This paper proposes an Information Gain (IG) -Decision Tree Matrix (DTM) Classifier (IG-DTMC) method for open-source code, which improves accuracy by optimizing the selection of software metrics. For the software metric subset selection, a subset of metrics is collected by the application of the IG model. Here, the relevant and irrelevant metrics are separated by the using those gathered subset of metrics. Then, high-dimensional (HD) data are handled by the classification method named, DTM. Here, different metrics are grouped into a network of classifiers by using the collected relevant metrics. Utilizing the resultant bound (lower and upper) matrix supports the Decision Tree (DT) Classifier in the collecting the similar software metrics in the same classifier. Here, the final weights of the accessible software metrics are visualized in the resultant bound (lower and upper) matrix. Using an open-source code dataset, the following metrics like accuracy (ACC) of software metric selection, the time required for software metric selection, software maintainability, software project code size, and the number of lines of code, the efficiency of the IG-DTMC method is assessed.
