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
Convolutional Neural Network Based Deep Learning Model for Accurate Classification of Durian Types Diana Diana; Tri Basuki Kurniawan; Deshinta Arrova Dewi; Mashal Kasem Alqudah; Musab Kasim Alqudah; Mohd Zaki Zakari; Eyna Fahera Binti Eddie Fuad
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.v6i1.480

Abstract

Durian recognition is significant among fans of the durian community since many people tend to get confused, especially if they are not familiar with durian species, which can lead them to be involved in durian fraud. The development of this prototype can detect and classify durian fruits into three categories, including Musang King, Black Thorn, and D24, which can significantly benefit consumers. The prototype in this research involves training using a dataset of durian images, specifically in Musang King, Black Thorn, and D24 varieties. Preprocessing techniques such as resizing and scaling data are applied to enhance the quality and consistency of the dataset. The models chosen to develop this prototype include VGG-16 and Xception, and each model is compared according to its accuracy percentage. The accuracy outcomes of VGG-16 and Xception models are 56.64% and 92%, respectively. The models used a total of 1,372 images of durian with three classifications. Based on the findings, further enhancement of the CNN models for durian classification can be done by implementing different architectures, techniques, and methods. Moreover, future models can consider real-time image capture and processing capabilities to enhance the practicality of the system for durian consumers. The prototype developed in this study demonstrates the feasibility of using deep learning techniques for accurate and efficient durian classification, paving the way for future advancements in automated fruit grading and quality control systems in the durian industry.
A New Data Preprocessing Framework to Enhance the Accuracy of Herbal Plants Classification Using Deep Learning Attapol Kunlerd; Atipat Ritthiron; Boonlueo Nabumroong; Sakchan Luangmaneerote; Anyawee Chaiwachirakhampon; Jakkrit Kaewyotha
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.v6i3.733

Abstract

This research proposes to solve the problem of herbal plant classification, which plays a key role in Thai pharmacy and traditional medicine. Moreover, there are limitations due to similar physical characteristics of plants and the reliance on specialists to classify herbal plants, which hinder the utilization of herbal plants by the general public at the local level. To solve this problem, this research presents a new preprocessing framework called P4, which integrates 7 techniques as follow: Image Cropping, Resizing, Normalization (0–1), Data Augmentation, Label Noise, Label Cleaning, and Dataset Quality Score (DQS). The prominent point of P4 technique is the combination of intentional mislabeling and label cleaning process, as well as, quantitative data quality assessment and additional expert review in order to filter out potentially inaccurate data before inputting to Deep Learning model. In the experiment, a dataset of 4,211 herbal images covering 30 herbal plant species is used and compared with 3 proposed techniques in previous research (P1–P3) with 5 deep learning architectures, namely DenseNet201, EfficientNetB7, ViT, Swin Transformer, and ConvNeXt. The experimental results showed that the P4 technique combined with DenseNet201 model provided the highest performance in herbal plant classification, with an Accuracy of 92%, Precision of 92%, Recall of 91%, and a training time of merely 22.92 minutes. This was a result of combining the good data quality from the P4 technique, which enhanced to increase efficiency in producing higher quality and more balanced data. When combined with the structural capability of DenseNet201 that supported feature reuse from previous layers, it increased the robustness to mislabeled data and was able to accurately distinguish plants with similar characteristics. The results of this experiment are able be applied as a guideline for future application in Thai traditional medicine support system and herbal plant learning system.
To Retrench or Invest? Evaluating the Turnaround and Recovery Strategies of Indonesia MNEs through Data Science Approaches Willy Abdillah; Delly Nofiani; Maria Paramastri Hayuning Adi; Eka Marlina
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.v5i3.270

Abstract

This study aims to investigate the turnaround and recovery strategies employed by Ecolab International Indonesia MNEs after facing significant financial decline caused by the COVID-19 pandemic. This research analyzed internal factors, external opportunities, and threats to define effective strategies using a multi-method approach. Qualitative interviews were conducted to identify key themes, supported by examining company publications, especially annual financial reports from 2018 to 2022, to understand economic trends before, during, and after the pandemic. Thematic analysis was utilized to analyze the results, involving coding interview transcripts using ATLAS.ti and validating these themes through member checking to ensure reliability. Our findings show that Ecolab's turnaround strategy (cost reduction, enhancing value in sustainability, restructuring leadership, and organizational culture) was essential in addressing immediate and long-term challenges. The recovery strategy (operational and financial strategies, strategic focus area, regulatory and market dynamics) helped the company navigate the pandemic's impact and its position for sustainable growth. This study breaks new ground by integrating sustainability into strategic frameworks, aligning with global trends. Offering fresh perspectives enhances the relevance and value of MNEs' corporate strategy research in emerging markets. Additionally, our findings provide actionable insights for other companies to effectively incorporate sustainable practices into their turnaround and recovery strategies, ensuring long-term growth and regulatory compliance.
Multimodal AI Framework for Sign Language Recognition and Medical Informatics in Hearing-Impaired Patients Pratya Nuankaew; Parin Khamthep; Patdanai Jaitem; Kuljira S. Nuankaew; Kaewpanya S. Nuankaew; Wongpanya S. Nuankaew
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.v7i1.1096

Abstract

This study assesses the feasibility of YOLO-based detectors for the recognition of Thai Sign Language (TSL) within clinical intake workflows. We benchmark YOLOv5 through YOLOv10 over 100 to 150 training epochs and evaluate metrics including Precision, Recall, mAP@50, mAP@50:95, alongside training and validation losses to gauge stability. The losses decrease steadily as detection metrics improve; YOLOv10 offers the optimal balance, with Precision at 0.953, Recall at 0.939, mAP@50 at 0.933, and mAP@50:95 at 0.492. The improvements observed at stricter IoU thresholds are modest, underscoring ongoing challenges in achieving accurate localization and generalization across varying lighting conditions, viewpoints, occlusions, and motion. YOLOv11 has been excluded from the primary results due to abnormal loss behavior. These findings endorse a multimodal pipeline that employs an image-based detector as the central perception component, supplemented with pose and key point cues, as well as OCR and NLP layers, to transform recognized signs into structured medical intents for triage and telemedicine applications. Future research will focus on expanding sequence-level evaluation, incorporating dialects and co-articulation in TSL, and developing compressed or distilled models to facilitate reliable on-device inference in resource-constrained environments. 
Classification of Political Party Conflicts and Their Mediation Using Modified Recurrent Convolutional Neural Network Slamet Riyadi; Muhamad Arief Previasakti Suradi; Cahya Damarjati; Hsing-Chung Chen; Ridho Al-Hamdi; Ahmad Musthafa Masyhur
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.v6i1.513

Abstract

The rapid proliferation of political information on the internet has exacerbated conflicts within political parties, including elite disputes, dualism, candidate controversies, and management issues, which can undermine political stability and public trust. To address these challenges, this study introduces the Modified Recurrent Convolutional Neural Network (M-RCNN), an enhanced RCNN model designed to improve classification accuracy and mitigate overfitting by incorporating additional layers and dropout mechanisms. The primary objective of this research is to provide an efficient and accurate framework for classifying political conflicts and mediation strategies, overcoming the limitations of traditional methods, particularly in handling imbalanced datasets and intricate data patterns. Using a dataset of 1,106 Indonesian news articles categorized into four conflict types—elite disputes, management, presidential, and legislative candidate conflicts—and four mediation strategies—leadership decisions, deliberation, legal channels, and none—the data underwent extensive preprocessing, tokenization, and an 80:20 training-testing split. The M-RCNN achieved a conflict classification accuracy of 98.0%, a precision of 99.0%, and a loss of 0.03, significantly outperforming baseline models, including CNN (85.0% accuracy), RNN with LSTM (88.0%), and standard RCNN (85.0%). For mediation strategy classification, the model demonstrated exceptional performance with an accuracy of 99.0%, a precision of 99.0%, and a loss of 0.01, highlighting its robustness and scalability. This study’s novelty lies in its ability to process imbalanced and complex datasets with unparalleled precision and efficiency, providing a practical framework for automated political conflict analysis and mediation. The findings underline the potential of the M-RCNN model to revolutionize political science applications by delivering reliable, fast, and accurate tools for analyzing and resolving political conflicts, thereby contributing to the advancement of artificial intelligence in promoting political stability and fostering public trust.
A Hybrid Machine Learning Model for Short-Term Mobile Traffic Forecasting Adewale Ajao Adeyinka; Joan Paul Ezra; John Simon Wejin; Wou Onn Choo; Maksetbay Mambetniyazov
Journal of Applied Data Sciences Vol 7, No 4: December 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i4.1121

Abstract

Today, mobile networks are essential for seamless communication. As mobile subscriber numbers increase, mobile network operators must devise strategies to meet the enormous demand for mobile network resources, such as spectrum. Therefore, the need for an effective prediction mechanism for efficient resource management and planning becomes paramount. To address the challenges of non-linearity, burstiness, and hidden temporal patterns in mobile traffic prediction, this paper develops and evaluates a hybrid machine learning model for short-term forecasting of mobile traffic congestion. Real-life base-station mobile data traffic for the first quarter of 2024 from northern and southern Kaduna State, Nigeria, was used. The 40573-row, 10-column entry dataset was preprocessed to remove duplicates, outliers, and missing values. From the processed dataset, 70% was used to train various machine learning models. Long Short-Term Memory (LSTM) and Accelerated Gradient Boosting (AGB), with RMSEs 12% and 15% lower than those of other models, were selected via voting to develop the proposed hybrid prediction model. Thirty per cent (30%) of the processed dataset was then employed to test and evaluate the prediction performance of the proposed model on nine base stations with both the highest and lowest traffic using Root Mean Square (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE) as evaluation metrics. From the results, LSTM-AGB outperformed LSTM and ARIMA, achieving the lowest RMSE, MAPE, and MAE values of 2354, 8.04%, and 196, for both uplink and downlink prediction evaluations. These findings demonstrate the effectiveness of the proposed model for short-term mobile traffic prediction by enabling proactive congestion management through early identification of imminent traffic surges, thereby enabling preemptive adjustments to resource allocation and network planning. The proposed hybrid model provides a scalable, data-driven approach that can support the efficient operation of modern mobile stations, especially in resource-constrained environments.
Unsupervised Learning Methods for Topic Extraction and Modeling in Large-scale Text Corpora using LSA and LDA Henderi Henderi; B Herawan Hayadi; Sofa Sofiana; Padeli Padeli; Didik Setiyadi
Journal of Applied Data Sciences Vol 4, No 3: SEPTEMBER 2023
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v4i3.102

Abstract

This research compares unsupervised learning methods in topic extraction and modeling in large-scale text corpora. The methods used are Singular Value Decomposition (SVD) and Latent Dirichlet Allocation (LDA). SVD is used to extract important features through term-document matrix decomposition, while LDA identifies hidden topics based on the probability distribution of words. The research involves data collection, data exploratory analysis (EDA), topic extraction using SVD, data preprocessing, and topic extraction using LDA. The data used were large-scale text corpora. Data explorative analysis was conducted to understand the characteristics and structure of text corpora before topic extraction was performed. SVD and LDA were used to identify the main topics in the text corpora. The results showed that SVD and LDA were successful in topic extraction and modeling of large-scale text corpora. SVD reveals cohesive patterns and thematically related topics. LDA identifies hidden topics based on the probability distribution of words. These findings have important implications in text processing and analysis. The resulting topic representations can be used for information mining, document categorization, and more in-depth text analysis. The use of SVD and LDA in topic extraction and modeling of large-scale text corpora provides valuable insights in text analysis. However, this research has limitations. The success of the methods depends on the quality and representativeness of the text corpora. Topic interpretation still requires further understanding and analysis. Future research can develop methods and techniques to improve the accuracy and efficiency of topic extraction and text corpora modeling.
Fine-Grained Sentiment Analysis Approach on Customer Reviews Based on Aspect-Level Emotion Detection Adi Suryaputra Paramita; Jusak Jusak
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.v6i3.964

Abstract

In the era of digital platforms, customer reviews constitute a vital resource for understanding user sentiment and perception toward products and services. Traditional sentiment analysis methods predominantly operate at the document or sentence level, often missing fine-grained emotional cues tied to specific product or service aspects. To address this limitation, this study proposes a novel Fine-Grained Sentiment Analysis (FGSA) framework that performs aspect-level sentiment classification using a joint learning approach. The proposed model employs a hybrid deep learning architecture that integrates transformer-based contextual encoders with Bidirectional Long Short-Term Memory (Bi-LSTM) layers. This design allows the model to capture both rich contextual semantics and sequential dependencies a combination that has not been widely adopted in existing FGSA research. Additionally, we introduce a new annotated dataset of 5,000 customer reviews spanning multiple domains (electronics, food and beverages, and general services), enabling robust training and evaluation. Experimental results show that the model outperforms standard baselines, achieving an F1-score of 82.0% for aspect extraction and an accuracy of 79.8% for sentiment classification. Further analysis reveals consistent patterns, such as positive sentiments linked to design and quality, and negative sentiments associated with customer service and delivery. These insights highlight the practical value of aspect-level sentiment modelling. The key contribution of this work is the integration of a transformer-Bi-LSTM joint architecture for aspect-based sentiment analysis, supported by a domain-diverse benchmark dataset. This framework enhances the interpretability and granularity of sentiment insights and sets a foundation for future research in multilingual and multimodal contexts.
Sustainable Educational Data Mining Studies: Identifying Key Factors and Techniques for Predicting Student Academic Performance Murnawan Murnawan; Sri Lestari; Rosalim Samihardjo; Deshinta Arrova Dewi
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.v5i3.347

Abstract

This research paper presents a systematic literature review of sustainable educational data mining (EDM) studies published between 2017 and 2022 with the objective of identifying the primary factors that affect student academic performance. The purpose of this study is to provide a comprehensive analysis of sustainable EDM research and identify the most important factors that influence student performance while highlighting commonly used data mining techniques in the EDM field. The results suggest that student demographics, previous grades and class performance, social factors, and online learning activities are the most common and widely used factors for predicting student performance in educational institutions. Furthermore, Decision Trees, Naive Bayes, and Random Forests are the most frequently used categories of data mining algorithms in the studies included in the dataset. The methodology used in this study is a systematic literature review, which is a widely used technique for literature review that provides a reliable and unbiased process for reviewing data from diverse sources. The findings of this study provide valuable insights into the factors influencing student performance in educational institutions and can be used by researchers to inform future research and identify relevant factors to consider when predicting student performance.
An Integrated Text Analytics and Ensemble Machine Learning Framework for Fake Review Detection in Online Marketplaces Eka Praja Wiyata Mandala; Sarjon Defit; Gunadi Widi Nurcahyo
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.v7i2.1143

Abstract

The increasing prevalence of fake reviews on e-commerce platforms undermines consumer trust and affects purchasing decisions, particularly for local products by limited visibility such as those by West Sumatra, Indonesia. This study proposes a hybrid approach combining text analytics and machine learning to enhance the detection of fake reviews. Four classification models—Naive Bayes, Random Forest, Logistic Regression, and K-Nearest Neighbor—were tested on a dataset of 1,500 labeled product reviews. Among these models, Random Forest had the highest starting accuracy of 0.8533. To enhance it, we created a better algorithm called EKAHypeRFor (Enhanced Knowledge Augmentation of Hyperparameter Random Forest). This method uses simple feature engineering and careful tuning of settings by RandomizedSearchCV. The enhanced model reached an accuracy of 0.8778, which is 2.45% higher than the original. It also includes a real-time review sorting tool, making it easy to use on online shopping sites. Tests by a confusion matrix and feature importance drawn the model works well and is easy to understand. This method is simple, fast, and accurate, helping to make online product reviews more trustworthy for small and medium businesses in the area.

Page 75 of 127 | Total Record : 1270