Using Supervised Clustering Technique to Classify Received Messages in 137 Call Center of Tehran City Council

Abstract

Supervised clustering is a data mining technique that assigns a set of data to predefined classes by analyzing dataset attributes. It is considered as an important technique for information retrieval, management, and mining in information systems. Since customer satisfaction is the main goal of organizations in modern society, to meet the requirements, 137 call center of Tehran city council is planning to reduce the waiting time of customers, forwarding their messages to the appropriate service manager to increase service quality. The city council is currently using a manual approach dispatching textual request messages. Since this process is very similar to supervised clustering concept, in this study, we applied the Naïve Bayes algorithm, which is one of the most common algorithms in the supervised clustering arena to classify received messages regarding their subjects. The performance results of the proposed technique indicate its high efficiency in clustering the received messages with 98% accuracy.