Garuda - Garba Rujukan Digital

SENTIMENT ANALYSIS USING MACHINE LEARNING FOR DIGITAL SERVICE DEVELOPMENT Rugaiyah Balqis; Jahda Rusti Putri; Mira Afrina; Ali Ibrahim; Fathoni Fathoni
JURTEKSI (jurnal Teknologi dan Sistem Informasi) Vol. 12 No. 2 (2026): Maret 2026
Publisher : Lembaga Penelitian dan Pengabdian Kepada Masyarakat (LPPM) STMIK Royal Kisaran

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33330/jurteksi.v12i2.4476

Abstract: The rapid growth of e-commerce mobile applications has generated large volumes of user reviews, making manual sentiment analysis increasingly impractical. This study aims to compare the effectiveness of three machine learning algorithms Support Vector Machine (SVM), Random Forest, and Naive Bayes for automated sentiment classification of Indonesian-language mobile application reviews. A dataset of 3,000 user reviews from the RupaRupa application on the Google Play Store was collected and preprocessed through normalization, tokenization, stopword removal, and stemming. TF-IDF vectorization was applied for feature extraction, while the Synthetic Minority Over-sampling Technique (SMOTE) was used to address class imbalance across three sentiment categories: positive, negative, and neutral. The results show that SVM achieved the highest accuracy of 90.02%, while Random Forest obtained the best F1-score of 88.08% when sufficient training data were available. Naive Bayes demonstrated relatively stable performance across varying training data sizes. Furthermore, TF-IDF keyword analysis revealed that negative reviews were primarily associated with delivery issues, technical problems, and pricing concerns. These findings demonstrate the effectiveness of machine learning approaches for sentiment classification and provide practical insights for improving mobile application services. Keywords: sentiment analysis; machine learning; SMOTE; TF-IDF; text classification Abstrak: Pertumbuhan pesat aplikasi mobile e-commerce telah menghasilkan volume ulasan pengguna yang sangat besar, sehingga analisis sentimen secara manual menjadi semakin tidak praktis. Penelitian ini bertujuan untuk membandingkan efektivitas tiga algoritma machine learning Support Vector Machine (SVM), Random Forest, dan Naive Bayes dalam melakukan klasifikasi sentimen otomatis terhadap ulasan aplikasi mobile berbahasa Indonesia. Dataset yang digunakan terdiri dari 3.000 ulasan pengguna aplikasi RupaRupa yang dikumpulkan dari Google Play Store. Data kemudian diproses melalui tahapan preprocessing yang meliputi normalisasi, tokenisasi, penghapusan stopword, dan stemming. Ekstraksi fitur dilakukan menggunakan metode Term Frequency–Inverse Document Frequency (TF-IDF), sedangkan ketidakseimbangan kelas ditangani menggunakan Synthetic Minority Over-sampling Technique (SMOTE) pada tiga kategori sentimen, yaitu positif, negatif, dan netral. Hasil penelitian menunjukkan bahwa SVM mencapai tingkat akurasi tertinggi sebesar 90,02%, sementara Random Forest memperoleh nilai F1-score terbaik sebesar 88,08% ketika tersedia data pelatihan yang memadai. Naive Bayes menunjukkan performa yang relatif stabil pada berbagai ukuran data pelatihan. Selain itu, analisis kata kunci berbasis TF-IDF mengungkapkan bahwa ulasan negatif terutama berkaitan dengan masalah pengiriman, kendala teknis aplikasi, dan isu harga. Temuan ini menunjukkan bahwa pendekatan machine learning efektif untuk klasifikasi sentimen serta memberikan wawasan yang bermanfaat dalam meningkatkan kualitas layanan aplikasi mobile. Kata Kunci: analisis sentimen; pembelajaran mesin; SMOTE; TF-IDF; klasifikasi teks.

Evaluating Pinterest User Experience and Usability Using AttrakDiff and PLS-SEM Septhia Charenda Putri; Ali Ibrahim; Yadi Utama; Endang Lestari Ruskan; Fathoni Fathoni
Journal of Information System and Informatics Vol 8 No 1 (2026): February
Publisher : Asosiasi Doktor Sistem Informasi Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.63158/journalisi.v8i1.1408

The rapid development of visual platforms such as Pinterest necessitates a comprehensive understanding of how functional and emotional aspects jointly influence users’ perception and engagement. This research addresses the gap in user experience (UX) evaluation of visually rich applications by examining the effects of Pragmatic Quality, Hedonic Quality-Stimulation, and Hedonic Quality Identity on the perceived Attractiveness of the Pinterest application. A quantitative approach was employed using the 28-item AttrakDiff instrument, based on data collected from a final sample of 524 valid respondents, predominantly aged 18–25 years, and using Pinterest several times a week. The data analysis was conducted using Partial Least Squares Structural Equation Modeling (PLS-SEM) with the aid of SmartPLS to examine the relationships among latent variables. The findings demonstrate that the structural model exhibits a high level of explanatory capability, with an R² value of 0.684. With all three UX dimensions exerting positive and statistically significant effects on Attractiveness. PQ shows the strongest influence (path coefficient = 0.457), followed by HQS (0.391) and HQI (0.112). These findings confirm that functional usability remains the primary driver of attractiveness on Pinterest, while hedonic qualities play a complementary role in enhancing user experience. Practically, this research suggests that designers and developers of visual platforms should prioritize efficient functionality while maintaining stimulating and identity-supporting elements to improve overall user appeal.

Analyzing the Impact of Review Sentiment on Carpentry Product Sales: Evidence from Tokopedia Agung Chandra Kharisma; Muhammad Haykal Alfariz Saputra; Ali Ibrahim; Mira Afrina
Journal of Information System and Informatics Vol 8 No 1 (2026): February
Publisher : Asosiasi Doktor Sistem Informasi Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.63158/journalisi.v8i1.1412

The rapid growth of e-commerce in Indonesia has increased the importance of consumer reviews as signals influencing purchasing decisions. This study examines the relationship between review sentiment and sales performance in the carpentry tools category on Tokopedia. Using a 2019 Kaggle dataset consisting of 1,826 reviews across approximately 60 products, we apply an NLP-based pipeline to classify review sentiment into positive, neutral, and negative categories. Sentiment labeling combines rating-based rules and a TF-IDF + Logistic Regression baseline, with additional evaluation using IndoBERT. Product-level metrics—including the proportion of positive sentiment (pos_share), average rating, and units_sold (sales proxy)—are analyzed using descriptive statistics, correlation analysis, and cross-sectional OLS regression. The findings reveal that, in this snapshot dataset, the association between positive sentiment share and log(units_sold + 1) is weak and statistically limited, suggesting that sales variation cannot be explained solely by sentiment polarity or average ratings without considering other commercial factors. These results highlight the importance of incorporating contextual variables and temporal design in future research. Practically, the study suggests that sellers should monitor not only sentiment polarity but also the informational richness of reviews to strengthen reputation management strategies.

Optimized Indonesia Language Preprocessing Framework for Gojek Reviews Sentiment Analysis Using Linear Support Vector Machine Jahda Rusti Putri; Hendra Wijaya; Ali Ibrahim; Ahmad Heryanto
Indonesian Journal of Artificial Intelligence and Data Mining Vol. 9 No. 2 (2026): July 2026
Publisher : Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar

This study analyzes sentiment in Gojek application user reviews using Natural Language Processing (NLP) and machine learning techniques to classify sentiments into positive, negative, and neutral categories. A dataset of 8,091 Bahasa Indonesia reviews from Kaggle (Gojek version 4.8) was processed, yielding 5,685 valid instances after cleaning, with sentiment distribution of neutral (52.7%), positive (26.8%), and negative (20.3%). An optimized Indonesian preprocessing pipeline was developed, incorporating text cleaning, slang normalization using a curated 1,247-word mapping dictionary, tokenization, stemming via Sastrawi, and stopword removal. Feature extraction employed TF-IDF Vectorizer (max_features=10,000; n-gram=(1,2)). Four algorithms, Naïve Bayes, Linear SVM, Logistic Regression, and Random Forest (tuned) were evaluated using stratified 80:20 split. Linear SVM and Random Forest achieved the highest accuracy at 93% (weighted F1-score: 93%), followed by Logistic Regression (92%) and Naïve Bayes (68%). Ablation study confirmed that slang normalization contributed the greatest performance gain (4.8%). Keyword-based aspect extraction on negative reviews identified three priority improvement areas: customer service responsiveness (38%), pricing transparency (32%), and application stability (24%).

IMPACT OF TEXT AUGMENTATION ON INDOBERT PERFORMANCE FOR HOSPITAL REVIEW SENTIMENT ANALYSIS Yoga Anugrah Pratama.SY; Ali Ibrahim
JITK (Jurnal Ilmu Pengetahuan dan Teknologi Komputer) Vol. 12 No. 1 (2026): JITK Issue August 2026
Publisher : LPPM Nusa Mandiri

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33480/jitk.v12i1.8327

Sentiment analysis of hospital patient reviews plays a critical role in evaluating healthcare service quality. However, limited labeled data and class imbalance often affect classification performance and reduce minority-class detection. This study empirically investigates the impact of text augmentation techniques on improving IndoBERT performance for sentiment classification of patient reviews at Dr. Mohammad Hoesin Palembang General Hospital. A total of 1,464 reviews were collected, preprocessed, and weakly labeled using a VADER-based approach, resulting in 1,168 positive and 296 negative instances. To address class imbalance, augmentation was applied exclusively to the training set using back-translation and Easy Data Augmentation (EDA), including synonym replacement, random insertion, random swap, and random deletion. IndoBERT was fine-tuned under consistent hyperparameter settings and evaluated using accuracy, precision, recall, F1-macro, and AUC. The baseline model achieved an F1-macro of 70.2%, indicating limited minority-class sensitivity. After augmentation, the random swap technique achieved the highest observed performance within the experimental setup, reaching 96.8% accuracy, 95.3% F1-macro, and 98.9% AUC. These results suggest improved performance within the experimental setting, particularly in minority-class detection. However, it should be noted that the labels were generated through a translation-based weak labeling approach, which may introduce noise and affect the accuracy of the labels. Therefore, the findings should be interpreted within the scope of this experimental setting.

User-Centered Design of Marker-Based Augmented Reality Learning Media to Improve Informatics Learning Effectiveness Chairunnisa Umi Kalsum; Ali Ibrahim; Ken Ditha Tania
Journal of Information System and Informatics Vol 8 No 4 (2026): August
Publisher : Asosiasi Doktor Sistem Informasi Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.63158/journalisi.v8i4.1704

Computer science education in schools still relies heavily on traditional teaching methods, resulting in low student motivation, particularly in topics related to computer hardware components. This study develops an interactive Augmented Reality (AR) learning medium with Marker-Based Tracking using a User-Centered Design (UCD) approach to improve Informatics Learning effectiveness among 73 junior high school students at SMP Negeri 28 OKU. The UCD method was implemented through four stages: establishing requirements, designing, prototyping, and evaluating. Learning effectiveness was measured using N-Gain, a paired sample t-test, and Cohen's d based on pre-test and post-test scores. User experience was evaluated using the User Experience Questionnaire (UEQ), covering six dimensions: Attractiveness, Perspicuity, Efficiency, Dependability, Stimulation, and Novelty. The N-Gain score reached 0.937, classified as "High." The paired t-test yielded a statistically significant difference between pre-test and post-test scores (p < 0.001), and Cohen's d was 1.757, indicating a "Very Large" effect size. UEQ evaluation results across all six dimensions exceeded the benchmark threshold for the "Excellent" category. These findings indicate improved learning outcomes in the evaluated class and positive user experience with the developed AR medium.

Evaluation of Information Systems on the SIMDAPRO using the Unfield Theory of Acceptance and Use of Technology (UTAUT) Method Muhammad Jidan Hasbiallah; Ali Ibrahim; Dwi Rosa Indah; Iin Seprina; Ricy Firnando
Journal of Information System and Informatics Vol 6 No 2 (2024): June
Publisher : Asosiasi Doktor Sistem Informasi Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51519/journalisi.v6i2.763

The Management Information System for Housing Profile Data and Settlement Areas (SIMDAPRO) is a web-based system managed by Department Housing settlement Areas South Sumatra Province (DISPERKIM).. This system is integrated and unified, thus accelerating and improving the process of proposing assistance from the South Sumatra Provincial Government. Additionally, it facilitates related parties in verifying proposals. To analyse the factors influencing to understand how people accept and use the information technology, This study employs the UTAUT model, comprising four primary constructs: the first one is performance expectancy, and the next is the second effort expectancy, and the next one is the third social influence, and the next one is the fourth facilitating conditions. It aims to examine how these constructs influence the behavioral intention of (SIMDAPRO) application users in South Sumatra Province. The research approach is quantitative accompanied by a survey method. The research sample consists of 34 respondents who were chosen through using purposive sampling. Data collection techniques include validation, questionnaires, and observation. with instrument tests conducted for validity and reliability. The findings reveal that all four constructs the first one is performance expectancy, and the next is the second effort expectancy, and the next one is the third social influence, and the next one is the fourth facilitating conditions. significantly and positively impact the behavioral intention of SIMDAPRO application users in South Sumatra Province.

Evaluating User Experience and Usability of the USEPT Website Using User Experience Questionnaire and System Usability Scale Method Fanny Olivia; Ali Ibrahim
Journal of Information System and Informatics Vol 6 No 4 (2024): December
Publisher : Asosiasi Doktor Sistem Informasi Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51519/journalisi.v6i4.910

In recent years, assessing the quality of the Sriwijaya University English Proficiency Test (USEPT) website and its user experience has become increasingly important due to the globalization of higher education and the need for students to develop international communication skills. This study aims to evaluate the UX and usability of the USEPT website by using the User Experience Questionnaire (UEQ) and System Usability Scale (SUS) to provide a comprehensive analysis. The results showed that all aspects of the UEQ scored with an average value ranging from -0.88 to 0.48 making it in the “bad” category especially on attractiveness, stimulation and novelty. While the average SUS score is 50 which categorizes usability at the “Not Acceptable” level, with a grade of “F” and adjective rating at the “poor” level. These findings illustrate the poor functionality of the website due to unsatisfactory user experience, thus requiring holistic improvements across all dimensions of UEQ. Recommendations for improvement include optimizing website navigation, increasing visual appeal by creating an attractive design, integrating interactive features to increase user engagement and satisfaction. This research makes a positive contribution to the development of the USEPT website in user satisfaction and can be a reference for website evaluation.

Harnessing SVM for Sentiment Analysis: Insights from Gojek's Instagram Engagement Muhammad Juan Savero; Ali Ibrahim; Yadi Utama; Endang Lestari
Journal of Information System and Informatics Vol 7 No 1 (2025): March
Publisher : Asosiasi Doktor Sistem Informasi Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51519/journalisi.v7i1.1041

The development of digital technology has changed the transportation industry, including online services such as Gojek. Understanding customer sentiment is key in improving user experience and designing more effective business strategies. This research analyzes Gojek user sentiment on Instagram using Support Vector Machine (SVM). Data is obtained through web scraping, then processed through text cleaning, tokenization, common word removal, and stemming. Features were extracted using Term Frequency-Inverse Document Frequency (TF-IDF) before being classified with SVM. The results showed that the SVM model achieved 70.82% accuracy in classifying user sentiment. Most positive comments highlight the convenience and efficiency of the service, while negative comments are more related to high tariffs, application constraints, and less responsive customer service. These findings provide insights for Gojek to improve marketing strategies, optimize customer service, and adjust fare policies based on user feedback. In addition, this analysis can help in predicting real-time customer satisfaction trends through sentiment monitoring on social media. As a development step, this research recommends further exploration with deep learning and Aspect-Based Sentiment Analysis (ABSA) to improve accuracy and understand the service aspects that have the most influence on customer satisfaction.

Taxpayer Classification Using K-Means Clustering to Support CRM Strategy Development: Case Study of Prabumulih City Samsat Bimmo Fathin Tammam; Ali Ibrahim; Dwi Rosa Indah; Ahmad Fali Oklilas; Yadi Utama
Journal of Information System and Informatics Vol 7 No 4 (2025): December
Publisher : Asosiasi Doktor Sistem Informasi Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.63158/journalisi.v7i4.1365

Effective management of taxpayer data is crucial for enhancing compliance and optimizing regional revenue. This study addresses the limited use of data-driven taxpayer segmentation in local Samsat institutions by applying K-Means Clustering to support targeted Customer Relationship Management (CRM) strategies. A dataset of 3,999 motor vehicle taxpayer records from September 2025 was processed through feature selection, scaling, and clustering. The analysis identified three distinct taxpayer groups based on payment timeliness, compliance consistency, and vehicle age. Cluster validity was confirmed using the Davies-Bouldin Index, yielding a value of -41.327 for k = 3, supported by ANOVA for statistical significance. The findings highlight how clustering can reveal taxpayer behavior patterns, guiding personalized services and compliance programs. This study's novelty lies in integrating clustering outcomes with practical CRM strategies for public agencies, offering a data-driven approach to improve taxpayer engagement and regional revenue. However, the study is limited by its focus on a single-period dataset and vehicle-related attributes.

Title

Found 10 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 10 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 10 Documents
Search