Subhan, Muh
Unknown Affiliation

Published : 2 Documents
Articles

Found 2 Documents
Search

Classification of Radical Web Content in Indonesia using Web Content Mining and k-Nearest Neighbor Algorithm Subhan, Muh; Sudarsono, Amang; Barakbah, Ali Ridho
EMITTER International Journal of Engineering Technology Vol 5 No 2 (2017)
Publisher : Politeknik Elektronika Negeri Surabaya (PENS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24003/emitter.v5i2.214

Abstract

Radical content in procedural meaning is content which have provoke the violence, spread the hatred and anti nationalism. Radical definition for each country is different, especially in Indonesia. Radical content is more identical with provocation issue, ethnic and religious hatred that is called SARA in Indonesian languange. SARA content is very difficult to detect due to the large number, unstructure system and many noise can be caused multiple interpretations. This problem can threat the unity and harmony of the religion. According to this condition, it is required a system that can distinguish the radical content or not. In this system, we propose text mining approach using DF threshold and Human Brain as the feature extraction. The system is divided into several steps, those are collecting data which is including at preprocessing part, text mining, selection features, classification for grouping the data with class label, simillarity calculation of data training, and visualization to the radical content or non radical content. The experimental result show that using combination from 10-cross validation and k-Nearest Neighbor (kNN) as the classification methods achieve 66.37% accuracy performance with 7 k value of kNN method[1].
Klasifikasi Mutu Buah Pala (Myristica Fragrans Houtt) Berbasis Pengolahan Citra Menggunakan Metode Deep Learning Arsitektur Faster R-CNN Subhan, Muh; Basri, Hasan
INTEK: Jurnal Penelitian Vol 6, No 2 (2019): Oktober 2019
Publisher : Politeknik Negeri Ujung Pandang

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1742.361 KB) | DOI: 10.31963/intek.v6i2.1566

Abstract

Fakfak is the number one nutmeg producing area in Indonesia with a land area of around 16,011 Ha. Nutmeg production is projected to continue to increase until 2020, recorded in 2011 the production of nutmeg in particular reached 12,884 tons or 25 percent of the total Indonesian nutmeg production. This increase was not followed by an increase in market share. One solution to get a wider nutmeg market is exports. But currently not all requests for nutmeg can be fulfilled, because the quality of the nutmeg does not meet the requirements requested. One factor is the surface defects in the nutmeg skin that affects the quality of the fruit, especially the appearance of the fruit. The sorting of nutmegs has so far been using conservative methods, namely by observation based on experience (self-taught). This manual method is felt to be less effective because it depends on the conditions and conditions of the sorting staff, different perceptions between each sorter, takes a long time, requires large costs and involves many workers. To deal with these problems, our previous research developed a method for classifying nutmeg seeds, using image processing methods with color and shape parameters combined with a neural network and sigmoid convolution classification algorithm as a validation method, an accuracy of 87%, but this has not been said to be optimal so we try to use the same approach with the latest method improvements using the R-CNN Faster obtained the best accuracy of 95% with a learning rate of 4000 with a processing time of 0.04 minutes per second.