Article Per Year (5 Year)
p-Index From 2021 - 2026
0.408
P-Index
This Author published in this journals
All Journal Journal of Artificial Intelligence and Digital Business Help: Journal of Community Service (HJCS)
Farah Diba Azkia
Universitas Bina Sarana Informatika
Humanities Computer Science & IT Economics, Econometrics & Finance Electrical & Electronics Engineering Engineering Languange, Linguistic, Communication & Media Public Health Social Sciences Other

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

Found 2 Documents
Search

 1 
Empowering Community Digital Literacy through Participatory Artificial Intelligence Training Using Participatory Action Research in South Jakarta Sarifah Agustiani; Riska Aryanti; Tri Wahyuni; Elah Nurlelah; Pristya Haliza Ramadhanti; Farah Diba Azkia
Help: Journal of Community Service Vol. 3 No. 1 (2026): June 2026
Publisher : PT Agung Media Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.62569/hjcs.v3i1.279

Abstract

Digital literacy has become a fundamental competency for communities in responding to the rapid advancement of Artificial Intelligence (AI) technologies. However, many community members still have limited knowledge and practical skills in utilizing AI productively, ethically, and responsibly. This community service program aimed to empower community digital literacy through participatory Artificial Intelligence training using a Participatory Action Research (PAR) approach in RT 05, Cikoko Urban Village, South Jakarta. The program was implemented through four stages of PAR, including problem identification, collaborative planning, participatory action, and reflection. Training activities consisted of interactive lectures, live demonstrations, guided hands-on practice, group discussions, and mentoring on AI applications, digital ethics, information verification, and online security. Program evaluation was conducted using observations, reflective discussions, and post-training questionnaires involving sixteen participants. The findings revealed three major outcomes. First, the participatory learning approach successfully increased community engagement, with 75% female participants and 69% of participants aged 12–20 years actively involved throughout the learning process. Second, the training achieved high participant satisfaction, with information delivery, training materials, presenter performance, and event organization each receiving an 81% satisfaction score. Third, the program significantly improved community digital literacy and readiness for AI adoption. Participants reported that the program provided substantial benefits (88%), increased their knowledge (81%), improved practical AI utilization skills (81%), and enhanced their overall satisfaction (81%), while sustainable technology utilization, practical relevance, systematic implementation, and willingness to participate in future activities each achieved 75% positive responses. 
Analisis Sentimen Wattpad di Play Store Menggunakan Naïve Bayes Berbasis TF-IDF dan SMOTE Farah Diba Azkia; Agus Junaidi; Arif Ismail Husin
RIGGS: Journal of Artificial Intelligence and Digital Business Vol. 5 No. 2 (2026): Mei-Juli
Publisher : Prodi Bisnis Digital Universitas Pahlawan Tuanku Tambusai

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31004/riggs.v5i2.12268

Abstract

Kajian ini menguji performa klasifikasi Multinomial Naive Bayes yang diintegrasikan dengan skema pembobotan TF-IDF untuk memilah ulasan aplikasi Wattpad ke dalam kategori Positif, Negatif, dan Netral, sembari menakar kontribusi Synthetic Minority Oversampling Technique (SMOTE) dalam memitigasi ketimpangan jumlah data. Sebanyak 2.829 ulasan valid hasil web scraping melalui library google-play-srcaper diproses melalui tahapan krusial, meliputi pembersihan teks, case folding, konversi kata tidak baku, tokenisasi, eliminasi 773 stopword, hingga stemming dengan PySastrawi. Transformasi data menjadi vektor TF-IDF dengan batasan max_features=5.000 serta ngram_range=(1,2) menghasilkan matriks berdimensi 2.829 × 4.499. Melalui skema pembagian stratified split 80:20, teknik SMOTE diaplikasikan pada sektor data latih untuk mendongkrak jumlah sampel dari 2.263 menjadi 4.056 agar komposisi antar kelas lebih proporsional. Hasil observasi memperlihatkan pergeseran metrik performa yang cukup signifikan pasca-intervensi SMOTE. Sebelum dilakukan penyeimbangan data, model memang mencatat akurasi 68,37% dengan presisi 62,83% dan F1-score 60,24%, namun sistem menunjukkan kelemahan fatal yakni kegagalan total dalam mengidentifikasi kelas Netral. Setelah SMOTE diterapkan, kendati terjadi penurunan akurasi ke angka 58,83%, model justru menunjukkan ketangguhan lebih baik melalui peningkatan presisi menjadi 65,25% dan F1-score ke level 61,22%, serta lonjakan recall pada kelas Netral dari titik nol menjadi 0,43. Secara keseluruhan, pemetaan sentimen didominasi oleh opini Negatif sebesar 59,51%, disusul opini Positif 26,01%, dan Netral 14,48%, di mana mayoritas keluhan pengguna berhulu pada persoalan teknis aplikasi seperti lonjakan iklan, hambatan login, serta rendahnya stabilitas sistem.
Co-Authors Agus Junaidi Agustiani, Sarifah Arif Ismail Husin Elah Nurlelah Pristya Haliza Ramadhanti Riska Aryanti Tri Wahyuni