Investigation of the Optimal Combinations of GPRS Services By Using Fuzzy Theory

摘要

The combination sale of various products can not only provide customers diversified choice, but also can further stimulate their requirements, and increase the sale quantities of products. Generally, people usually investigate the optimal combination of products by collecting customers' demand intentions through the market survey when these products are not launched yet or are only launched for a short period of time. However, the traditional discrete condition probability in market basket analysis can not treat the data of demand intentions reasonably, and hence make consequence inflexible and biased. Therefore, in the studies, the fuzzy set theory is adopted to solve the problem. We suggest some appropriate membership functions to describe the demand intentions of customers, and derive a fuzzy condition probability model from the viewpoint of market basket in data mining. A solution procedure is proposed to compute the degree of relations of various product combinations, represented by condition probabilities. Besides, we apply this method to investigate the optimal combinations of GPRS (General Packet Radio Service) mobile phone services. The results show that our proposed model is more accurate and flexible than discrete condition probability to provide efficient product combination.