Text mining implementation in complaint management: A case study at Surabaya city office for population administration and civil registration (COPACR)

There is no doubt that customer satisfaction and service excellent have become one of the prime drivers for company’s continuous improvement, not least for the Surabaya City Office for Population Administration and Civil Registration (COPACR). To provide service excellent, managing customer complaints is critical. As a modern city, Surabaya has awarded many prizes nationally and internationally in many categories, for instance green and eco-city, child-friendly city, and cyber city. All of those prizes do not stop Surabaya City Government to further enhance the services for Surabaya people through the implementation of e-government. Surabaya COPACR as part of Surabaya City Government has implemented e-government to better serve Surabaya people. However, there are still many enquiries and complaints from users. Currently those complaints have been handled directly by staffs in order to provide solution, yet not analyzed thoroughly to find the root causes. In fact, finding the root causes of complaints would be beneficial to prevent repeated problems. In this paper, text mining technique are implemented to find better ways in managing complaints. Using clustering, especially the K-means algorithm, large and unpatterned complaint data has been grouped and displayed in cloud forms based on the identified root causes. Two main conclusions derived from the case study confirm that firstly, the implementation of text mining is powerful to obtain the root causes of the complaints in an efficient way. However, secondly, there are several weaknesses in the COPACR’s complaint receival that must be improved to better produce response and analysis. In a nutshell, text mining implementation will not bring maximum benefit without the improvement in complaint receiving and recording.