Finding the Effectiveness of Software Team Members Using Decision Tree

This paper presents steps taken in finding the effectiveness of software team members using decision tree technique. Data sets from software engineering (SE) students were collected to establish pattern relationship among four predictor variables-prior academic achievements, personality types, team personality diversity, and software methodology-as input to determine team effectiveness outcome. There are three main stages involved in this study, which are data collection, data mining using decision tree, and evaluation stage. The results indicate that the decision tree technique is able to predict 69.17 % accuracy. This revealed that the four predictor variables are significant and thus should consider in building a team performance prediction model. Future research will be carried to obtain more data and use a hybrid algorithm to improve the model accuracy. The model could facilitate the educators in developing strategic planning methods in order to improve current curriculum in SE education.