Article Per Year (5 Year)
p-Index From 2021 - 2026
2.795
P-Index
This Author published in this journals
All Journal Jurnal Edukasi dan Penelitian Informatika (JEPIN) JIEET (Journal of Information Engineering and Educational Technology) Kinetik: Game Technology, Information System, Computer Network, Computing, Electronics, and Control Indonesian Journal of Electrical Engineering and Computer Science Journal of Informatics and Computer Science (JINACS) J-Icon : Jurnal Komputer dan Informatika Journal of Emerging Information Systems and Business Intelligence (JEISBI) Jurnal Ilmu Teknologi Informasi Indonesia (JITIFNA)
Wiyli Yustanti
Universitas Negeri Surabaya
Chemical Engineering, Chemistry & Bioengineering Civil Engineering, Building, Construction & Architecture Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Education Electrical & Electronics Engineering Energy Engineering Industrial & Manufacturing Engineering Languange, Linguistic, Communication & Media Library & Information Science

Published : 23 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 23 Documents
Search

 1   2   3 
A COMPARATIVE STUDY OF SUPERVISED FEATURE SELECTION METHODS FOR PREDICTING UANG KULIAH TUNGGAL (UKT) GROUPS Windy Chikita Cornia Putri; Wiyli Yustanti; Ervin Yohannes
J-Icon : Jurnal Komputer dan Informatika Vol 13 No 2 (2025): October 2025
Publisher : Universitas Nusa Cendana

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35508/jicon.v13i2.23893

Abstract

The manual classification of Uang Kuliah Tunggal (UKT) groups at Indonesian public universities is laborious, subjective, and error-prone, especially given the explosion of socio-economic data captured via online admission portals. In this study, we evaluate five feature selection techniques Chi-Square filter, Random Forest importance, Recursive Feature Elimination, LASSO embedded selection, and Exploratory Factor Analysis on a dataset of 9,369 applicants described by 53 socio-economic variables. Six classifiers (Decision Tree, Random Forest, SVM-RBF, K-Nearest Neighbor, and Naïve Bayes) were tuned via stratified 5-fold cross-validation within an 80:20 train-test split. Performance was measured by accuracy, macro-F1, and training time, and differences in weighted-average accuracy across feature-selection scenarios were assessed using the Friedman test (χ² = 15.06, p = 0.010). Results show that reducing to 13 features via LASSO (weighted-average accuracy 0.730) or Chi-Square (0.678) significantly outperforms both the full feature baseline (0.624) and the EFA baseline (0.303), while cutting computational costs by over 40%. We conclude that supervised feature selection particularly LASSO and Chi-Square enables simpler, faster, and more transparent UKT prediction without sacrificing accuracy. The novelty of this study lies in comparing five feature-selection methods within a standardized preprocessing pipeline on real UKT data from UNESA, resulting in a 13-feature subset aligned with the current UKT policy. This finding is ready to be integrated into an automated UKT verification system to enhance decision accuracy and efficiency.
Bitcoin Transaction Multivariate Forecasting Analysis Deep Learning Model Walk Forward Validation Muhammad Dafi Bagas; Wiyli Yustanti
Journal of Emerging Information Systems and Business Intelligence (JEISBI) Vol. 7 No. 3 (2026): Vol. 07 Issue 03
Publisher : Universitas Negeri Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26740/jeisbi.v7i3.77536

Abstract

The volatile and non-linear movement of Bitcoin prices makes price prediction a complex problem in time series analysis. This study aims to compare the performance of several deep learning models, namely Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Transformer, and Temporal Fusion Transformer (TFT), in predicting Bitcoin closing prices based on multivariate data. The dataset consists of daily historical data from 2020 to 2025, including Open, High, Low, Close, and Volume features. Model evaluation was conducted using the Walk Forward Validation (WFV) approach with 5 folds and was compared with the Cross Validation (CV) method. Three data split scenarios were applied: 70:30, 80:20, and 90:10. Model performance was measured using Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), Mean Absolute Percentage Error (MAPE), Symmetric MAPE (sMAPE), and the coefficient of determination (R²). Furthermore, the Wilcoxon Signed-Rank Test was employed to analyze the statistical significance of performance differences between validation methods. The results indicate that the GRU model under the 90:10 data split scenario achieved the best performance, with a median MAE of 0.0116 and RMSE of 0.0179, along with an R² value of 0.8622. This model demonstrated lower prediction errors and greater stability compared to the other models. Meanwhile, the Wilcoxon test results showed no significant difference between Walk Forward Validation and Cross Validation (p-value > 0.05), indicating that both validation methods produce statistically equivalent performance. Based on these findings, the GRU model is recommended as the most optimal model for Bitcoin price prediction under the experimental configuration used in this study.
Engagement Patterns of Educational Content on TikTok Amalia Putri; Wiyli Yustanti
Journal of Emerging Information Systems and Business Intelligence (JEISBI) Vol. 7 No. 3 (2026): Vol. 07 Issue 03
Publisher : Universitas Negeri Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26740/jeisbi.v7i3.78027

Abstract

The rapid development of digital technology has contributed to the increasing use of social media, particularly TikTok, which functions not only as an entertainment platform but also as a medium for distributing educational content through hashtags such as #edukasi. However, not all educational content generates the same level of engagement, making further analysis necessary to understand the interaction patterns formed within the platform. This study aims to analyze engagement patterns of educational content on TikTok and identify the dominant hashtags appearing alongside #edukasi within each cluster. The methods employed in this study include K-Means Clustering to group content based on engagement characteristics and Social Network Analysis (SNA) to examine relationships among hashtags. The findings indicate the formation of two clusters with different engagement characteristics, namely high-engagement and low-engagement clusters. Network analysis reveals that the low-engagement cluster forms several communities associated with topics such as facts, health, and children’s education, while the high-engagement cluster is dominated by hashtags related to educational toys, such as #mainananak and #mainanedukasi. These results demonstrate that the combination of clustering methods and social network analysis is effective in identifying engagement patterns and hashtag relationships in educational TikTok content.
Importance Performance Analysis (IPA) of Google Reviews Sentiments Based on SERVQUAL Dimension for Public Health Center Service in Surabaya Octania Sriwahyuni; Wiyli Yustanti
Journal of Emerging Information Systems and Business Intelligence (JEISBI) Vol. 7 No. 3 (2026): Vol. 07 Issue 03
Publisher : Universitas Negeri Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26740/jeisbi.v7i3.78201

Abstract

Google Reviews can serve as a digital mirror to gauge how the public evaluates the services of community health centers. This study focuses on analyzing the sentiment of Google Reviews for community health centers in Surabaya, mapping reviews into the SERVQUAL dimensions using Gap Analysis and Importance–Performance Analysis (IPA), and identifying the most influential keywords via TF-IDF within a GUI system. This study applies the Knowledge Discovery in Databases (KDD) workflow. Data was obtained by scraping reviews from 63 community health centers in Surabaya. Subsequently, sentiment was determined based on user ratings, then classified into the five SERVQUAL dimensions, and analyzed using the GAP analysis and Importance-Performance Analysis (IPA). The results indicate that positive public perceptions predominate. However, all dimensions still show negative scores, suggesting that service quality has not yet fully met user expectations. In the IPA analysis, Responsiveness, Assurance, and Empathy are categorized in Quadrant II as aspects that require maintenance, while Tangibles and Reliability fall into Quadrant III as low-priority aspects. Notably, no dimension is in Quadrant I. Additionally, TF-IDF successfully captures keywords such as “queue,” “long,” “friendly,” “clean,” and “procedure,” and has been successfully implemented in the GUI for automatic classification. Building on these results, this study confirms that digital reviews combined with sentiment analysis, SERVQUAL, and Importance-Performance Analysis (IPA) can serve as a more objective, practical, and sustainable evaluation tool for Public health center services.
Analisis Kepuasan Pengguna Aplikasi Neobank Melalui Analisis Sentimen Dan Klasifikasi Multilabel Dengan Logistic Regression Pada Kerangka Eucs Stevi Aprilianti Cahyani; Wiyli Yustanti
Jurnal Ilmu Teknologi Informasi Indonesia Vol. 2 No. 2 (2026): JITIFNA - Juli
Publisher : CV. SINAR HOWUHOWU

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.70134/jitifna.v2i2.1720

Abstract

The rapid development of digital banking services in Indonesia has led to an increased adoption of mobile banking applications as the primary medium for financial transactions. Neobank, as one of the mobile banking applications with a high number of downloads, exhibits a discrepancy between usage intensity and user satisfaction, as reflected in its ratings and user reviews. Various user reviews indicate issues related to technical constraints, system accuracy, ease of use, and service timeliness. This condition highlights the importance of conducting a comprehensive evaluation of user satisfaction. This study aims to analyze user satisfaction with the Neobank application through the integration of sentiment analysis and multilabel classification based on the End-User Computing Satisfaction (EUCS) framework. The dataset consists of 10,392 user reviews collected from the Google Play Store and App Store. The research stages include text preprocessing, sentiment labeling using a pre-trained Indonesian RoBERTa Base Sentiment Classifier, and EUCS dimension labeling using a fuzzy string matching approach based on keywords. Subsequently, a multilabel classification model was developed using Logistic Regression with a One-vs-Rest (OVR) approach and TF-IDF features.The evaluation results on the test data demonstrate a precision of 0.940, a recall of 0.809, and an F1-score of 0.869 (macro average). The k-fold cross-validation results (k=5) indicate stable model performance, with the highest average F1-score achieved in the Ease of Use dimension (0.927). The model was then implemented into a web-based system to perform sentiment prediction, EUCS multilabel classification, and generate improvement recommendations using a Large Language Model (LLM).
Pengembangan Standarisasi Kategori Keluhan Aplikasi Mypelindo Berbasis Latent Dirichlet Allocation Dan Klasifikasi Muwirotul Hasanah; Wiyli Yustanti
Jurnal Ilmu Teknologi Informasi Indonesia Vol. 2 No. 2 (2026): JITIFNA - Juli
Publisher : CV. SINAR HOWUHOWU

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.70134/jitifna.v2i2.1793

Abstract

The number of complaints received through the MyPelindo app has increased in line with its high usage for operational and administrative activities by employees. The large volume of complaint data received makes manual identification and grouping of complaints less effective and time-consuming. Therefore, the aim of this study is to extract and standardise complaint topics using the Latent Dirichlet Allocation (LDA) method, build a classification model based on the obtained topics, and implement this model in a website-based system to automatically predict the category of user complaints. The study was conducted using a text mining approach with a Knowledge Discovery in Databases (KDD) methodology, which includes the preprocessing stage, topic modelling using LDA and topic standardisation based on cosine similarity. The results of the standardisation were then used as category labels in the classification process. The best model was then implemented in a web-based system using the Flask framework. The results of the study showed that the optimal number of topics obtained was eight, with the highest coherence score of 0.353. The standardisation process successfully simplified the eight topics into three main categories: Ganti Perangkat & Error Umum, Masalah Login & Akun, and Absensi & Koreksi Data. Based on the classification evaluation results, the best model was Logistic Regression, achieving an accuracy rate of 96.05% and an F1-score of 0.96. The implementation of the web-based system also successfully predicted the category of complaints based on user input.
Web-based Profanity Detection Using a Combination of Lexicon and Support Vector Machine: Web-based Profanity Detection Using a Combination of Lexicon and Support Vector Machine Ainandita Riwipapusa; Wiyli Yustanti
Journal of Emerging Information Systems and Business Intelligence (JEISBI) Vol. 6 No. 4 (2025): Vol. 06 Issue 04
Publisher : Universitas Negeri Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26740/jeisbi.v6i4.71408

Abstract

Advances in information and communication technology, particularly the internet and social media, have made it easier for people to express their opinions openly, but have also increased the potential for the spread of profanity and hate speech. This study proposes a web-based profanity detection solution by combining lexicon-based methods and Support Vector Machine (SVM). The Knowledge Discovery in Database (KDD) process was implemented for data extraction and analysis, starting from Twitter data collection, preprocessing (cleaning, case folding, tokenizing, stemming), transformation using TF-IDF, to manual labeling. The SVM model was trained using a 3-fold cross-validation scheme, and evaluation was conducted using a classification report and confusion matrix. The results of the study showed a model accuracy of 93% on the test data with an average F1-score of 0.93, as well as optimal performance in detecting sentences categorized as profanity. The developed web application prototype successfully ran all profanity word detection and sensing features automatically, as proven by the black box testing results. The analysis test also ran smoothly, with a test using 10 sentences containing profanity words achieving 100% accuracy, and a test using 10 sentences without profanity words achieving 95% accuracy. This system is expected to contribute to creating a more positive digital space through adaptive and accurate profanity word detection.
Classification Agorithm Analysis For Predicting The Type Of Senior High School On Alumni Smp 2 Balong Ponorogo: Classification Agorithm Analysis For Predicting The Type Of Senior High School On Alumni Smp 2 Balong Ponorogo Nabiilah Winda Kurnia Putri; Wiyli Yustanti
Journal of Emerging Information Systems and Business Intelligence (JEISBI) Vol. 6 No. 4 (2025): Vol. 06 Issue 04
Publisher : Universitas Negeri Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26740/jeisbi.v6i4.71676

Abstract

This study aims to analyze the performance of various classification algorithms in predicting the type of Senior High School (SLTA) that students choose based on academic scores and achievements. The study was conducted at SMPN 2 Balong Ponorogo using the SEMMA (Sample, Explore, Modify, Model, Assess) approach. Secondary data from 1,113 students were used and processed through the stages of data exploration, normalization, feature selection (using Pearson Correlation, Mutual Information, Random Forest, and Lasso Logistic Regression), and dimension reduction using Principal Component Analysis (PCA). Eight classification algorithms were tested, namely Logistic Regression, K-Nearest Neighbors, Support Vector Machine, Random Forest, XGBoost, LightGBM, CatBoost, and Naïve Bayes. Model evaluation is done using accuracy, precision, recall, F1-score, and confusion matrix metrics. The results show that the Random Forest and KNN models with the Hybrid Feature Selection approach provide the best performance, with the F1-score value reaching 84%. This research contributes to data-based decision making for student guidance in choosing the right further education pathway.
Multivariate Time Series Forecasting on Sales Using Recurrent Neural Network (Case Study: Aqiqah Almeera) Susi Purwani; Wiyli Yustanti
Journal of Emerging Information Systems and Business Intelligence (JEISBI) Vol. 6 No. 4 (2025): Vol. 06 Issue 04
Publisher : Universitas Negeri Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26740/jeisbi.v6i4.72100

Abstract

Sales forecasting is a crucial component in business decision-making, particularly in inventory management and marketing strategies. Accurate sales predictions can help companies maintain stock balance, design effective promotions, and minimize the risk of losses. This study examines the application of Multivariate Time Series Forecasting using Recurrent Neural Networks (RNNs) to more accurately predict product sales. By considering multiple variables such as product price, inventory levels, promotional activities, and temporal features, this approach aims to capture complex and interrelated patterns in historical data. RNNs are chosen for their ability to handle sequential data and learn temporal relationships among variables, thereby improving prediction accuracy. This research adopts a quantitative method with a causal-associative approach, utilizing secondary data from the company’s sales records over the past two years. The data is analyzed using various preprocessing techniques such as data normalization, feature encoding, and correlation analysis for optimal feature selection before being fed into the RNN model. The model is trained using specific validation techniques to prevent overfitting. Model performance is evaluated using MAE and RMSE metrics to measure prediction accuracy and reliability. The results of this study are expected to produce an accurate and practical sales forecasting system that can be implemented by business practitioners to support more efficient, data-driven, and well-targeted decision-making processes.
Implementation of Xception Algorithm with Convolutional Block Attention Module (CBAM) for Waste Type Detection in Visual Images Ahmad Khoiru Shofa; Wiyli Yustanti
Journal of Emerging Information Systems and Business Intelligence (JEISBI) Vol. 7 No. 1 (2026): Vol. 07 Issue 01
Publisher : Universitas Negeri Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26740/jeisbi.v7i1.72502

Abstract

The increasing volume of waste each year poses a serious challenge in waste management, particularly in the waste sorting process, which remains suboptimal. The lack of public awareness and limited manual sorting facilities are major obstacles to creating an effective waste management system. To address this issue, this study developed a waste classification system based on visual images by utilizing the Xception algorithm integrated with the Convolutional Block Attention Module (CBAM) to improve classification accuracy. The dataset used in this study includes various categories of organic and anorganic waste. The experiments involved several stages, including the integration of CBAM into the Xception architecture, testing different data splitting schemes for training and validation, and hyperparameter tuning using the Random Search method with 10 combinations. The model was trained using the Keras and TensorFlow libraries, and the trained model was saved in the .h5 format commonly used for deploying deep learning models into web applications. The results showed that the addition of CBAM improved the model's accuracy from 88.38% to 91.29% without significantly increasing training time. Furthermore, the best hyperparameter combination obtained from tuning was Dense = 128, Dropout = 0.3, Optimizer = Adam, and Learning Rate = 0.0001. When retrained using this configuration, the model achieved a highest accuracy of 93.37%. The best-performing model was then integrated into a Flask-based web application. This application allows users to upload images of waste through a simple web interface and instantly receive the predicted waste type classification. With the implementation of this technology, the system is expected to assist the public in sorting waste more easily and to increase active participation in environmentally conscious waste management. Keyword: Waste Classification, Xception, CBAM, Deep Learning, Flask
Co-Authors Agus Prihanto Ahmad Khoiru Shofa Ainandita Riwipapusa Aliyah Alifi Amalia Putri Andhika Edo Prasetyo Anita Qoiriah Dhiany Hana Amalia Ervin Yohannes Hapsari Peni Agustin Tjahyaningtijas Irhamah - Miftakhul Alief Faturrahman Mohammad Hasan Machfoed Muhammad Dafi Bagas Muwirotul Hasanah Nabiilah Winda Kurnia Putri Naim Rahmawati Naim Rochmawati Nur Iriawan Octania Sriwahyuni Rahadian Bisma Salsabilla Putri Saharani Siti Mahmudah Putri Yanna Stevi Aprilianti Cahyani Susi Purwani Tama Priyadi Windy Chikita Cornia Putri Yoyok Prastyo YUNI YAMASARI