Classification of Science, Technology and Medicine (STM) Domains with PSO and NBC
| Penulis |
|---|
Dosen:
Eksternal:
|
| Tanggal Terbit |
| 07 Agustus 2018 |
| Kategori |
| Jurnal Internasional [Lainnya] |
| Penerbit |
| IEEE Xplore |
| Kota / Negara |
| Medan / Indonesi |
| Halaman |
| - |
| URL |
| https://ieeexplore.ieee.org/document/8674271 |
| Abstrak |
| Science, Technology, and Medicine (STM) is a field of research that has a characteristic in each document. These characteristics are different from most documents that are used as a corpus in mining text research. Documents derived from Newswire are more dominant in previous research. However, in this study will try to classify documents from STM field. Complex technical terms, symbols, position information, and the number of citations would be a challenge itself. Previous studies have used the Naive Bayes Classifier (NBC) classification method. There are also those who apply Particle Swarm Optimization to assist its classification. From the Newswire field generated a fairly high accuracy Therefore, it would be applied to the optimization method with PSO and combine it with NBC method. This study produced accuracy value in classification model without using PSO equal to 82,73%. While in the classification model using PSO, the accuracy value is 87.27%. This shows that the use of PSO optimization is very influential on the classification. |
