User preferences to attributes of books for personalized recommendation

Personalized recommendation and user preference to books have aroused lots of attention both domestically and abroad and fruitful achievements have been made in this area. Nevertheless, the previous preference models mainly focus on the content of books and ignore the attribute of price of a book. In an actual book recommendation system, the price influences the consumers' buying behavior to a large extent and should be considered seriously. Based on the existing preference models, in addition to the attributes of the publisher, publishing date, discount, and the other consumers' reviews, this study takes the attribute of price into consideration to a particular book. The dummy variable regression method is employed to analyze the data mainly from Amazon. Through data analysis and investigation, this paper tries to explore the importance of its different attributes in the user preference to a book. In accordance with the data analysis, some suggestions are provided for book recommendation system design and relevant website modification.