Klasifikasi Teks Berbasis Ontologi Untuk Dokumen Tugas Akhir Berbahasa Indonesia

Abstract

In this research, the classification of the final project document in Informatics Engineering UMM. The problem faced is the difficulty of finding relevant information and the difficulty of categorizing the TA documents according to areas of interest if it should be done manually. The purpose of this research is to get information based on abstract TA according to the category and make it easier to classify the TA document according to the field of interest. The categories used are the areas of interest in the study program: RPL, Computer Networking, Intelligent Game and Data Science. The data used TA documents as many as 500 data. The stage is to build and model ontology rules according to the data obtained with reference data of UMM Informatics Engineering Curriculum 2017 sourced from the Association for Computing Machinery (ACM) of the IEEE Computer Society. Ontology aims to classify the objects that exist in the collection without requiring the data train. To support the classification process used dao method. The dao method is used to calculate the similarity between documents from an existing node on the ontology by looking at the closest distance. Stage testing system using accuracy obtained result of 87%. This shows that ontologists are able to classify documents without using train data.