Abstract
Sebelas Maret University has been publishing many academic articles. Classifying many articles at a time is not a simple task. The more articles need to be classified, the more energy and time needed. Naive Bayes Classifier method can be used to classify academic articles in short time. Naive Bayes Classifier classifies each article based on the field of study by analyzing its title and abstract. One of feature selection method Document Frequency Improved (DFM) is implemented for improving the classification performance. This study used of 292 articles as training data and 100 articles as testing data. It tested by applying 5 threshold value from 1 to 2,5 with each threshold executed 5 times. The best results showed at 2 threshold level with the average value of accuracy, precision, recall, and f-measure respectively are 87,8%, 76,6%, 76,2%, and 76,0%.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
