An investigation of the accuracy of code and process metrics for defect prediction of mobile applications

Mobile applications have been around for a long time and proved to be a new excited market where everyone want to engage themselves. They have become more important than Web pages nowadays. Companies are giving more preference to mobile apps as compared to Web sites because of their user friendliness , better visibility and ease of social networking. This paper compares static code metrics and process metrics for predicting defects in an open source mobile applications. Correlation coefficient, mean absolute error and root mean squared error with process metrics as predictors are significantly better than with code metrics as predictors. Also the combined model based on process and code metrics is better than the model based on code metrics. It is shown that process metrics based defect prediction models are better for mobile applications in all 7 machine learning techniques used for modelling.