A Implementation of the Support Vector Machine Algorithm in the Student Thesis Subject Classification System at the University of Lampung

Hadi, Rizky

Journal of Applied Science, Engineering and Technology (J. ASET)
Vol. 2 No. 2 (2022): December 2022

The Technical Implementation Unit (UPT) of the University of Lampung Library is responsible for archiving student scientific work (thesis). Filing is done manually, using the DDC or Dewey Decimal Classification system, thus allowing for the tendency of inaccurate Subject selection and long grouping durations. This study aims to apply the Support Vector Machine (SVM) algorithm in classifying thesis subjects stored in the Unila student scientific article repository. The application of the SVM algorithm uses the machine learning life cycle method which includes the data collection process, data pre-processing, data splitting, model training, to the model evaluation process. The data used are Unila student thesis titles totaling 1707 data. The results of this study are a thesis subject classification model based on the title with the SVM model where the accuracy of the training data in the model training process is 0.95, and the evaluation model process with an accuracy rate of 0.65.