MANAJEMEN PENENTUAN JUDUL DAN DOSEN PEMBIMBING SKRIPSI DENGAN ALGORITMA OLIVER DAN ALGORITMA NAIVE BAYES CLASSIFIER

Abstract

A study program is always required to produce quality students, the quality of graduates is also seen from the research results or projects that have been produced. Buying titles and supervisors is currently still done manually by the head of the study program. So that sometimes there are redundancies in titles because they are only based on the memory of the head of the study program. Guidance is also done without any consideration. This study aims to assist the process of checking titles and ordering books to facilitate the managerial process of existing theses. The methods used for the process of checking the title and fishing are Oliver's algorithm and the Naive Bayes Classifier algorithm. As for the selection of supervising lecturers in the focus areas of lecturers (competence), functional positions and lecturer home base. With reference to the training data for supervisor 1 as many as 217 and training data for supervisor 2 as many as 177, while the test data is 10 data. The results of Oliver's algorithm test with a comparison of the proportion of the degree of similarity of titles accepted and rejected by 80%: 20% and the results of the Naive Bayes Classifier algorithm test with the level of adjustment of supervisor 1 of 80%: 20% and the level of conformity of supervisor 2 of 40%: 60 %.

Keywords

• Naive Bayes Classifier, Oliver's Algorithm, Superviser, Thesis Title