An Environment for detection of Bugs through SVM

Mining technique finds hidden patterns from the data stored in the repositories and turn it into useful information and knowledge. Most open source software development projects include an open bug repository in which users of the software can gain full access. Training data size for bug priority classification using SVM.In this paper, calculate the precision and recall of different features categories for each class. The bug report having status value NEW, UNCONFIRMED, ASSIGNED bugs are not included in our training .because the priority level of these classifications may not be authentic. A trigger’s task is to manage the bug repository so that it contains only real bugs and important bugs are addressed quickly. We also use the precision and recall that can be used to measure the accuracy of the classifier. For a bug priority classification we need high precision and recall especially for higher priorities.


INTRODUCTION
Software has become critical to advancement in almost all areas of human endeavor. The art of programming only is no longer sufficient to construct large programs [15]. There are serious problems in the cost, timeliness, maintenance and quality of many software products. BUG: Software bug is the common term that are used to describe an error,mistakes,failure or fault in a computer program or system that produces an incorrect or unexpected result [4].

Types of Bugs:
1. Arithmetic bugs: Arithmetic overflow and underflow, Division by zero. 2. Logical Bugs: Infinite loops and infinite recursions, 3. Syntax bugs: Use of wrong operators. 4. Multithreaded programming bugs: Deadlocks [11] 5. Interfacing bugs: Incorrect hardware handling. 6. Performance bugs: High computational complexity of algorithms. The results show that our approach can significantly improve the efficiency of bug assignment the bug resolver is often identified with fewer tossing events [10].
Bug triage: Bug triage is an important step in the process of bug fixing. The goal of bug triage is to assign a new coming bug to the correct potential developer [3]. The existing bug triage approaches are based on machine learning algorithms, which build classifier from the training sets of bug reports. In practice, these approaches are suffered from the large scale and low quality training sets [7]. The training set reduction with both future selection and support vector machine for bug triage. We combine feature selection with instance selection to improve the accuracy of bug triage. We evaluate the training set reduction on the bug data of eclipse [13]. For the training set, 70% words and 50% bug reports are removed after the training set reduction. The experiment result show that the new and small training sets can provide better accuracy than the original one.

Error reported by developer
Open bug repository collects the bug reports Then we resolve the bug that whether it is not a duplicate bug (Bug triage) Eclipse bug report Mining the bug data that is present in bug repositories Classification algo are used SVM

Feature categorization
Find the best category according to priority Result indicate the feasibility of classification technique Coding J u n e 0 2 , 2 0 1 5 Static analysis phases can be used to improve the quality of dynamic bug detection tools created with SUDS and could be used to expand to find defect statically [11].The instrumentation engine is designed in a manner that allows users to create their own [9].A machine learning approach to develop a bug priority recommender which automatically assigns an appropriate priority level to newly arrived bugs, so that they are resolved in order of importance and an important bug is not left untreated for a long time [6]. Our result shows that the role of users goes beyond simply reporting bugs: their active and ongoing participation is important for making progress on the bugs they report. Based on the results, we suggest for ways in bug tracking systems can be improved [12]. We demonstrate our approach on the large open source project Mozilla that offers great opportunities to compare results and validate our approach.

METHODOLOGY:
In this research I am present a classification based approach to train the bug classifier through SVM for managing open bug repositories. Trigger is the method that resolves the important bug first. Support vector machines build non linear classification models from training data for each class. SVM performs classification tasks by constructing hyper planes in a multidimensional Space that separates cases of different class labels. SVM has shown the promising results for classification. To evaluate the bug priority we use precision and recall measures. Support Vector Machine present a comparison to evaluate which category classifier performs Better. It can be check according to the priority. The highest accuracy is achieved with SVM when categorical and text features are combined for training. . For SVM the results of bug priority classifier are better when the categorical and text features are combined for training the classifiers. That is also known as CSTF feature.CSTF is a combination of categorical, summary and text features that contains the relevant information that is useful for determining the bug priority.

RESULTS:
1. Click on www.eclipse.org and then click on tabular report.in order to generate the reports.

2.
Then choose the status of bug.