association rule Implementation of Data Mining to Determine Book Placement Recommendation Based on Lending Pattern Using Association Rule

Implementation of Data Mining to Determine Book Placement Recommendation Based on Lending Pattern Using Association Rule

Abstract

Library book lending data is increasing, thus a processing to make the lending transaction record data into information is required to help library visitor find books by finding relation with book borrowed at the same time. The relation of borrowed book item was found by analyzing library book lending data from 2014 to March 2019. The data was cleaned to select the attributes of id member, book code, book title and inconsistent writing, then the data was grouped into one single transaction during book lending in the library and transformed into tabular data to calculate the itemset of books borrowed at the same time. Association rules were made using data which had been grouped and transformed into one tabular data transaction during lending, resulting in 2225 transaction data with 0.01 support and confidence by putting the limit of 50 association rules with the highest role being lending communication science book with psychology book with support x confidence of 8.17%