Preference Identification Based on Big Data Mining for Customer Responsibility Management

Junzo Watada ； Arunava Roy ； Pandian Vasant

《International Journal of Intelligent Technologies and Applied Statistics》 13卷1期 (2020/03) Pp. 1-24

https://doi.org/10.6148/IJITAS.202003_13(1).0001

摘要

The customer relationship management (CRM) plays a pivotal role in running a company. Therefore, it is important to identify the customer preference of products and services provided by an industry and/or a company. The numerous research works are published on how to identify the customer preference and recognize the right customer. It is essential to select ideal customers in business. Moreover, a challenging issue is to analyze service quality. Also, determining the factors is essential, which influence consumer's perception of service quality. In this paper, we propose the hybrid optimization method of big data mining and provide a heuristic method to determine the consumers' preference even in the case where Arrow's contradict situations occur in big data analysis. Both the methods play a central role in analyzing a CRM case and the experiment provides the comparison of its efficiency with conventional methods. The paper consists of brief introduction of latent semantic indexing (LSI) and probabilistic LSI (PLSI). The next section is spent on how to identify group preference and raise the contradict problems in group decision making presented by Kenneth Arrow. The following section provides heuristic method to solve such conflict situation by an optimization approach. Finally, we will provide our application to gas station companies in China and show how the methods work.

關鍵字

Web mining ； Attribute reduction matrix (ARM) ； Latent semantic indexing (LSI) ； Singular value decomposition (SVD) ； Customer relationship management (CRM) ； Probabilistic latent semantic indexing (PLSI)