Article Per Year (5 Year)
p-Index From 2021 - 2026
0.444
P-Index
This Author published in this journals
All Journal Jurnal Surya Informatika ISLAMIC PEDAGOGY: Journal of Islamic Education
Safitri, Fadilah
Unknown Affiliation
Computer Science & IT Education

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

 1 
Penerapan Model Pembelajaran Make a Match dalam Meningkatkan Hasil Belajar Aqidah Akhlak Siswa Madrasah Tsanawiyah Negeri 5 Kota Jambi Safitri, Fadilah; Surayya, Ely; Wiratama, Arif; Habibi, M. Fadli
ISLAMIC PEDAGOGY: Journal of Islamic Education Vol. 1 No. 2 (2023)
Publisher : Sekolah Tinggi Ilmu Syariah Ummul Ayman Pidie Jaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52029/ipjie.v1i2.163

Abstract

Class VIII-C student learning outcomes experience classical mastery. This is motivated by how active students are and how extensive understanding of the material impacts learning outcomes. For this reason, the purpose of this study was to find out whether the application of the Make a Match model to learning aqeedah morals can improve student learning outcomes in class VIII-C Madrasah Tsanawiyah Negeri 5 Jambi City. This type of research is classroom action research. The class that was used as the class under study was class VIII-C as a whole the students in that class. Data collection techniques using multiple choice questions, interviews, observation and documentation. Based on the results of applying the Make a Match learning model, it is categorized as very good. This can be seen from the scores of student learning outcomes after the learning process in each cycle has increased. Based on the learning outcomes scores obtained by students from cycle I, the average value was 73,3 or 17 students had completed learning, cycle II had an average score of 86,3 or 27 students had completed learning. So it can be concluded that the average increase in learning outcomes after the application of the Make a Match learning model that occurs from cycle I to cycle III is 79.8, which means that through the application of the Make a Match learning model it is able to improve learning outcomes to improve student learning outcomes.
Perbandingan Performa Algoritma Deep Learning dan Machine Learning untuk Klasifikasi Text Aduan di Sosial Media Nasim, Muhamad; Safitri, Fadilah; Ibrahim, Enggal Meureksa; Pratama, Arya Surya
Jurnal Surya Informatika Vol. 16 No. 2 (2026): - UPCOMING ISSUE - Jurnal Surya Informatika, Vol 16. No. 2, November 2026
Publisher : Universitas Muhammadiyah Pekajangan Pekalongan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.48144/suryainformatika.v16i2.2509

Abstract

Perkembangan media sosial menghasilkan volume data aduan publik berbahasa Indonesia yang besar dan beragam sehingga diperlukan metode klasifikasi otomatis yang efektif dan efisien. Penelitian ini bertujuan membandingkan performa algoritma machine learning, yaitu Naive Bayes dan Support Vector Machine (SVM), dengan algoritma deep learning, yaitu Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), dan Generative Adversarial Network (GAN), dalam melakukan klasifikasi teks aduan publik. Kebaruan penelitian ini terletak pada evaluasi penggunaan GAN sebagai model klasifikasi melalui komponen Discriminator pada konteks aduan publik berbahasa Indonesia. Dataset yang digunakan terdiri atas 5.000 data teks yang diperoleh dari media sosial, dengan distribusi 2.127 data Complaint (42,5%) dan 2.873 data Not Complaint (57,5%). Tahapan penelitian meliputi preprocessing teks, pembagian data menggunakan stratified sampling dengan proporsi 80% data latih dan 20% data uji, serta evaluasi model menggunakan Accuracy, Precision, Recall, dan F1-Score berdasarkan rata-rata tiga kali pengujian. Hasil penelitian menunjukkan bahwa Naive Bayes menghasilkan performa terbaik dengan Accuracy sebesar 77,50%, Precision 77,41%, Recall 77,50%, dan F1-Score 77,31%. Model berikutnya adalah CNN dengan F1-Score 77,09%, SVM sebesar 74,67%, LSTM sebesar 73,66%, dan GAN sebesar 41,98%. Rendahnya performa GAN dipengaruhi oleh ketidakstabilan proses pelatihan adversarial, sedangkan LSTM mengalami kecenderungan overfitting pada dataset berukuran sedang. Penelitian ini menunjukkan bahwa Naive Bayes lebih efektif untuk klasifikasi aduan publik dengan jumlah data terbatas karena memiliki akurasi kompetitif dan kompleksitas komputasi yang lebih rendah. Penelitian selanjutnya dapat mengembangkan dataset yang lebih besar, memperluas kategori aduan, menerapkan teknik augmentasi data, serta mengeksplorasi model berbasis transformer seperti BERT dan IndoBERT
Co-Authors Ely Surayya Habibi, M. Fadli Ibrahim, Enggal Meureksa Nasim, Muhamad Pratama, Arya Surya Wiratama, Arif