triage. For example if we find any bug while working on project we have to search in google for solution if we cannot find solution so much time and money is wasted. So in our project we are creating web application for commercial companies. In this website employee can query about bugs. Notification will go to rest of employee in the company if they are able to solve the bug they can update the solution to the bug repository. So finding solution to the bugs is easy. Noisy data may mislead the data analysis techniques while redundant data may increase the cost of data processing. In bug repositories, all the bug reports are filled by developers in natural languages. The low-quality bugs accumulate in bug repositories with the growth in scale. Illustration of reducing bug data for bug triage. presents the framework of existing work on bug triage. Before training a classifier with a bug data set, we add a phase of data reduction, which combines the techniques of instance selection and feature selection to reduce the scale of bug data. In bug data reduction, a problem is how to determine the order of two reduction techniques. based on the attributes of historical bug data sets, we propose a classification method to predict reduction. © 2016, International Journal of Pharmacy and Technology. All rights reserved.