<![CDATA[The government's efforts to increase public interest in reading are by creating public libraries which are also implemented in Magelang City. However, interest in reading in Indonesia is still relatively low due to several factors. One of the shortcomings in the current library is the lack of interaction between library staff and library members to provide book recommendations that members might like. This study aims to design and build a book recommendation system based on the history of borrowing by members using the Apriori method at the Magelang City Library. With this system, it is hoped that library staff can provide book recommendations that might be interesting based on member interests. In this study, the apriori method was chosen because it was considered appropriate with the background of the method which works based on the frequency of items selected simultaneously. The apriori method which is often categorized as market basket analysis can be implemented for several purposes such as predicting items to be purchased in a store, arranging the layout of books or goods, so this study will try to implement the apriori method into a book recommendation system. The system is made by collecting data which is generally carried out from the observation process which is continued with system design and system implementation. The results of creating a book recommendation system using the apriori method built using a web programming language have a good level of accuracy if based on the calculation value. The only difference lies in the system output which will be automatically sorted based on the highest confidence value.]]>
