Article Per Year (5 Year)
p-Index From 2021 - 2026
0.444
P-Index
This Author published in this journals
All Journal Jurnal Pengabdian Sains dan Humaniora Journal of Applied Informatics Research
Dina Zatusiva Haq
Universitas Pembangunan Nasional “Veteran” Jawa Timur
Computer Science & IT Education Electrical & Electronics Engineering Engineering Other

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

Found 2 Documents
Search

 1 
Enhancing Clickbait Headline Identification Performance Without Preprocessing Through Feature Reduction and Sentiment Analysis Moch Deny Pratama; Anisa Nur Azizah; Misbachul Falach Asy'ari; Dimas Novian Aditia Syahputra; M Adamu Islam Mashuri; Binti Kholifah; Rifqi Abdillah; Adinda Putri Pratiwi; Dina Zatusiva Haq
Journal of Applied Informatics Research Vol. 1 No. 1 (2025): July
Publisher : Universitas Negeri Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26740/jair.v1i1.44659

Abstract

This study addresses the challenge of identifying clickbait headlines without relying on conventional text preprocessing, which can be resource-intensive and may degrade contextual integrity. To enhance detection performance, we examine three feature extraction methods: TF-IDF, Word2Vec, and Headline2Vec, an embedding technique designed for short texts like headlines. These features are optimized using feature selection algorithms, including Pearson Correlation Coefficient (PCC), Neighborhood Component Analysis (NCA), and Relief, to reduce dimensionality and enhance relevant signal retention. Sentiment polarity is also integrated as a complementary feature. A comparative evaluation is conducted using several machine learning classifiers, namely Support Vector Classifier (SVC), Random Forest, LightGBM, and XGBoost, across all combinations of feature extraction and selection methods. Results show that the optimal configuration Headline2Vec with Relief and SVC achieves the highest accuracy at 94.40%, outperforming other approaches. This demonstrates the effectiveness of combining semantic vectorization and feature selection for clickbait detection in the absence of traditional preprocessing. The findings support the development of streamlined and scalable classification models capable of maintaining high accuracy while reducing preprocessing overhead, making the proposed method particularly suitable for real-time and large-scale content moderation and news verification systems.
An Instant Online CV Creation Workshop Using Generative AI and a Web-Based Platform to Improve Digital Literacy and Job Readiness for Vocational High School Students Angga Lisdiyanto; Addien Haniefardy; Laqma Dica Fitrani; Agus Wibowo; Ikhwan Abdillah; Nurul Fuad; Winarti; Yerezqy Bagus; Dina Zatusiva Haq; Yoga Ari Tofan; Vinza Hedi Satria
Jurnal Pengabdian Sains dan Humaniora Vol. 5 No. 1 (2026): 2026 May Edition
Publisher : Fakultas Keguruan dan Ilmu Pendidikan-Universitas Timor

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32938/jpsh.v5i1.10943

Abstract

Graduates of Vocational High Schools (SMK) consistently contribute the most to Indonesia's Open Unemployment Rate (TPT), reaching 8.63% as of August 2025. A major issue is students' limited ability to develop relevant digital personal branding aligned with modern recruitment standards, such as having an attractive and accessible online Curriculum Vitae (CV). This community service project (PkM) aims to equip 12th-grade students at SMK Al-Amin Mojowuku, Kedamean, Gresik, with skills to create instant website-based CVs using three free tools: generative AI (DeepSeek), image hosting service (ImgBB), and HTML publishing platform (Tiiny.host). Conducted offline on April 28, 2026, with 28 participants, the workshop employed project-based learning combined with AI-assisted learning. The activity involved needs analysis, module development, workshops through lectures and practical exercises, and output evaluation. Results showed all participants successfully published personal CV websites with various themes such as manga comics, anime, and floral motifs. Quantitative indicators included a 100% task completion rate, high active engagement, and positive feedback on material relevance. This activity effectively improved digital literacy, creativity and prepared students for digital-focused recruitment processes
Co-Authors Addien Haniefardy Adinda Putri Pratiwi Agus Wibowo Angga Lisdiyanto Anisa Nur Azizah Binti Kholifah Dimas Novian Aditia Syahputra Ikhwan Abdillah Laqma Dica Fitrani M Adamu Islam Mashuri Misbachul Falach Asy'ari Moch Deny Pratama Nurul Fuad Rifqi Abdillah Vinza Hedi Satria Winarti Yerezqy Bagus Yoga Ari Tofan