Hani Rahmawati
Universitas Muhammadiyah Prof. DR. HAMKA

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

Found 2 Documents
Search

Perbandingan Algoritma NBC dan SVM dalam Analisis Sentimen Terhadap Dampak Kesehatan Rokok Elektrik Hani Rahmawati; Isa Faqihuddin Hanif
Progresif: Jurnal Ilmiah Komputer Vol 21, No 2 (2025): Agustus
Publisher : STMIK Banjarbaru

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35889/progresif.v21i2.2903

Abstract

There is no optimal method for accurately classifying public opinion, so an analytical approach is needed that is able to capture the nuances of public sentiment regarding the health impacts of e-cigarettes. This study examines public perception of the health impacts of electronic cigarettes using two classification algorithms: NBC and SVM. Data sourced from social media X (formerly Twitter) underwent stages of data cleaning, sentiment labeling, TF-IDF weighting, and data balancing through the SMOTE technique. Performance evaluation was conducted using four key metrics: accuracy, precision, recall, and f1-score. NBC achieved 80.5% accuracy with high recall despite low precision. In contrast, SVM recorded superior performance with 95.2% accuracy and more consistent balance between precision and recall. Therefore, the Support Vector Machine (SVM) algorithm is recommended as a more effective method for analyzing public sentiment regarding electronic cigarettes.Keywords: Electronic Cigarette; Entiment Analysis; Naïve Bayes Classifier; Support Vector Machine.AbstrakBelum adanya metode yang optimal untuk mengklasifikasikan opini publik secara akurat, sehingga diperlukan pendekatan analitik yang mampu menangkap nuansa sentimen masyarakat terhadap dampak kesehatan rokok elektrik. Studi ini mengkaji persepsi publik terhadap dampak kesehatan rokok elektrik dengan menerapkan dua algoritma klasifikasi: NBC dan SVM. Data yang bersumber dari media sosial X (eks Twitter) diproses melalui tahapan pembersihan data, pelabelan sentimen, pembobotan menggunakan TF-IDF, serta penyeimbangan data menggunakan teknik SMOTE. Evaluasi performa dilakukan menggunakan empat metrik utama: accuracy, precision, recall, dan f1-score. NBC memperoleh akurasi sebesar 80,5% dengan recall tinggi meskipun precision-nya rendah. Sebaliknya, SVM mencatat performa superior dengan akurasi 95,2% serta keseimbangan precision dan recall yang lebih konsisten. Oleh karena itu, algoritma Support Vector Machine (SVM) direkomendasikan sebagai metode yang lebih efektif dalam menganalisis sentimen publik terhadap rokok elektrik.Kata kunci: Analisis Sentimen; Rokok Elektrik; Naïve Bayes Classifier; Support Vector Machine.
Membangun Budaya Disiplin pada Penggunaan Teknologi Informatika dalam Masyarakat Muslim Hani Rahmawati; Salsa Billa Permana Putri; Azizah Salma Nida; Salsa Dwi Agistina; Nurjanah
Islamic Education Vol. 3 No. 3 (2024): July : journal Islamic Education
Publisher : CV. Maryam Sejahtera

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.68352/education.v3i3.1160

Abstract

The advancement of information technology has significantly impacted Muslim communities, presenting both opportunities and challenges in maintaining Islamic values. This research aims to explore strategies for building a culture of discipline in the use of information technology among Muslim communities. Using a literature review method, this study analyzes various sources related to Muslim perceptions and behaviors towards technology, challenges and opportunities faced, and effective strategies for building a culture of discipline. The results show that while technology is viewed positively for supporting religious activities, concerns about its negative impacts persist. Key challenges include digital literacy gaps, value conflicts, and potential technology addiction. Identified strategies include Islamic value-based digital literacy education, parental supervision, utilization of technology for da'wah, development of Islamic digital communities, and religious moderation campaigns. Implementing these strategies requires a comprehensive approach considering theological, ethical, social, and practical aspects. This research concludes that building a culture of discipline in technology use requires collaboration among various stakeholders and integration of Islamic values with modern digital literacy.