Prediction of defect density Conclusion

             Stated that get the data from different companies and build a model to detect the defects by using transfer learning method. In this described the Transfer Naïve Bayes algorithm, in the algorithm uses the training data and test data, in training data used the data which are to be test and later on the data are weighted and model was built and showed that this algorithm is best than the AUC that is the Area Under the Receiver Operating Characteristic Curve, the classification method reduce the cost of testing and increase the software efficiency. stated that there are the statistical significance tests to compare six machine learning algorithms and using design metric and requirement metric and the combination of them. In this used k star algorithm and confusion matric that have true positive, true negative, false positive and false negative aspects and recall and precision to describe the results of detection of defects.

            Stated that there were proposed a data driven methodology to detect the defects and improve the quality of the software and reduce the cost of the product, in this build an estimation model to detect the defects at design phase of System Development Life Cycle and increase the performance of the System Development Life Cycle. In the data driven methodology data drive from different software and detect which technique is best to handle the defects in the software. stated that the reuse of the code in the software may cause different defects. The defects that occur in previous software delivers in the next software and the percentage of software defect increases if the defective code use in multiple software then the quantity of defective codes increased to handle this problem cloning of the codes is very necessary the code clone detect the plagiarism in the software and it is helpful to reduce the defects in the software.

                Said that the design pattern at web based application and its uses. In this used the features of the application as the quality of the application if the use of the application is valuable for user than the application can be said to be qualitative and also described that the usability is the feature of design patterns and described the three programming patterns by using reengineering and the uses of qualitative results. Stated that the graph mining approach to design the best pattern for object oriented systems in this was used the sub-graph , reuse patterns, anti-pattern, domain specific pattern and clones at design levels and also described the draw backs of this methods such as the uses of multiple graphs is overloaded for Random Access Memory. The object oriented methodology is used for past twenty five years and this is very efficient approach.

MATERIALS AND METHODS                       

            The software design pattern is helpful to design the patterns of the software. If the software is design in well manner way this is very great opportunity for software designers, developers, testers, implementers and many others to understand the software requirements and many other things related to the software and the quality of the software is increase. There was many other benefits to the use of the software design pattern like the cost and effort to develop the software was decreased and the performance, efficiency and reliability of the software increased.

In this research prepare a questionnaire in well manner and conduct a survey from different software houses and collect the data and create hypothesis that is null hypothesis and alternate hypothesis. After collecting the data apply some statistical analysis to check the validity of the results. The research questions are:

Q1: Helpful to the developers versus fault density in classes

Q2: Design pattern increases the quality versus helpful in the documentation

Q3: Easy in testing versus effectiveness of software

Q4: Defective free software versus defect density in design motif

RESULTS AND DISCUSSION

Statistical Analysis:

            This chapter describes the results and discussions about the survey that was conducted and describe how much the density of defects can be measurable and predictable to develop the qualitative software due to the design pattern. In the statistical analysis use the chi square test that shows the association between the variables

                The analysis represents that there is a significant association between these variables because the p value is less than the 0.05 that is the value of significant level. In this the hypothesis Ho is rejected and the hypothesis H1 is accepted. So that after analysis it is proven that the design pattern is helpful for the developers and the classes that is participating in the design and that is not participating in the design is helpful for the developers to predict the defects in the design.

                The analysis represents that there is a highly significant association between these variables because the p value is 0.00 that is less than the 0.05 that is the value of significant level. In this the hypothesis Ho is rejected and the hypothesis H1 is accepted. It is proved that the design pattern is helpful in making the documentation and makes less effort upon the documentation and concentrates to the development of the software and increases the quality of the software.

            The analysis represents that there is a significant association between these variables because the p value is 0.003 that is less than the 0.05 that is the value of significant level. In this the hypothesis Ho is rejected and the hypothesis H1 is accepted. After the statistical test it is proved that due to the design pattern the testing for the developers of the software become easier and in this way the effectiveness and usage of the software is valuable.

            The analysis represents that there is a significant association between these variables because the p value is 0.005s that is less than the 0.05 that is the value of significant level. In this the hypothesis Ho is rejected and the hypothesis H1 is accepted. After the statistical test it is proven that in the design pattern the design motif is valuable to detect the defects in the software and if the most probably defects predict firstly than most chances is occur to develop the defect free or less defective software.

CONCLUSION

        The design pattern provides the easy understanding of the design of the developers and the developers can easily communicate with each other due to design pattern because the design pattern are provide the solutions of the problems that are faced during the development of the software, coding of the software and making the documentation of the software. In this research, conduct a survey that represents the how much the use of design patterns in common area and how much the design pattern are helpful to design the software, develop the software, predict the defects in the software and improve the quality of the software. The research is qualitative. The results and discussion presents the descriptive and statistical analysis test. In the descriptive test describe the percentages and accumulative percentages that mostly agree that the design pattern are helpful to detect the defects in the software, increase the quality of the software and helpful to the developer by proving the easiness, understanding of the design, making the documentation of the software and saving the cost, time and effort of the developers.