cover
Contact Name
Husni Teja Sukmana
Contact Email
husni@bright-journal.org
Phone
+62895422720524
Journal Mail Official
jads@bright-journal.org
Editorial Address
Gedung FST UIN Jakarta, Jl. Lkr. Kampus UIN, Cemp. Putih, Kec. Ciputat Tim., Kota Tangerang Selatan, Banten 15412
Location
Kota adm. jakarta pusat,
Dki jakarta
INDONESIA
Journal of Applied Data Sciences
Published by Bright Publisher
ISSN : -     EISSN : 27236471     DOI : doi.org/10.47738/jads
One of the current hot topics in science is data: how can datasets be used in scientific and scholarly research in a more reliable, citable and accountable way? Data is of paramount importance to scientific progress, yet most research data remains private. Enhancing the transparency of the processes applied to collect, treat and analyze data will help to render scientific research results reproducible and thus more accountable. The datasets itself should also be accessible to other researchers, so that research publications, dataset descriptions, and the actual datasets can be linked. The journal Data provides a forum to publish methodical papers on processes applied to data collection, treatment and analysis, as well as for data descriptors publishing descriptions of a linked dataset.
Articles 1,270 Documents
Hybrid MINLP-FNS Framework for Solving Large-Scale LIRP in E-Retail Logistics Kristian Telaumbanua; Syahril Efendi; Poltak Sihombing; Maya Silvi Lydia
Journal of Applied Data Sciences Vol 7, No 3: September 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i3.1114

Abstract

This research proposes a hybrid optimization framework to address the large-scale, multi-echelon Location-Inventory-Routing Problem (LIRP) in e-retail logistics. The proposed method combines a Mixed-Integer Nonlinear Programming (MINLP) model with a tailored Feasible Neighborhood Search (FNS) algorithm to solve complex decision-making problems involving facility location, inventory control, and vehicle routing simultaneously. Distinct from conventional models, the framework is integrated with a Business Intelligence (BI) environment to enable real-time decision support and dynamic data processing. Experimental evaluations were conducted using realistic logistics scenarios involving four distribution echelons. The results show that the hybrid MINLP-FNS approach achieves a total logistics cost reduction of 13.6% and a runtime improvement of 48% compared to the baseline MINLP-only model. It also significantly outperforms traditional GRG-based methods in scalability and computational stability across large datasets. These findings demonstrate that the proposed hybrid framework offers a more effective and scalable solution for complex logistics optimization, while its BI integration ensures practical applicability in real-world operations. This study contributes a novel data-driven framework that advances current research in intelligent supply chain and optimization systems.
Study on Image Classification Method Based on Small Sample Learning Dongxue Wang
Journal of Applied Data Sciences Vol 3, No 2: MAY 2022
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v3i2.57

Abstract

Image classification as according to their different features of reflected in the image information, make a distinction between different categories of target image processing methods, and especially for quantitative analysis using the computer, each of the images or image pixels, or regional planning is one of the several categories, in lieu of visual interpretation of the person, It has important practical value for the study of image classification method. However, the current study of image classification method based on small sample learning cannot effectively follow the development needs of society and industry, so it is urgent to carry out effective reform. Based on this, this paper first analyzes the problems existing in the research system construction of image classification method in small sample learning, and then gives the construction strategy of image classification method system according to these problems.
Applied Regression Modelling to Recommend Green Business Development in Vietnam Lu Phi Nga; Nguyen Quoc Huy; Phan Thanh Tam
Journal of Applied Data Sciences Vol 5, No 3: SEPTEMBER 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i3.294

Abstract

The green businesses engaged in environmentally friendly business activities, minimizing negative impacted on the ecosystem and used resources effectively. Green businesses contributed to the country's sustainable development goals and brought practical benefits to the businesses themselves. Therefore, the article aimed to identify the key factors affecting green business development and proposed policy implications for green business development. Based on the study goal, the authors surveyed 400 managers of green enterprises in Vietnam and applied simulation modelling based on the quantitative research method. The findings explored five critical factors affecting green business development: (1) financial incentives, (2) corporate strategy and culture, (3) technological innovation, (4) mechanisms and policies, and (5) regulatory environment. The novelty of this study is that it increased the level of reputation and trust of customers, partners, and the community. Green businesses built a positive brand image to affirm their social responsibility and pioneering role in environmental protection. Green businesses could also attract and retain talent when employees feel proud and attached to the business's mission. A significant benefit of green businesses was their contribution to protecting the living environment for humans and animals. The article’s contributions helped to reduce greenhouse gas emissions, air, water, and soil pollution, reduce waste, and increase energy and resource efficiency. Green business development can also support nature conservation activities, habitat restoration, and biodiversity maintenance. The article's contributions proposed recommendations to management agencies in developing and implementing mechanisms and policies to support and promoted businesses to build in a green direction.
Comparative Study of CNN-Based Architectures for Early Brain Tumor Diagnosis Lakshmi D; Pragash C; Malathy Batumalay; Karthick Manoj R
Journal of Applied Data Sciences Vol 7, No 1: January 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i1.920

Abstract

This study presents a comprehensive comparative analysis of Convolutional Neural Network (CNN)-based deep learning architectures for early brain tumor detection and classification using multi-modal medical imaging. The primary objective is to evaluate and integrate advanced deep neural network models, including EfficientNet-B2, VGG16, U-Net, and a hybrid CNN-LSTM, to enhance diagnostic accuracy, precision, and robustness. The proposed framework involves five key stages: image acquisition from MRI, CT, PET, and ultrasound modalities; preprocessing through normalization, skull stripping, noise reduction, and registration; segmentation of tumor regions; feature extraction; and classification using optimized deep learning algorithms. Experimental evaluation demonstrates that the hybrid CNN-LSTM model achieved the highest overall performance, with an accuracy of 98.81%, precision of 98.90%, recall of 98.90%, and F1-score of 99%. The EfficientNet-B2 model followed closely with 98.73% accuracy, 98.73% precision, 99.13% recall, and 98.79% F1-score, confirming its strength in efficient feature utilization and computational scalability. In contrast, VGG16 and U-Net achieved accuracies of 93.27% and 88%, respectively, indicating limited adaptability to complex tumor morphologies. The findings reveal that CNN-based hybrid models outperform traditional architectures by effectively capturing both spatial and temporal dependencies in MRI data, leading to improved interpretability and clinical reliability. The novelty of this research lies in its methodological integration of convolutional and recurrent layers within a unified diagnostic framework, establishing a reproducible, high-performance model for early brain tumor detection. The study contributes to the advancement of intelligent medical imaging systems by demonstrating that hybrid deep learning architectures can significantly reduce diagnostic uncertainty and enable more precise, automated clinical decision support for early intervention.
Improved Performance of Hybrid GRU-BiLSTM for Detection Emotion on Twitter Dataset M. Khairul Anam; Munawir Munawir; Lusiana Efrizoni; Nurul Fadillah; Wirta Agustin; Irwanda Syahputra; Tri Putri Lestari; Muhammad Bambang Firdaus; Lathifah Lathifah; Atalya Kurnia Sari
Journal of Applied Data Sciences Vol 6, No 1: JANUARY 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v6i1.459

Abstract

This study addresses emotion detection challenges in tweets, focusing on contextual understanding and class imbalance. A novel hybrid deep learning architecture combining GRU-BiLSTM with SMOTE is proposed to enhance classification performance on an Israel-Palestine conflict dataset. The dataset contains 40,000 tweets labeled with six emotions: anger, disgust, fear, joy, sadness, and surprise. SMOTE effectively balances the dataset, improving model fairness in detecting minority classes. Experimental results show that the GRU-BiLSTM hybrid with an 80:20 data split achieves the highest accuracy of 89%, surpassing BiLSTM alone, which obtained 88%, and other state-of-the-art models. Notably, the proposed model delivers significant improvement in detecting the emotion of joy (recall: 0.87, F1-score: 0.86). In contrast, the surprise category remains challenging (recall: 0.24). Compared to existing research, this study highlights the effectiveness of combining SMOTE and hybrid GRU-BiLSTM, outperforming models such as CNN, GRU, and LSTM on similar datasets. The incorporation of GloVe embeddings enhances contextual word representations, enabling nuanced emotion detection even in sarcastic or ambiguous texts. The novelty lies in addressing class imbalance systematically with SMOTE and leveraging GRU-BiLSTM's complementary strengths, yielding superior performance metrics. This approach contributes to advancing emotion detection tasks, especially in conflict-related social media data, by offering a robust, context-sensitive, and balanced classification method.
Data-Driven Evaluation of Task-Specific Performance in Refined Watercolor Brush Prototypes Using Repeated-Measures Analysis Suraphan Chantanasut
Journal of Applied Data Sciences Vol 7, No 3: September 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i3.1574

Abstract

Watercolor brush performance emerges from the interaction between prototype configuration, bristle characteristics, material transfer, and the requirements of specific painting tasks. This study presents a data-driven evaluation of ten refined watercolor brush prototypes by integrating professional expert validation with controlled repeated-measures performance testing. The study used a sequential two-stage design. In Stage 1, 15 professional watercolor artists evaluated five round-handled and five flat-handled refined prototypes on material quality, grip comfort, movement control, water retention and color distribution, and professional-use potential. The overall expert evaluation mean was 4.65/5.00. In Stage 2, 24 experienced watercolor practitioners tested all ten prototypes across precision, graded-wash, and mixed technical tasks, producing 720 primary trials. Objective outcomes included task completion time, line-deviation error, stroke-width variability, wash-gradient error, wash smoothness, and paint consumption; task-specific usability was recorded after each trial. All 720 outputs were independently rated by three blinded watercolor experts, yielding 2,160 technical-quality ratings. Round configurations showed markedly lower precision error (0.442 vs. 1.243 mm) and shorter precision-task times (48.64 vs. 75.53 s), whereas flat configurations showed lower wash-gradient error (3.43 vs. 13.32 CIE L* units), higher wash smoothness (95.72% vs. 83.35%), and shorter wash-task times (39.40 vs. 67.42 s). Mixed-task differences were substantially smaller. A linear mixed-effects model confirmed a strong family-by-task interaction in completion time, while trial order was not significant (p = 0.512). Subjective control and satisfaction followed the objective task-specific pattern, whereas hand fatigue and wrist discomfort did not differ meaningfully between families. Blinded rating reliability was excellent (ICC(2,1)=0.969; ICC(2,3)=0.990). The findings indicate complementary task-specific specialization rather than universal superiority of one prototype family and demonstrate the value of repeated-measures, multimetric evaluation for data-driven artistic-tool design.
Research on Deep Learning-Based Algorithm and Model for Personalized Recommendation of Resources Yu’e Liu
Journal of Applied Data Sciences Vol 4, No 2: MAY 2023
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v4i2.85

Abstract

Resource recommendation system is a new type of management system, which uses personalized information to solve business needs such as customer consultation and product recommendation, and provides users with high quality services and achieves accurate marketing, so nowadays resource recommendation system has a pivotal role in modern resource management. In this paper, I study the algorithm and model of resource personalized recommendation based on deep learning, taking human resource recommendation as an example.
Multi-Label Classification of Indonesian Voice Phishing Conversations: A Comparative Study of XLM-RoBERTa and ELECTRA Ahmad Hidayat; Sarifuddin Madenda; Hustinawaty Hustinawaty
Journal of Applied Data Sciences Vol 6, No 3: September 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v6i3.858

Abstract

Mobile phones have become a primary means of communication, yet their advancement has also been exploited by cybercriminals, particularly through voice phishing schemes. Voice phishing is a form of social engineering fraud carried out via telephone conversations to illegally obtain personal or financial information. The complexity of voice phishing continues to increase, as a single conversation may involve multiple fraudulent schemes simultaneously, necessitating the application of multi-label classification to comprehensively identify all motives of fraud. Previous studies have predominantly utilized single-label approaches and foreign-language data, making them less relevant to the Indonesian language context and unable to produce speaker segmentation outputs for conversational analysis. This study contributes by developing a multi-label voice phishing classification system specifically for Indonesian telephone conversations to address this gap. Audio data were collected from open sources and simulated recordings, resulting in a total of 300 samples labeled into six categories: five phishing modes and one non-phishing category. The proposed system consists of a preprocessing pipeline that includes noise reduction, speaker segmentation, automatic transcription, and text cleaning to preserve the context of two-way conversations. Two machine learning models based on transformer architectures, XLM-RoBERTa and ELECTRA, are employed to identify various fraud schemes that may occur simultaneously within a single conversation. The dataset was split into training, validation, and testing sets with two division ratios for performance evaluation. Several combinations of hyperparameters were tested to obtain the most optimal model configuration. Evaluation was conducted using a supervised learning approach and various performance metrics. The experimental results show that XLM-RoBERTa achieved the highest average accuracy of 97.04 ± 1.15% and the highest average F1-score of 92.66 ± 2.59%. These results highlight the novelty of applying multi-label classification in the Indonesian language context for voice phishing detection, contributing to more effective fraud identification in real-world telephony systems.
Machine Learning Algorithm Optimization using Stacking Technique for Graduation Prediction Herianto Herianto; Bambang Kurniawan; Zupri Henra Hartomi; Yuda Irawan; M Khairul Anam
Journal of Applied Data Sciences Vol 5, No 3: SEPTEMBER 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i3.316

Abstract

Graduating on time is crucial for academic success, impacting time, costs, and education quality. Hang Tuah University Pekanbaru (UHTP) is currently struggling to meet its goal of achieving a 75% on-time graduation rate. This study introduces an innovative approach using machine learning techniques, particularly ensemble learning with Stacking Machine Learning Optuna SMOTE (SMLOS), to address this issue. Our primary objective is to enhance data classification accuracy to predict student graduation timelines effectively. We employ algorithms such as K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Decision Tree (C4.5), Random Forest (RF), and Naive Bayes (NB). These were combined with meta-models, including Logistic Regression (LR), Adaboost, XGBoost, LR+Adaboost, and LR+XGBoost, to create a robust prediction model. To address class imbalance, we applied the Synthetic Minority Over-sampling Technique (SMOTE) and utilized Optuna for hyperparameter tuning. The findings reveal that SMLOS with the Adaboost meta-model achieved the highest accuracy of 95.50%, surpassing previous models' performances, which averaged around 85%. This contribution demonstrates the effectiveness of using SMOTE for class imbalance and Optuna for hyperparameter optimization. Integrating this model into UHTP's academic information system facilitates real-time monitoring and analysis of student data, offering a novel solution for promoting a Smart Campus through more accurate student performance predictions. This technique is not only beneficial for predicting student graduation but can also be applied to various machine learning tasks to improve data classification accuracy and stability.
Utilization of K-means Clustering for Classifying Diabetes Risk Populations According to Health Behaviors and 3Es-2Ss Health Literacy Supaporn Yodmunee; Wongpanya S. Nuankaew; Thapanapong Sararat; Pratya Nuankew
Journal of Applied Data Sciences Vol 7, No 2: May 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i2.1042

Abstract

This study focused on classifying populations at risk for diabetes using K-means clustering integrated with the 3Es–2Ss health literacy framework: eating, exercise, emotion, smoking cessation, and alcohol cessation. Biological, behavioral, and health literacy data were analyzed. The dataset was collected from 126 participants identified as at-risk individuals in Ngao District, Lampang Province, Thailand. This relatively small, community-based sample provides valuable insights into local health behaviors but limits the generalizability and statistical power of the findings to broader populations. The K-means clustering analysis, guided by the Elbow method, identified k = 4 as the optimal number of clusters, yielding four distinct groups with different socio-demographic and health characteristics. These clusters revealed variations in health profiles, economic status, and behavioral literacy within the Thai population. Despite the small sample size and limited generalizability, missing data and inconsistencies were systematically addressed through data cleaning and normalization to maintain analytical reliability. The results suggest that K-means clustering can serve as an effective decision-support tool for public health planning, particularly for Non-Communicable Disease (NCD) prevention and diabetes management at the local level.

Page 83 of 127 | Total Record : 1270