Improving Design Quality of Software Using Machine Learning Techniques

Design patterns commonly pre-owned in Object-Software oriented Development. It provides Information helping to overcome recurrent design challenges (problems). Recognition of patterns can be described as identification and representation of information based on knowledge obtained from statistical data derived from patterns. Detection of design patterns provides the advantages of an extensive investigation of object-oriented software applications. When using various pattern detection techniques depending on static research issues to identify patterns it detects similar pattern structures. For this, we seek to identify patterns in design of Software by using classification-based methods and information metrics. The JHotDraw, QuickUML, and Junit of object-oriented software are used to detect software problems in the patterns of designing a Software. The datasets are manual creations which are designed utilizing software criteria for classifying learners. It is carried out in two steps, namely the development of a statistics-oriented dataset and the identification of trends in software trends in software design. Tests are carried out by Usage of 3 open source computer programs to test the process, then the tests are examined.