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
A Hybrid TF-IDF and Knowledge Graph-Enhanced Retrieval-Augmented Generation Framework with Large Language Models for Domain-Aware Question Answering Lilyani Asri Utami; Hilda Rachmi; Syarif Hidayatulloh
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.v7i2.1136

Abstract

This study aims to develop a domain-aware legal Question-Answering (QA) system tailored for Indonesia’s Micro, Small, and Medium Enterprises (MSMEs) by proposing a hybrid Retrieval-Augmented Generation (RAG) framework that integrates Term Frequency–Inverse Document Frequency (TF-IDF), Knowledge Graph (KG), and Large Language Model (LLM) components. In this framework, TF-IDF contributes by performing lexical-level retrieval to identify the most relevant documents based on keyword weighting; the KG enriches this retrieval by providing semantic relationships among legal entities, enabling deeper contextual understanding; and the LLM generates coherent responses conditioned on both lexical and semantically grounded evidence. Together, these components work synergistically to strengthen factual grounding during retrieval and improve contextual reasoning during generation. Methodologically, the system processes a curated dataset of 1,400 legal question–answer pairs collected from national legal repositories, including legislation, government regulations, and MSME digitalization guidelines. The process includes text preprocessing, keyword extraction using TF-IDF, semantic enrichment through a KG that maps legal entities and their relationships, and answer generation via an LLM powered by the RAG pipeline. The system was evaluated using Precision, Recall, F1-Score, Bilingual Evaluation Understudy (BLEU), and Recall-Oriented Understudy for Gisting Evaluation (ROUGE) metrics, validated by five legal experts. Results show an accuracy improvement from 76.5% to 83.5% after integrating KG, with Precision of 0.853, Recall of 0.877, and F1-Score of 0.865. The generative evaluation yielded a BLEU score of 0.9276 and ROUGE-L of 0.9301, indicating strong linguistic and semantic alignment between system outputs and expert-authored references. The study concludes that this approach offers a practical foundation for building AI-based legal assistance tools and highlights future opportunities for expansion to other legal domains and multilingual RAG applications.
High-Accuracy Stroke Detection System Using a CBAM-ResNet18 Deep Learning Model on Brain CT Images Imam Tahyudin; R Rizal Isnanto; Anton Satria Prabuwono; Taqwa Hariguna; Eko Winarto; Nazwan Nazwan; Ades Tikaningsih; Puji Lestari; Rofik Abdul Rozak
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.v6i1.569

Abstract

Stroke is a brain dysfunction that occurs suddenly as a result of local or overarching damage to the brain, lasts for at least 24 hours, and causes about 15 million deaths each year globally. Immediate medical treatment is essential to reduce the potential for further brain damage in stroke patients. Medical imaging, especially computed tomography (CT scan), plays a crucial role in the diagnosis of stroke. This study aims to develop and evaluate a deep learning architecture based on Convolutional Block Attention Module (CBAM) and ResNet18 for stroke classification in CT images. This model is designed through data preprocessing, training, and evaluation stages using a cross-validation approach. The results showed that the CBAM-ResNet18 integration resulted in a high accuracy of 95% in distinguishing stroke and non-stroke cases. The accuracy rate reached 96% for nonstroke identification (class 0) and 94% for stroke (class 1), with recall rates of 96% and 93%, respectively. Outstanding classification ability is demonstrated by an Area Under the Curve (AUC) value of 0.99. In comparison, the standard ResNet18 model shows significant fluctuations in validation loss and difficulty in generalization, with training accuracy only reaching 64-68%. On the other hand, CBAM-ResNet18 showed a significant performance improvement with a validation accuracy of 95%, a validation loss of 0.0888, and good generalization on new data. However, the limitations of the dataset and the interpretation of the results indicate the need for further validation to ensure the generalization of the model. These results show the great potential of the CBAM-ResNet18 architecture as an innovative tool in stroke diagnostic technology based on CT imaging analysis. This technology can support faster and more accurate clinical decision-making, as well as open up opportunities for further research related to the development of artificial intelligence-based systems in the medical field.
Gold Prices Time-Series Forecasting: Comparison of Statistical Techniques Indra Maryati; Christian Christian; Adi Suryaputra Paramita
Journal of Applied Data Sciences Vol 4, No 4: DECEMBER 2023
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v4i4.135

Abstract

The fluctuation of gold prices throughout the year makes it difficult for both investors and regular individuals to predict the future value. The goal of this research is to utilize various statistical techniques, such as linear regression, naive bayes, and various types of smoothing algorithms, to predict the price of gold. The data used in this study was obtained from Kaggle and is from a 70-year time period. The results showed that using a single exponential smoothing method had the highest accuracy and precision, with a good MAPE score of 7.12%. This study is unique in that it compares multiple algorithms using data over a long time period, and it can be useful for investors and traders in making decisions related to gold prices. Additionally, it can also serve as a reference for future research studies.
A Data-Driven MINLP Approach for Enhancing Sustainability in Blockchain-Enabled e-Supply Chains Afif Badawi; Syahril Efendi; Tulus Tulus; Herman Mawengkang
Journal of Applied Data Sciences Vol 6, No 4: December 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v6i4.889

Abstract

Modern e-supply chains are characterized by increasing complexity and a critical need for enhanced sustainability, transparency, and traceability. Blockchain technology emerges as a significant enabler, offering decentralized, immutable ledgers and smart contracts that can support more secure, verifiable, and environmentally responsible operations through trustworthy data. Despite blockchain's potential, a notable gap exists in the availability of quantitative, data-driven optimization models that rigorously assess the operational and sustainability impacts of its integration into e-supply chains, particularly for complex, non-linear system interactions. This study aims to address this gap by presenting an in-depth analysis of a specific Mixed-Integer Non-Linear Programming (MINLP) optimization model. The goal is to clarify its structure, evaluate its application for an e-supply chain incorporating blockchain features (like transaction costs and conceptual smart contract enforcement for compliance) and sustainability objectives (such as carbon emission reduction), and derive practical insights. The methodology involves a detailed exposition of the MINLP model, followed by its application to a defined e-supply chain scenario. The analytical approach includes computational experiments focusing on a base case analysis to demonstrate model functionality. The broader evaluative framework for this study encompasses benchmarking the model’s performance against a conventional system and conducting sensitivity analyses on key parameters to understand performance trade-offs. The initial base case analysis demonstrates the model's capability to optimize supplier selection and operational plans while adhering to sustainability constraints, such as maintaining carbon emissions at or below 300 kg CO₂ per period, and accounting for blockchain-specific costs like a per-supplier usage fee of 500. The structure of the model and preliminary insights suggest its potential to achieve improved environmental impact compared to conventional systems, balanced against associated blockchain implementation costs. This research provides a detailed examination of a complex MINLP structure, offering a data-driven analytical approach for assessing blockchain's role in sustainable e-supply chains. It furnishes a foundational framework and insights that can guide managerial decisions and strategic planning for industries transitioning towards greener, more transparent, and digitally advanced supply chain operations.
Environment Sentiment Analysis of Bali Coffee Shop Visitors Using Bidirectional Encoder Representations from Transformers (BERT) and Generative Pre-trained Transformer 2 (GPT2) Model Ni Putu Widya Yuniari; Ni Made Satvika Iswari; I Made Surya Kumara
Journal of Applied Data Sciences Vol 5, No 4: DECEMBER 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i4.302

Abstract

Bali is one of the provinces with the most abundant natural and cultural wealth in Indonesia. One commodity that supports it is coffee. Bali Coffee is not only a gastronomic identity, but also a cultural identity which makes it have added value to be developed into various business lines. One business derivative that is quite promising is a coffee shop. However, these favorable conditions also need to be maintained to ensure good quality reaches consumers. One thing that can do is analyze reviews from customers. One of the most popular methods is Sentiment Analysis. This technique allows business to analyze customer reviews on social media. It can be a feedback to maintaining and improving quality and good relationships with customers. This research aims to create a machine learning model to analyze customer reviews at several coffee shops in Bali which are divided into three labels, namely: positive, negative and neutral. The methods used are: scraping, cleaning, stopword removal, embedding, undersampling, and modeling. The algorithms used are Bidirectional Encoder Representation from Transformer (BERT) and Generative Pre-trained Transformers (GPT). The performance metrics used in this research are precision, recall, accuracy and loss. This research succeeded in creating a sentiment analysis model for coffee shop customers in Bali. The BERT model obtained an accuracy value of 78% without undersampling with a loss in the 10th iteration of 0.27. Meanwhile, the BERT model with undersampling obtained an accuracy value of 32.85% with a loss in the 10th iteration of 0.16. The GPT2 model without undersampling gets an accuracy of 78% with a loss in the 10th iteration of 0.25. Meanwhile, the GPT model with undersampling obtained an accuracy value of 32.85% with a loss in the 10th iteration of 0.15.
Monkeypox Disease Classification Based on Skin Images Using Hierarchical Swin Transformer-Based Convolutional Neural Network Approach Putu Desiana Wulaning Ayu; Sasiwimol Sukket; Sutikno Sutikno; Putu Manik Prihatini; Gede Angga Pradipta; Dandy Pramana Hostiadi
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.v7i2.1313

Abstract

Monkeypox diagnosis can initially be conducted through expert physical examination based on characteristic lesions. However, laboratory confirmation using PCR is still essential, these tests are often hampered by limitations such as high costs, lengthy processing times, and a general lack of detailed symptom knowledge among patients. In light of these issues, image-based diagnostic methods offer a more efficient solution, given that monkeypox manifests as visible lesions on the skin that can be accurately detected using a deep learning. This study employs Transformer network-based deep learning for classifying skin diseases. To improve model robustness and mitigate the limitations of the relatively small dataset, we designed a comprehensive data augmentation pipeline that incorporates both positional and color transformations, including rotation, horizontal and vertical flipping, zooming, shearing, and brightness, contrast, hue, and saturation adjustments. Furthermore, a k-fold cross-validation strategy was employed, where the entire dataset was partitioned into k equal-sized folds to ensure a reliable and unbiased evaluation of the model performance. The Swin Transformer leverages advanced transformer network to analyze images, emphasizing hierarchical relationships within images. Swin Transformer enhances the convolutional Transformer architecture by substituting the standard multi-head-self-attention (MSA) mechanism with a shifted window-based MSA module It enhances efficiency over traditional transformer models by incorporating a shifted window mechanism, which reduces computational demands. The average global accuracy achieved was 0.99 (99%), which is further supported by the AUC values obtained for each disease category. The model achieved an AUC of 1.00 for chickenpox, cowpox, and hand-foot-mouth disease (HFMD), indicating excellent discriminative capability for these classes. Meanwhile, the remaining classes, including healthy skin, measles, and monkeypox, achieved AUC values of 0.99 and 0.98, respectively. These results demonstrate that the proposed Hierarchical Swin Transformer model provides highly reliable classification performance across all skin disease categories included in the dataset.
Blended Teaching Model Optimization for Innovation and Entrepreneurship Courses through Data Analytics in Higher Education Liu Yang; Thosporn Sangsawang; Naruemon Thepnuan; Nawaphas Chankham; Thidarat Kulnattarawong
Journal of Applied Data Sciences Vol 6, No 2: MAY 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v6i2.734

Abstract

This study aimed to (1) develop a blended teaching model for Innovation and Entrepreneurship courses in Chinese higher education, and (2) assess the effectiveness of the proposed model. The sample consisted of 17 Chinese experts selected through purposive sampling and 30 higher education students from China. The research employed statistical analysis techniques including mean, standard deviation, coefficient of variation, and t-test to analyze the data. Results demonstrated significant improvements in students' entrepreneurship skills. In the experimental group, the pre-test mean score increased from 2.21 to 3.78 post-intervention, while the control group showed a slight improvement from 2.32 to 2.84. The standard deviation of learning outcomes decreased from 0.884 to 0.564, indicating a more consistent student performance. A statistically significant difference was observed (p = 0.003), confirming the effectiveness of the blended teaching model. These findings highlight the potential of blended learning in enhancing the quality of innovation and entrepreneurship education.
Statistical Approach to Evaluating the Efficacy of Career Guidance Programs on University Graduate Employability in China Li Guo; Thosporn Sangsawang; Piyanan Pannim Vipahasna; Matee Pigultong; Sulaganya Punyayodhin; Kanokwan Darboth
Journal of Applied Data Sciences Vol 5, No 1: JANUARY 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i1.172

Abstract

This study aimed to develop a career guidance model for improving employment ability among Chinese undergraduate students and assess the impact of this model on students’ employment ability. The research involved 17 Chinese experts and 100 instructors from 10 universities in Sichuan, China. The Delphi technique was employed to gather expert perspectives, while data on employment ability were collected using the College Student Employment Ability Questionnaire. The Cronbach's coefficient of the questionnaire is .869, and Cronbach's α .80 indicates excellent internal consistency, affirming the authenticity and credibility of the data in this study. Based on the statistical criteria defined from the results of the fourth-round inquiries, each Course needs to meet any two of the following conditions: arithmetic x ̅ 3.5964, Full Score Rate .1020, and Cronbach's α .3883 to be preliminarily retained. The results of the third-round expert inquiries show that the course offerings meet the Arithmetic x ̅ 3.3548 criteria, Full Score Rate .1987, and Cronbach's α .5590. The study found a significant improvement in students’ employment ability after participating in the model, with the average score increasing from 16.11 to 20.33. These results underscore the effectiveness of targeted career guidance in enhancing undergraduate students’ employment prospects. Most experts have passed all courses and course content by this round, with viable ideas identified. Career Education and Orientation received the highest response percentage (90.67%), followed by self-assessment (89.50%), industry-oriented skill development (87.50%), mentor support and networking (85.50%), industry insights and trend analysis (89.50%), job search and application assistance (90.80%), continuous review and improvement (87.50%), and follow-up counseling and support (89.50%).
Exploring User Experience of an Interactive LMS for Green Entrepreneurship: An Empirical Study of the GEJUR Platform Florentina Kurniasari; Jansen Wiratama; Elissa Dwi Lestari; Andrey Andoko; Antonius Sony E N; Muhammad Fadhlan; Yohanes Brian Caesaryano Lala; Fayed Abdul Hakim; Akhmad Zainal Mubarak
Journal of Applied Data Sciences Vol 6, No 4: December 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v6i4.1041

Abstract

This article aims to assess the design, development, and User Experience (UX) of GEJUR, a web-based interactive Learning Management System (LMS) designed to foster green entrepreneurship among youth in Nusa Tenggara Timur (NTT), Indonesia. This study defines green entrepreneurship as entrepreneurial activities that combine economic value creation with environmentally responsible practices that promote sustainability in business. This research enhances the domains of Human–Computer Interaction (HCI) and entrepreneurship education by illustrating how user-centered digital platforms may cultivate youth business competencies while promoting the Sustainable Development Goals (SDGs). GEJUR incorporates interactive learning modules, e-mentoring, and specific cultural features like weaved motifs (pola tenun), providing contextual relevance alongside essential LMS functions. A mixed-methods approach was employed, integrating surveys, Focus Group Discussions (FGDs), semi-structured interviews, and usability testing. Quantitative data were gathered via the System Usability Scale (SUS), whilst qualitative data documented user impressions of obstacles and anticipations. The usability evaluation, conducted with 33 individuals, resulted in a mean SUS score of 48.9, which corresponds to approximately 71.9% of the benchmark score of 68. This positions GEJUR beneath the established threshold, classified as having “poor” usability, although it remains functionally workable. Participants effectively accomplished critical tasks including authentication, course access, and quizzes, and offered favorable feedback regarding the system’s cultural integration. The results highlighted critical areas for enhancement, such as interface uniformity, minimization of response times, and more explicit user feedback. The study finds that GEJUR offers a viable yet improvable foundation for digital entrepreneurship training platforms in underprivileged areas. Subsequent study ought to broaden testing to encompass bigger and more heterogeneous user populations and implement sophisticated UX evaluation criteria beyond the SUS. Through iterative refinement, GEJUR can develop into a scalable platform that facilitates youth empowerment, sustainable business practices, and extensive socio-economic advancement.
Efficient Fruit Grading and Selection System Leveraging Computer Vision and Machine Learning Deshinta Arrova Dewi; Tri Basuki Kurniawan; Rajermani Thinakaran; Malathy Batumalay; Shabana Habib; Muhammad Islam
Journal of Applied Data Sciences Vol 5, No 4: DECEMBER 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i4.443

Abstract

Automated fruit grading is crucial to overcoming the time and accuracy challenges posed by manual methods, which are often limited by subjective human judgment. This study introduces an intelligent grading system leveraging computer vision and AI to improve speed and consistency in assessing fruit quality. Using high-resolution imaging and advanced feature extraction, including grayscale processing, binarization, and enhancement, the system achieves non-destructive, efficient sorting for fruits like apples, bananas, and oranges. Grayscale processing reduces image complexity while preserving essential details, binarization isolates the fruit from its background, and enhancement highlights critical features. Notably, the Edge Pixel method proved most effective, achieving 79.20% accuracy in grading, while the Grayscale Pixel method reached 93.94% accuracy for fruit types. Edge Pixel also achieved 80.32% in differentiating grading types, showcasing its ability to capture essential shapes and edges. Fruits are classified into four grades: Grade_01 (highest quality), Grade_02 (minor imperfections), Grade_03 (notable defects but consumable), and Grade_04 (unfit for consumption). A specialized dataset supports model training, ensuring practical real-world application. The study concludes that this automated system offers significant improvements over traditional grading, providing a scalable, objective, and reliable solution for the agricultural sector, ultimately enhancing productivity and quality assurance.