TY - GEN
T1 - Applying agglomerative fuzzy K-means to reduce the cost of telephone marketing
AU - Hsu, Ming Jia
AU - Hsu, Ping Yu
AU - Dashnyam, Bayarmaa
PY - 2011
Y1 - 2011
N2 - 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 pre-identification 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.
AB - 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 pre-identification 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.
KW - Agglomerative Fuzzy K-Means
KW - Clustering analysis
KW - Data mining
KW - Internet behavior
UR - http://www.scopus.com/inward/record.url?scp=80054863954&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-24918-1_22
DO - 10.1007/978-3-642-24918-1_22
M3 - 會議論文篇章
AN - SCOPUS:80054863954
SN - 9783642249174
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 197
EP - 208
BT - Integrated Uncertainty in Knowledge Modelling and Decision Making - International Symposium, IUKM 2011, Proceedings
T2 - 2011 International Symposium on Integrated Uncertainty in Knowledge Modelling and Decision Making, IUKM 2011
Y2 - 28 October 2011 through 30 October 2011
ER -