Article Per Year (5 Year)
p-Index From 2020 - 2025
0.61
P-Index
This Author published in this journals
All Journal Sistemasi: Jurnal Sistem Informasi Journal of Information Technology and Computer Science (JOINTECS) JITK (Jurnal Ilmu Pengetahuan dan Komputer) Jurnal Tekno Kompak
I Made Artha Agastya
Universitas AMIKOM Yogyakarta
Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Electrical & Electronics Engineering Engineering Languange, Linguistic, Communication & Media Library & Information Science Mathematics

Published : 4 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 4 Documents
Search

 1 
PENGARUH STEMMER BAHASA INDONESIA TERHADAP PEFORMA ANALISIS SENTIMEN TERJEMAHAN ULASAN FILM I Made Artha Agastya
Jurnal Tekno Kompak Vol 12, No 1 (2018): Februari
Publisher : Universitas Teknokrat Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33365/jtk.v12i1.70

Abstract

Bahasa Indonesia memiliki banyak variasi akhiran, awalan, dan sisipan. Stemming adalah bagian dari prapengolahan dari analisis sentimen yang mendeteksi dan menghilangkan imbuhan tersebut. Pengaruh dari stemming pada analisis sentimen masih belum jelas karena dataset yang digunakan tidak terdistribusi secara bebas. Untuk mendapatkan pengaruh dari stemming terhadap analisis sentimen maka dilakukan percobaan dengan dataset ulasan film yang sudah diterjemahkan ke Bahasa Indonesia. Stemmer Sastrawi sebagai algoritma stemming terbaru digunakan pada penelitian ini. Dataset dibagi menjadi 5 (lima) kategori yang mana 100 data, 250 data, 500 data, 750 data, dan 1000 data. Hasil yang diperoleh menunjukan bahwa stemmer tidak memberikan peningkatan akurasi yang stabil. Bahkan waktu yang diperlukan untuk menyelesaikan analisis sentimen memerlukan waktu meningkat hingga 310 kali lipat. Kenyataan ini sangat buruk karena stemming dapat mengurangi efisiensi dari analisis sentimen.
Implementasi Deep Learning untuk Klasifikasi Motor Imagery pada Sinyal EEG I Made Artha Agastya; Robert Marco2; Nila Feby Puspitasari
JOINTECS (Journal of Information Technology and Computer Science) Vol 8, No 2 (2024)
Publisher : Universitas Widyagama Malang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31328/jointecs.v8i2.6413

Abstract

Electroencephalography (EEG) adalah teknik yang digunakan untuk merekam aktivitas listrik otak melalui sensor yang ditempatkan pada kulit kepala. Salah satu area penelitian yang menarik dalam analisis EEG adalah motor imagery (MI), yaitu kemampuan untuk membayangkan suatu gerakan tanpa adanya stimulus visual eksternal. Pengolahan sinyal EEG yang kompleks dalam skenario MI memerlukan pendekatan komputasi yang canggih untuk mengenali pola-pola yang terbentuk selama proses pembayangan tersebut. Penelitian ini bertujuan untuk membandingkan kinerja empat arsitektur deep learning populer—yaitu EEGNet, EEGConformer, EEGInception, dan EEGITNet—dalam mengklasifikasikan data EEG pada konteks motor imagery. Hasil pengujian menunjukkan bahwa EEGConformer dan EEGNet adalah model yang paling efektif, dengan akurasi rata-rata masing-masing sebesar 72,41% dan 71,88%, serta performa yang stabil di berbagai subjek. Di sisi lain, EEGInception dan EEGITNet mencatatkan akurasi yang lebih rendah, terutama EEGInception dengan akurasi rata-rata sebesar 55,59%. Temuan ini mengindikasikan bahwa arsitektur sederhana seperti EEGNet tetap kompetitif, meskipun model yang lebih kompleks seperti EEGConformer memberikan sedikit keunggulan dalam performa. Penelitian ini juga menyoroti pentingnya faktor spesifik-subjek dalam meningkatkan performa model, yang dapat diatasi melalui pendekatan adaptif atau personalisasi. Hasil penelitian ini diharapkan dapat memberikan wawasan lebih mendalam terkait model yang paling akurat dalam tugas motor imagery EEG dan berkontribusi pada pengembangan aplikasi berbasis Brain-Computer Interface (BCI).
Sentiment Analysis of Banking Application Reviews on Google Play Store using Support Vector Machine Algorithm Martinus Juan Prasetyo; I Made Artha Agastya
Sistemasi: Jurnal Sistem Informasi Vol 13, No 6 (2024): Sistemasi: Jurnal Sistem Informasi
Publisher : Program Studi Sistem Informasi Fakultas Teknik dan Ilmu Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32520/stmsi.v13i6.4536

Abstract

Banking applications are increasingly important in facilitating daily financial transactions. However, to ensure service quality, developers need to understand user feedback. Reviews on the Google Play Store provide important insights related to satisfaction, complaints, and suggestions. Therefore, this study aims to develop the Sentiment Analysis Model for Banking Application Reviews Using Support Vector Machine (SVM). Data collected from three popular banks in Indonesia is used to train and test models. This research also contributes to providing multi -bank dataset which can be a benchmark. Various scenarios of the distribution of training and test data are explored, and repeated tests are carried out with different random state values to get stable results. The results showed that the SVM model was able to achieve good accuracy, with BRI Mobile dataset reaching the highest accuracy of 92.97%, followed by a combined dataset of 90.05%, BCA Mobile 89.73%, and Livin Mandiri 87.46%. Negative reviews are dominated by technical complaints, while positive reviews highlight the ease and reliability of the application. This study shows that the approach used has succeeded in producing competitive performance, and application developers are advised to focus on improving technical aspects, such as fixing login, verification, and transaction problems, in order to increase user satisfaction.
FOREST FIRE LOCATION AND TIME RECOGNITION IN SOCIAL MEDIA TEXT USING XLM-ROBERTA Hafidz Sanjaya; Kusrini Kusrini; Kumara Ari Yuana; Arief Setyanto; I Made Artha Agastya; Simone Martin Marotta; José Ramón Martínez Salio
JITK (Jurnal Ilmu Pengetahuan dan Teknologi Komputer) Vol. 10 No. 4 (2025): JITK Issue May 2025
Publisher : LPPM Nusa Mandiri

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33480/jitk.v10i4.6194

Abstract

Forest fires have become a serious global threat, significantly impacting ecosystems, communities, and economies. Although remote sensing technology shows potential, limitations such as time delays, limited sensor coverage, and low resolution reduce its effectiveness for real-time forest fire detection. Additionally, social media can serve as a multimodal sensor, presenting multilingual text data with rapid and global coverage. However, it may encounter challenges in obtaining location and time information on forest fires due to limitations in datasets and model generalization. This study aims to develop a multilingual named entity recognition (NER) model to identify location and time entities of forest fires in social media texts such as tweets. Utilizing a transfer learning approach with the XLM-RoBERTa architecture, fine-tuning was performed using the general-purpose Nergrit corpus dataset containing 19 entities, which were relabeled into 3 main entities to detect location, date, and time entities from tweets. This approach significantly improves the model's ability to generalize to disaster domains across multiple languages and noisy social media texts. With a fine-tuning accuracy of 98.58% and a maximum validation accuracy of 96.50%, the model offers a novel capability for disaster management agencies to detect forest fires in a scalable, globally inclusive manner, enhancing disaster response and mitigation efforts.
Co-Authors Arief Setyanto Hafidz Sanjaya, Hafidz José Ramón Martínez Salio Kumara Ari Yuana Kusrini Kusrini Martinus Juan Prasetyo Nila Feby Puspitasari Simone Martin Marotta