Abstract :
Koperasi Wanita "Cempaka" is one type of active credit unions, by utilizing funds from members in the form of deposits and loans. Given the large number of candidates who enroll each year, the cooperative still less selective in the acceptance of candidates who only viewed from the aspect of work and wages, thus causing bad credit. To reduce the occurrence of bad credit, it is necessary to forecast repayments candidate member status, to determine prospective members including bad credit or good credit. This research applies data mining classification techniques using naïve Bayes method to determine the class of repayments which class or classes jammed smoothly. Attributes are used in this study consists of seven variables, namely age, status, number of children, employment, salaries, expenses, and the status of the house. System testing performed six times testing. The data used are taken from the data Loan Cooperative Members Woman "Cempaka" the year 2011-2016 as many as 610 data. Based on the test results showed that in the fourth test produces the highest accuracy reached 84.64%.