Applying agglomerative fuzzy K-means to reduce the cost of telephone marketing

  • Authors:

  • Ming-Jia Hsu;Ping-Yu Hsu;Bayarmaa Dashnyam

  • Affiliations:

  • Department of Business Administration, National Central University, Jhongli City, Taiwan;Department of Business Administration, National Central University, Jhongli City, Taiwan;Department of Business Administration, National Central University, Jhongli City, Taiwan

  • Venue:
  • IUKM'11 Proceedings of the 2011 international conference on Integrated uncertainty in knowledge modelling and decision making
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

This research utilizes marketing research database the Taiwan telecom itself has together with Agglomerative Fuzzy K-Means to proceed fuzzy clustering analysis. The database content includes online behaviors and basic properties of clients, such as online motive, online frequency, salary, and gender. First, we use descriptive statistics to determine the difference in online behavior among different client clusters; these differences among clusters comprise indexes. Next, we compare the obtained indexes with experts' judgments to verify the precision of each index. These indexes can be used to estimate client's mobile online hours and the adaptive tariff plan. In addition, while approaching different cases, sales personnel can specifically query on significant questions within the index. Moreover, using these preidentification indexes, prolonged question analysis, especially on illogical answers, can be avoided. This can result in time saving and increase the number of cases handled, causing an overall improvement in industry performance.