cover
Article Per Year (5 Year)
All
Home Page
OAI Link
Editorial Team
Contact
Reviewer
Google Scholar
Contact Name
-
Contact Email
-
Phone
-
Journal Mail Official
-
Editorial Address
-
Location
,
INDONESIA
JOURNAL OF APPLIED INFORMATICS AND COMPUTING
Published by Politeknik Negeri Batam
ISSN : -     EISSN : 25486861     DOI : 10.3087
Core Subject : Science,
Computer Science & IT
Journal of Applied Informatics and Computing (JAIC) Volume 2, Nomor 1, Juli 2018. Berisi tulisan yang diangkat dari hasil penelitian di bidang Teknologi Informatika dan Komputer Terapan dengan e-ISSN: 2548-9828. Terdapat 3 artikel yang telah ditelaah secara substansial oleh tim editorial dan reviewer.
Arjuna Subject : -
Articles 1,006 Documents
 87   88   89   90   91 
ResNet50-Based Mobile Application for Big Five Personality Detection Using Handwriting Adelia Salsabila Arifin; Salamun Rohman Nudin
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12902

Abstract

Handwriting reflects a person's unique traits and has long been studied in the field of graphology to uncover personality characteristics. However, traditional graphological analysis is subjective, time-consuming, and prone to inter-rater differences. This study aims to develop a PenaKepribadian mobile application using ResNet50 transfer learning to automatically identify Big Five personality traits (Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism). The HiEnWrite dataset contains 327 English handwritten images annotated by a certified graphologists, was utilized with an 80:20 train-test split. Three optimizers such as SGD, RMSprop, and Adam were comparatively evaluated. Adam achieved the best performance with a training PCC of 0.5872 with 91.90% accuracy and a testing PCC of 0.4719 with 89.95% accuracy, outperforming both SGD and RMSprop. A testing PCC of 0.4719 indicates moderate correlation, suggesting promising yet improvable results. Robustness testing across varying lighting conditions, paper backgrounds, and writing media showed consistent performance, with mean prediction deviations ranging from 0.070 to 0.126. All Black Box Testing scenarios returned valid results. These findings confirm that ResNet50 transfer learning effectively extracts handwriting features for personality prediction, though further improvements remain necessary before high-stakes deployment. This research contributes to personality computing and opens avenues for efficient, automated, and accessible personality assessment systems.
Comparison of ARIMA, LSTM and GRU Models for Seismic b-Value Prediction in Southern Sumatera Rendinis Rendinis; Makhsun Makhsun; Choirul Basir
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12910

Abstract

Indonesia experiences high seismic activity due to its location at the convergence of three major tectonic plates, making continuous monitoring of earthquake potential crucial. A fundamental parameter in seismic hazard analysis is the b-value, which reflects the stress conditions, structural heterogeneity, and magnitude distribution within the Earth’s crust. Predicting b-value fluctuations remains challenging due to its highly volatile nature. This study aims to analyze and forecast the seismic b-value in Southern Sumatra by comparing a classical statistical model, ARIMA , with two advanced machine learning architectures, LSTM and GRU. Historical earthquake catalogs from BMKG and NEIC-USGS spanning 1960–2025 were utilized. The data underwent a declustering process using the Reasenberg method to eliminate foreshocks and aftershocks, yielding 15,844 independent events. The monthly b-value was then calculated using Maximum Likelihood Estimation. Furthermore, PSO was applied to tune the hyperparameters of the deep learning models. Evaluation reveals that ARIMA yields the highest predictive accuracy, achieving a Mean Absolute Error (MAE) of 0.11, Root Mean Square Error (RMSE) of 0.16, and Mean Absolute Percentage Error (MAPE) of 13.85%. In contrast, GRU (MAPE 14.30%) and LSTM (MAPE 16.22%) produced smoother predictions but struggled to capture extreme short-term fluctuations. The findings conclude that for volatile and limited-size time series data like regional b-values, the linear approach of ARIMA remains significantly more effective than complex deep learning models.
Content-Based Filtering Recommendation System for E-Commerce Products Using Sentence-BERT and Cosine Similarity Muhammad Abbad Raid Alfath; Arif Nur Rohman
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12920

Abstract

The rapid growth of e-commerce catalogs complicates product discovery, particularly for items with complex technical specifications like luxury watches. Conventional keyword searches and TF-IDF methods often fail to capture underlying semantic relationships. The primary objective of this study is to propose a robust e-commerce recommendation system utilizing Content-Based Filtering enhanced with Sentence-BERT (SBERT) semantic embeddings. This study employs an experimental comparative research design. The methodology involves aggregating product attributes into descriptive sentences and pre-processing them to minimize representation bias. These sentences are transformed into high-dimensional embeddings using the lightweight all-MiniLM-L6-v2 SBERT model, with similarities calculated via the Cosine Similarity algorithm. The system's performance is comparatively evaluated against a baseline TF-IDF method. Main outcomes and experimental results across 20 testing scenarios demonstrate that SBERT significantly outperformed the baseline, achieving an average Precision@5 of 93.00%, a Recall@5 of 2.71%, and a highly efficient latency of 0.38 ms. In conclusion, SBERT provides a superior, scalable solution for recommending complex products by accurately capturing the semantic similarity of technical specifications and textual representations of visual characteristics. The approach inherently resolves the item cold-start problem, and its successful integration into a web application confirms its feasibility for real-time similarity computation in modern e-commerce platforms.
Anti-Data Leakage Pipeline for Differentiated Thyroid Cancer Recurrence Prediction: Integrating SMOTE, Optuna-based Optimization, and Bootstrap BCa Validation Deri Rosadi; Sindhu Rakasiwi
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12922

Abstract

Thyroid cancer recurrence prediction remains a critical clinical challenge, as early identification of high-risk patients enables targeted monitoring and intervention. This study presents a comparative evaluation of six machine learning classifiers (XGBoost, LightGBM, CatBoost, Logistic Regression, Random Forest, and Decision Tree) using the UCI Differentiated Thyroid Cancer Recurrence dataset which consists of 383 patient records and 16 clinical features. To prevent performance overestimation, a rigorous anti-data leakage pipeline was implemented, encapsulating SMOTE, Optuna-based hyperparameter optimization, and Isotonic Calibration within the cross-validation process. Furthermore, model stability was assessed using Bias-Corrected and accelerated (BCa) Bootstrap validation with 2,000 iterations. Experimental results demonstrate that XGBoost achieved the best overall performance with an F1-score of 0.9545, an AUC-ROC of 0.9967, and the lowest Brier Score of 0.0183. Bootstrap BCa analysis confirmed XGBoost as the most stable model, with a 95% CI F1-score width of 0.1429 and unbiased estimation. These findings suggest that XGBoost, integrated within a zero-leakage pipeline and validated through Bootstrap BCa, is a promising candidate for post-treatment clinical decision support in differentiated thyroid cancer management.
Adaptive Retrieval-Augmented Generation with Domain Specific Fine Tuning For Smart MSME Digital Transformation Mawar Hardiyanti; Sri Hartati Wijono; Dwi Poetra Sedjati
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12924

Abstract

Micro, Small, and Medium Enterprises (MSMEs) play a significant role in the Indonesian economy; however, the adoption of digital technologies among MSMEs remains limited, reducing operational efficiency and business competitiveness. This study proposes an Adaptive Retrieval-Augmented Generation (ARAG) framework integrated with WhatsApp to support MSME digital transformation through contextual conversational AI assistance. The proposed system combines adaptive retrieval mechanisms with domain-specific fine-tuning using IndoBERT and a knowledge base containing 50,000 MSME operational documents. A mixed-methods approach was employed, consisting of system development, comparative evaluation, and field validation involving 200 MSMEs. Experimental results demonstrated that ARAG achieved an average response accuracy of 86.80%, outperforming rule-based, TF-IDF, and generic large language model baselines. The system also achieved a Retrieval Precision@5 of 0.874, an end-to-end F1-score of 0.841, and a lower hallucination rate compared to generic LLM approaches. Field validation showed a 22.7% improvement in operational efficiency, a 17.4% increase in digital adoption rates, and a System Usability Scale (SUS) score of 84.6, categorized as excellent usability. The findings indicate that retrieval grounding and domain adaptation contribute substantially to improving contextual relevance and practical usability in MSME-oriented conversational AI systems. Therefore, the proposed ARAG framework demonstrates strong potential as a practical and scalable digital assistance solution for supporting Indonesian MSME digital transformation.
Mapping the Persistent Danger Zones of Dengue Hemorrhagic Fever in Semarang City: A Spatio-Temporal Analysis Based on INLA Natanael Anggit Wicaksono; Amiq Fahmi
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12928

Abstract

This study maps the spatio-temporal risk dynamics of Dengue Hemorrhagic Fever (DHF) across 16 districts in Semarang City (2016–2025). Traditional epidemiological approaches using raw incidence rates often ignore spatial autocorrelation and struggle with overdispersion anomalies. To address this, we implemented a Hierarchical Bayesian framework using Integrated Nested Laplace Approximations (INLA) with a Negative Binomial distribution and a Besag-York-Mollié (BYM2) spatial architecture. We specified the spatial topology through a manually validated binary adjacency matrix to minimize subjectivity in defining regional boundaries. Our structured model improved computational performance significantly, reducing the Deviance Information Criterion (DIC) by 34.31% and the Root Mean Square Error (RMSE) by 15.30% compared to a baseline Poisson regression model. Using Geopandas and NetworkX for visualization, we identified Tembalang and Banyumanik districts as absolute Epicenter Nodes with an Exceedance Probability of 1.000. Spatial spillover network analysis demonstrated the propagation of epidemiological pressure from these epicenters to surrounding buffer zones, synchronized with the seasonal peak in the first quarter. This framework provides a precise computational foundation for vector control strategies, shifting from localized reactive approaches to preventive cluster mitigation.
Deep Learning-Based Rice Grain Classification with Class Imbalance Handling Using Weighted Sampling and Data Augmentation Muhammad Aulia Anhar; Nova Rijati
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12930

Abstract

Rice grain quality assessment is important for maintaining product consistency and quality standards in the food industry. Manual inspection is still widely used in practice, but it is often subjective and time-consuming especially in fine-grained classification problems where the visual differences between classes are relatively subtle. Another common challenge in rice grain datasets is class imbalance where the number of normal samples is much larger than the number of defective classes. In this paper, we investigate various imbalance handling strategies for deep learning-based rice grain classification using the GrainSet dataset, which contains 30,962 rice grain images divided into eight quality classes with a highly imbalanced class distribution. The proposed approach combines weighted sampling, data augmentation, and hyperparameter optimization to address class imbalance, while Focal Loss and Class-Balanced Loss are used as comparison methods for performance evaluation. Three deep learning architectures, namely MobileNetV3, ResNet50, and ViT-Small, were evaluated using Accuracy, Macro F1-score, and Weighted F1-score metrics. Experimental results show that the proposed approach achieved the most consistent overall performance. Among the tested models, ResNet50 achieved the best result with 98.87% accuracy and a Macro F1-Score 0.9799. The results also show that convolution-based architectures are more stable than transformer-based models for texture-oriented datasets with limited training data. Furthermore, the combination of sampling-based balancing, augmentation and suitable hyperparameter configuration contributed to better recognition performance on minority classes.
Identification and Mitigation of Web Application Vulnerabilities in Healthcare Systems Saeful Diyan Pratama; Aris Tri Joko Harjanto; Bambang Agus Herlambang
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12951

Abstract

The rapid adoption of web-based applications in healthcare systems has increased exposure to security threats, particularly at the application layer. Despite the implementation of various security mechanisms, many systems remain vulnerable due to improper input validation, weak authentication controls, and insecure database interactions. This study aims to identify, validate, and mitigate critical web application vulnerabilities in a healthcare system, focusing on nonce reuse vulnerabilities in token-based authentication mechanisms, stored cross-site scripting (XSS), and SQL injection. The research employs an empirical approach through controlled security testing, including vulnerability identification, exploitation validation, and mitigation evaluation. The results demonstrate that all identified vulnerabilities are actively exploitable, affecting authentication integrity, data confidentiality, and system reliability. Furthermore, the implementation of targeted mitigation strategies, such as token validation, input sanitization, and parameterized queries, substantially reduced the observed exploitability of the identified vulnerabilities within the tested scenarios. These findings highlight that application-layer security weaknesses remain a significant risk in healthcare systems and require systematic and integrated mitigation approaches. The study suggests that adopting secure-by-design principles and continuous security testing may improve system resilience against application-layer attacks. The implications of this research emphasize the need for proactive security practices in web-based healthcare applications to prevent exploitation and protect sensitive data from evolving cyber threats.
An Integrated Topic–Sentiment Analysis of User Reviews in Vidio Application Using BERTopic and IndoRoBERTa Nalendra Whisnu Pinilih; Ika Novita Dewi; Farrikh Alzami
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12952

Abstract

User reviews on digital platforms provide valuable insights into user experience; however, the large volume and unstructured nature of such data make systematic analysis challenging. In the case of the Vidio application, user feedback frequently reflects concerns related to advertisements, subscription systems, and technical performance. Nevertheless, existing researches often apply sentiment analysis and topic modeling separately, limiting the ability to understand how specific discussion themes are associated with user sentiment. To address this limitation, this research proposes an integrated topic–sentiment analysis approach for analyzing user reviews of the Vidio application collected from the Google Play Store. After filtering and quality control, 8,854 reviews were retained for further analysis using BERTopic for topic modeling and IndoRoBERTa for sentiment classification. The topic modeling process was optimized through parameter tuning, resulting in an improvement of the coherence score from 0.4076 to 0.6878, indicating better semantic consistency among the identified topics. Meanwhile, the sentiment classification model achieved an accuracy of 72%, although its performance was affected by class imbalance, particularly in identifying neutral sentiment. The analysis identified seven primary topics, where advertising-related issues emerged as the dominant topic and were strongly associated with negative sentiment, followed by concerns regarding subscription mechanisms and login accessibility. In contrast, content-related topics, particularly sports broadcasts, were consistently associated with positive sentiment. Furthermore, statistical evaluation confirmed a significant relationship between topic categories and sentiment distribution. Overall, the findings demonstrate that integrating topic modeling and sentiment analysis provides a more comprehensive understanding of user opinions and can support improvements in application quality and user experience.
A Sentiment Analysis of Free Nutritious Meal Program on Platform X: Comparing Naive Bayes, SVM, Random Forest, and IndoBERT Alrijal Nur Ilham; Etika Kartikadarma
Journal of Applied Informatics and Computing Vol. 10 No. 3 (2026): June 2026
Publisher : Politeknik Negeri Batam

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30871/jaic.v10i3.12959

Abstract

The Free Nutritious Meal Program (MBG), launched by the Indonesian government in January 2025, generated various public responses on social media, particularly on platform X. This study aims to analyze public sentiment toward the MBG Program and compare the performance of four sentiment classification methods: Naive Bayes, Support Vector Machine (SVM), Random Forest, and IndoBERT. The dataset was collected through tweet crawling using the keywords “MBG” and “Makan Bergizi Gratis” during the period of July–December 2025, resulting in 1,906 Indonesian-language tweets. The preprocessing stage included cleaning, case folding, normalization, tokenization, stopword removal, and stemming. Sentiment labeling was performed using the InSet Lexicon and produced 1,113 negative tweets and 793 positive tweets. Manual validation on part of the dataset was conducted by two independent annotators and achieved a Cohen’s Kappa score of 0.78, indicating substantial agreement. For classical machine learning models, feature extraction was carried out using TF-IDF, while IndoBERT used contextual text representations without stemming. Class imbalance in classical models was handled using SMOTE, whereas IndoBERT applied class weighting. The experimental results show that IndoBERT achieved the best performance with an accuracy of 92.93%. Among the classical models, SVM produced the highest performance with an accuracy of 92.15%, followed by Naive Bayes and Random Forest. Word frequency analysis also revealed that positive sentiment was mainly associated with support for the program and nutrition-related topics, while negative sentiment was dominated by concerns about food safety, budget management, and criticism of the program. Based on the findings, IndoBERT is more effective in understanding the context of Indonesian-language tweets. However, TF-IDF-based classical models, especially SVM, still provide competitive performance with lower computational requirements, making them suitable for sentiment analysis in public policy studies.

Page 89 of 101 | Total Record : 1006

 87   88   89   90   91 

Filter by Year

2017 2026


Reset
Filter By Issues
All Issue Vol. 10 No. 4 (2026): August 2026 Vol. 10 No. 3 (2026): June 2026 Vol. 10 No. 2 (2026): April 2026 Vol. 10 No. 1 (2026): February 2026 Vol. 9 No. 6 (2025): December 2025 Vol. 9 No. 5 (2025): October 2025 Vol. 9 No. 4 (2025): August 2025 Vol. 9 No. 3 (2025): June 2025 Vol. 9 No. 2 (2025): April 2025 Vol. 9 No. 1 (2025): February 2025 Vol. 8 No. 2 (2024): December 2024 Vol. 8 No. 1 (2024): July 2024 Vol. 7 No. 2 (2023): December 2023 Vol. 7 No. 1 (2023): July 2023 Vol. 6 No. 2 (2022): December 2022 Vol 6 No 1 (2022): July 2022 Vol 5 No 2 (2021): December 2021 Vol. 5 No. 2 (2021): December 2021 Vol 5 No 1 (2021): July 2021 Vol. 5 No. 1 (2021): July 2021 Vol. 4 No. 2 (2020): Desember 2020 Vol 4 No 2 (2020): Desember 2020 Vol 4 No 1 (2020): Juli 2020 Vol 3 No 2 (2019): Desember 2019 Vol 3 No 1 (2019): Juli 2019 Vol 2 No 2 (2018): Desember 2018 Vol. 2 No. 1 (2018): Juli 2018 Vol 2 No 1 (2018): Juli 2018 Vol 1 No 2 (2017): Desember 2017 Vol 1 No 1 (2017): Juli 2017 More Issue