Measuring and predicting the service quality of information systems and technology: an integrated approach of decision tree and random forest

The purpose of this research is to provide a new approach to measure, predict and ultimately improve service quality of all data and indicators affecting service quality simultaneously. In this research, two indicators of performance and gap of service quality have been used. The data mining model provided based on clustering has been used to segment and target users, and based on decision trees used to estimate service quality, and random forests have been used to predict service quality gap. The results of service with the decision tree showed that among the five indicators of service quality gap, guarantee, empathy and responsiveness indicators have more importance and among the five performance indicators, responsiveness time indicators, responsiveness ratio and problem-solving time have more importance. Knowledge and improving the quality of services extracted in this model can be implemented in various service organisations with different indicators.