Title: A clustering method of marketing effective data based on relation matrix fusion

Authors: Linlin Zhou

Addresses: Henan College of Transportation, Zhengzhou 450000, China

Abstract: In order to overcome the problems of traditional clustering methods, such as low recall rate, low clustering accuracy and poor clustering efficiency, an effective marketing data clustering method based on relation matrix fusion is proposed. The relationship matrix fusion process is designed, and the effective data in the marketing data is selected according to the fusion results. Then, the feature units of effective marketing data are extracted, and the data clustering problem is transformed into a linear programming problem by calculating the EMD distance between the data. Finally, data clustering is completed according to the results of data integration. The experimental results show that the recall rate of effective marketing data is between 94.5% and 98.3%, the clustering accuracy is between 95.1% and 98.7%, and the maximum number of iterations is 900, which proves that the method achieves the design expectation.

Keywords: marketing data; valid data; relation matrix fusion; EMD distance; data clustering; Earth mover's distance.

DOI: 10.1504/IJISE.2023.132271

International Journal of Industrial and Systems Engineering, 2023 Vol.44 No.3, pp.380 - 390

Received: 10 Jun 2021
Accepted: 23 Aug 2021
Published online: 14 Jul 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article