Date of this Version
The library is one of the most important facilities because it manages collections of written works, printed works, and recorded works and can provide information resources as well as be a driving force for the advancement of an educational institution. Conventional libraries will have piles of book borrowing transaction data recorded in the agenda book, which is only an archive, and the placement of books far apart, which causes members to take longer to find books when borrowing books of different types, is an issue that must be addressed. To overcome these two issues, a recommendation for an intelligent system is required. This study was carried out in one of Indonesia's vocational schools, with data collected through observation and interviews with librarians. The goal of this study is to examine the borrowing pattern of books using association data mining techniques. The association method used is a priori, and it will result in recommendations for association rules. The result of the association rule with reference to the 2-itemset with the highest value is a combination of Religion book and Physical Education book with 8 percent support and 100 percent confidence, whereas the association rule with 3-itemset reference resulted in 4 rules with 6 percent support and 100 percent confidence. The result is an application that can generate association rules for book recommendations and book placement recommendations.