Product features characterization and customers' preferences prediction based on purchasing data

Big data of online product purchases is an emerging source for obtaining customers' preferences of product features for new product development. This paper proposes a framework and associated method for product features characterization and customers' preference prediction based on online product purchase data. Specifications and components of products are firstly analyzed and the relationships between product specifications and components are then established for features characterization. The customers preferred specifications, features and their combinations are predicted for development of new products. The features characterization and customers' preferences prediction of toy cars were used as an example of illustrating the proposed method.