News Category Classification using Support Vector Machine Algorithm
| Main Authors: | Juliana, Nisa Eka, Khansa, Faridah Dewi, Azis, Aaz M Hafidz, Gunawan, Rafli Indra, Cahya, Nurul Dwi |
|---|---|
| Format: | Article info application/pdf Proceeding |
| Bahasa: | eng |
| Terbitan: |
UIN Sunan Gunung Djati Bandung
, 2021
|
| Subjects: | |
| Online Access: |
https://conferences.uinsgd.ac.id/index.php/gdcs/article/view/98 https://conferences.uinsgd.ac.id/index.php/gdcs/article/view/98/60 |
Daftar Isi:
- Nowadays many have used web-based systems to convey information and news in real time. However, in dividing news into these categories, some are still done manually, so it takes a long time. Of the several existing techniques, the technique most often used for classification of news content is the Support Vector Machine (SVM). In complex problems or problems with many parameters, this method is very good to use. The SVM algorithm performs supervised learning classifications or has inputs and outputs that have been formed into a mathematical relationship model that can classify and predict existing data. There are 2224 datasets and 5 categories with 70% of the data being trained and 30% of the data being tested. This study produces text classifications in the form of technology, business, sports, entertainment, and political categories from digital news content. The classification results obtained an accuracy value of 98.35% with an average precision of 90%, a recall of 98%, an F1-score of 98% and a Support of 668.