Analysis of Online Consumer Behavior - Design of CRISP-DM Process Model

The basis of the modern marketing of a business entity is to know the behavior of its customers. Advanced artificial intelligence methods, such as data mining and machine learning methods, penetrate data analysis. The application of these methods is most appropriate in the case of online sales of any goods in large quantities and various industries. They are very often used in the sale of electronics, PCs or clothes. However, it is also possible to apply them to the agricultural industry, not only in B2C, but also in B2B in the sale of seeds, agricultural products, or agricultural machinery. Appropriate combinations of offers and knowledge of customers can bring the selling entity higher profits or competitive advantages. The main goal of our study is to design a CRISP-DM process model that will enable small businesses to analyze online customers' behavior. To reach the main goal we perform a data analysis of the online sales data by using machine learning methods as clustering, decision tree and association rules mining. After evaluating the proposed model, we discuss its use of the proposed model in the field of internet sales in the agricultural sector