چكيده به لاتين
In today's digital world, businesses in the retail industry face new challenges every day, and with the rise of online stores, competition is getting tougher for them. Keeping in touch with the customer, knowing the characteristics of each customer, and making suitable offers are challenges that have occupied the minds of the activists of this industry. The answer to this problem lies in customer relationship management and customer segmentation. Each customer segment is a group of customers who have common needs and desires or, in a word, common behavior. In this research, data mining methods are used to segment customers and customer purchase data is examined; Then, the behavioral characteristics of customers are extracted from the RFMTP method, which is an improvement of the RFM method, and various Clustering methods are implemented on it. Next, among the models, the K-means model was chosen due to its better efficiency and accuracy, and the customers were divided into four segments: loyal, worthy, new customers, and churners based on the information of each cluster, and data analysis was provided according to each segment. Then, by using the combined model of decision tree and fuzzy system, the rules of each cluster were extracted so that it is possible to make decisions for future customers and determine the relevant section of each customer. Finally, with the FP-Growth method and the discovery of associative rules, the customers' shopping carts of each section were examined to be used for product suggestions.
