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
Kota surabaya,
Jawa timur
INDONESIA
Journal of Information Systems Engineering and Business Intelligence
Published by Universitas Airlangga
ISSN : -     EISSN : -     DOI : -
Core Subject : Science,
Computer Science & IT
Jurnal ini menerima makalah ilmiah dengan fokus pada Rekayasa Sistem Informasi ( Information System Engineering) dan Sistem Bisnis Cerdas (Business Intelligence) Rekayasa Sistem Informasi ( Information System Engineering) adalah Pendekatan multidisiplin terhadap aktifitas yang berkaitan dengan pengembangan dan pengelolaan sistem informasi dalam pencapaian tujuan organisasi. ruang lingkup makalah ilmiah Information Systems Engineering meliputi (namun tidak terbatas): -Pengembangan, pengelolaan, serta pemanfaatan Sistem Informasi. -Tata Kelola Organisasi, -Enterprise Resource Planning, -Enterprise Architecture Planning, -Knowledge Management. Sistem Bisnis Cerdas (Business Intelligence) Mengkaji teknik untuk melakukan transformasi data mentah menjadi informasi yang berguna dalam pengambilan keputusan. mengidentifikasi peluang baru serta mengimplementasikan strategi bisnis berdasarkan informasi yang diolah dari data sehingga menciptakan keunggulan kompetitif. ruang lingkup makalah ilmiah Business Intelligence meliputi (namun tidak terbatas): -Data mining, -Text mining, -Data warehouse, -Online Analytical Processing, -Artificial Intelligence, -Decision Support System.
Arjuna Subject : -
Articles 301 Documents
 27   28   29   30   31 
Explainable AI-Driven Deep Learning Model for Transparent and Accurate Pneumonia Detection Using Chest X-Rays Cynthia Hayat; Florensa Rosani Br Purba; Eka Widrian Suradji
Journal of Information Systems Engineering and Business Intelligence Vol. 12 No. 3 (2026): October
Publisher : Universitas Airlangga

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20473/jisebi.12.3.566-579

Abstract

Background: Pneumonia remains a national health concern with high morbidity and mortality among vulnerable populations. Chest X-ray is the most widely used diagnostic modality due to its accessibility and low cost. However, its interpretation heavily relies on radiologist expertise and often exhibits substantial interobserver variability, potentially leading to inconsistent diagnoses. Although deep learning models have shown remarkable accuracy in medical image analysis, their adoption in clinical practice is limited by the lack of transparency. XAI has therefore emerged as a promising approach to enhance interpretability and clinical trust.  Objective: This study aims to develop a ResNet-50-based explainable deep learning framework for pneumonia detection in CXR by integrating XAI techniques in the form of Grad-CAM and LIME  Methods: A CNN model with a ResNet-50 backbone was trained and tested using a CXR dataset from clinical sources at partner hospitals in Indonesia. Grad-CAM enhances transparency by highlighting relevant lung areas, and LIME offers local feature based explanations by identifying the most important image regions for model decisions. Performance was evaluated using metrics such as accuracy, precision, recall, F1-score, IoU, Dice Score, mean Average Precision (mAP@0.5), and the agreement level with radiologists measured by Cohen’s Kappa test.  Results: The model achieved an accuracy of 93.2%, precision of 92.5%, recall of 94.1%, F1-score of 93.3%, IoU of 78.4%, Dice score of 87.9%, and mAP@0.5 of 89.7%, with an average inference time of 120 ms per image. Grad-CAM and SHAP visualizations showed up to 91% of the area matched the radiologist’s assessment, indicating nearly perfect agreement, with a Kappa value of 0.90.  Conclusion: Combining ResNet-50 with Grad-CAM creates a pneumonia detection model that is accurate, easy to understand, and matches doctors' opinions. This approach could help doctors trust artificial intelligence more and encourage wider use of technology in healthcare, especially in resource limited settings.  Keywords: Pneumonia detection; Chest X-ray; Deep learning; Explainable AI; ResNet-50; Grad-CAM; Medical image analysis 
Unveiling Trends in Natural Language Processing Research for Public Policy: A Bibliometric Analysis and Dynamic Topic Modeling Approach Shiddiq Sugiono; Adhityo Nugraha Barsei
Journal of Information Systems Engineering and Business Intelligence Vol. 12 No. 3 (2026): October
Publisher : Universitas Airlangga

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20473/jisebi.12.3.726-743

Abstract

Background: Natural Language Processing (NLP), a subfield of Artificial Intelligence (AI), has advanced rapidly and can now analyze large volumes of textual data to extract meaningful information. This is highly relevant for supporting evidence-based, responsive, and accountable public policy through efficient analysis of public opinion, policy documents, and academic literature.  Objective: This study maps research trends in the application of NLP to public policy through a bibliometric analysis and Dynamic Topic Modeling (DTM) approach, identifying the current topic structure, the temporal evolution of research themes, and the topics that may inform future research agendas.  Methods: Following the PRISMA 2020 protocol, publications from 2016–2025 were retrieved from the Scopus and Web of Science databases and screened through a multi-stage selection process. A total of 1,134 publications were included in the final bibliometric analysis. The resulting abstracts were analyzed using BERTopic-based DTM. The best-performing hyperparameter configuration, chosen after comparing several candidates, achieved coherence and topic diversity scores, as well as an outlier ratio, within generally acceptable ranges, yielding a set of primary topics plus a smaller outlier topic.  Results: Publications grew from fewer than 70 per year before 2021 to 377 in 2025, led by China (317), the United States (181), and India (158). AI Governance & Public Services was the largest topic (27.69% of documents) but showed a slowing topic-probability trend, while Green Energy & Low-Carbon Transition, Water Resource Governance, and Climate Policy & Environmental Awareness emerged as the clearest hot topics. COVID-19-related topics declined after 2022, following a marked increase in NLP research on public health during the pandemic.  Conclusion: NLP research in public policy is shifting from governance- and pandemic-related themes toward AI-driven public services and socio-environmental sustainability. Future public health policy research should prioritize post-COVID-19 recovery and healthcare resilience, while sustainability-linked topics are expected to continue expanding.  Keywords: Natural Language Processing, Dynamic Topic Modeling, Public Policy, Government, BERTopic 
Particle Swarm Optimization-Based CatBoost for Livestock Biogas Prediction and Feature Importance Analysis Syafrial Fachri Pane; Ahmad Rifki Ayala
Journal of Information Systems Engineering and Business Intelligence Vol. 12 No. 3 (2026): October
Publisher : Universitas Airlangga

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20473/jisebi.12.3.688-703

Abstract

Background: Accurate estimation of livestock-based biogas potential is essential for renewable energy planning in agricultural regions with heterogeneous livestock populations and varying microclimatic conditions. Existing studies commonly rely on facility-scale or aggregated datasets with limited spatial resolution, reducing their applicability for regional assessment.   Objective: This study aims to develop an accurate, interpretable machine learning framework to predict daily livestock biogas potential at the sub-district level by integrating livestock population, feed management, microclimate variables, and to improve predictive performance through hyperparameter optimization.   Methods: A quantitative regression approach was applied using sub-district-level data from West Java, Indonesia. The dataset consisted of 450 validated observations derived from livestock statistics, biogas factors based on the Intergovernmental Panel on Climate Change-based, feed usage indicators, and spatially mapped temperature data. The preprocessing included missing value imputation, outlier handling, and min–max scaling. Eleven regression models were evaluated using K-Fold Cross-Validation. Particle Swarm Optimization (PSO) was used to optimize CatBoost hyperparameters, while interpretability was assessed using SHapley Additive exPlanations (SHAP), Mean Decrease in Impurity, and Permutation Importance.  Results: Boosting-based models demonstrated superior predictive performance compared with linear and distance-based methods. The PSO-optimized CatBoost model achieved the highest performance with a coefficient of determination (R²) value of 0.746, Root Mean Square Error (RMSE) of 0.162, and Mean Absolute Error (MAE) of 0.084, outperforming all other models across optimization strategies. The feature importance analysis consistently identified total livestock population, beef cattle population, and average district temperature as the most essential predictors, while feed-related variables moderately contributed to the model predictions.  Conclusion: Livestock scale and regional microclimate are key biogas potential determinants. The proposed PSO-optimized CatBoost framework provides an accurate and interpretable approach for estimating livestock-based biogas production, offering useful insights for regional renewable energy planning.  Keywords: CatBoost, Feature importance analysis, Livestock biogas prediction, Machine learning, Particle Swarm Optimization 
Integrating Prompt-Guided Pipeline of YOLOv8 and Segment Anything Model for Cervical Precancerous Lesion Segmentation Akhiar Wista Arum; Siti Nurmaini; Muhammad Naufal Rachmatullah; Bambang Tutuko; Firdaus Firdaus; Annisa Darmawahyuni; Ade Iriani Sapitri; Anggun Islami; Ghita Athalina; Muhammad Alpim Alfa Rolis
Journal of Information Systems Engineering and Business Intelligence Vol. 12 No. 3 (2026): October
Publisher : Universitas Airlangga

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20473/jisebi.12.3.629-644

Abstract

Background: Segmenting precancerous cervical lesions from cervicography images remains challenging due to irregular lesion morphology, blurred boundaries, low contrast, and visual similarity between abnormal and normal cervical tissue. Automated segmentation methods are needed to support more objective and efficient cervical cancer screening.  Objective: This study aimed to develop and evaluate a prompt-guided integration framework that integrates variants of YOLOv8 with the Segment Anything Model (SAM) for the detection and segmentation of precancerous cervical lesions.  Methods: A total of 1,350 cervicography images were used, consisting of 300 primary clinical images (TeleOTIVA Dataset) and 1,050 public images from the International Agency for Research on Cancer (IARC). The dataset was divided into 80% training data and 20% validation data. Five variants of YOLOv8—namely YOLOv8n, YOLOv8s, YOLOv8m, YOLOv8l, and YOLOv8x—were evaluated for detecting the cervical region, columnar region, and lesion classes. The best YOLOv8 model was then used to generate bounding box prompts for SAM-based segmentation using the ViT-B, ViT-L, and ViT-H encoders.  Results: YOLOv8m achieved the best overall detection performance, with an average mAP of 65.7%. In the segmentation stage, ViT-L SAM achieved the best performance. The optimal model configuration depends on the object being segmented. YOLOv8m–ViT-L SAM delivered the best performance in lesion segmentation, with an IoU of 0.591 and a Dice coefficient of 0.705. Statistical analysis shows that YOLOv8m–ViT-L SAM achieved the best average score for lesion segmentation and significantly outperformed the evaluated comparison models.   Conclusion: The proposed YOLOv8m-SAM pipeline demonstrates the feasibility of detection-guided segmentation using prompts generated by YOLOv8. However, more detailed lesion segmentation remains challenging and requires further improvement through prompt refinement, domain adaptation, and validation on larger, multicenter datasets.  Keywords: Cervicography, Precancerous Cervical Lesions, Segmentation, Segment Anything Model, YOLOv8 
Optimizing Fine-Tuning Memory Usage with QLoRA for Automated Pedagogical Feedback Generation in Bahasa Indonesia Martinus Maslim; Albert William Wibisono; Yulius Denny Prabowo
Journal of Information Systems Engineering and Business Intelligence Vol. 12 No. 3 (2026): October
Publisher : Universitas Airlangga

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20473/jisebi.12.3.580-595

Abstract

Background: While recent advances in open-source large language models (LLMs) have enabled automated feedback generation for educational assessment, most other research is focused on English and requires high-end Graphics Processing Unit (GPU) resources. However, there is limited empirical evidence of the efficiency of LLMs in Bahasa Indonesia feedback tasks in low-resource settings like Indonesia, given the strict hardware limitations and differences in data sources  Objective: This study investigates whether 4-bit quantization combined with Low-Rank Adaptation (LoRA)-based task adaptation can enable open-weight LLMs to generate Indonesian pedagogical feedback under strict 8GB Video Random Access Memory (VRAM) constraints, comparing the performance across various model architectures and supervision data (human-curated vs synthetic).   Methods: A quantitative and experimental design using a 2 × 2 factorial structure was used to compare Llama-3-8B-Instruct and Gemma-2-9B-Instruct on 2 datasets of equal size (1,950 human-curated and 1,950 synthetic feedback pairs). For model fine-tuning, we implemented 4-bit Quantized Low-Rank Adaptation (QLoRA) quantization on consumer-grade GPUs with 8GB VRAM Performance was assessed by semantic similarity using Sentence-BERT (SBERT), similarity of the model with BLEU-4, ROUGE-L, peak VRAM usage, and training time.  Results: QLoRA successfully enabled fine-tuning within 8GB VRAM, with peak memory usage between 7.0 and 7.8GB. Gemma-2-9B-Instruct achieved the highest semantic alignment (SBERT similarity = 0.77 on synthetic and 0.74 on human data), while Llama-3-8B-Instruct showed competitive performance (0.76 and 0.71, respectively). Human-curated datasets produced higher lexical alignment (ROUGE-1 up to 44.8) than synthetic datasets, though synthetic supervision remained semantically consistent.  Conclusion: QLoRA enables efficient fine-tuning of large language models on consumer-grade hardware, making Indonesian feedback generation feasible in low-resource settings. While 4-bit quantization reduces memory requirements, LoRA enables task-specific pedagogical adaptation. Instruction-tuned models consistently outperform language-adapted base models. Future work should include human evaluation and broader educational contexts to improve generalizability.  Keywords: Automated Feedback Generation, Large Language Models, QLoRA Fine-Tuning, Bahasa Indonesia, Educational Assessment 
Bridging Algorithms and Robust Systems through Adaptive Machine Learning in Electrooculography-Based Assistive Technology: A Systematic Literature Review Deddy Prayama; Muhammad Ilhamdi Rusydi; Budi Rahmadya; Minoru Sasaki
Journal of Information Systems Engineering and Business Intelligence Vol. 12 No. 3 (2026): October
Publisher : Universitas Airlangga

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20473/jisebi.12.3.791-809

Abstract

Background: Electrooculography (EOG) has emerged as a promising biosignal modality for assistive technology because it enables eye movement-based human–computer interaction for individuals with motor impairments. However, research on EOG-based assistive systems remains fragmented across signal processing, machine learning, adaptive learning, multimodal integration, and deployment-oriented implementation. Objective: This study aims to systematically review adaptive machine learning approaches in EOG-based assistive technology, identify current research trends and challenges, and develop a conceptual framework for bridging adaptive intelligence and real-world deployment. Methods: A Systematic Literature Review (SLR) was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guideline. Literature searches were performed in IEEE Xplore, Scopus, PubMed, Springer Nature, Sage, and ScienceDirect. After applying inclusion and exclusion criteria and quality assessment procedures, 60 primary studies published between 2015 and 2025 were selected for qualitative descriptive and thematic synthesis. Results: The findings show a transition from conventional signal processing, threshold-based recognition, and handcrafted feature extraction toward deep learning, transfer learning, domain adaptation, and multimodal adaptive systems. The review also identifies a persistent deployment gap, where high-performing models often require computational resources that may not be suitable for portable, low-cost, and real-time assistive devices. To address this issue, this study proposes the Adaptive EOG Deployment Framework (AEDF), which integrates signal adaptivity, model adaptivity, and deployment adaptivity. Conclusion: Adaptive machine learning is a key enabler of next-generation EOG-based assistive technology. The proposed AEDF provides a unified perspective for developing robust, generalizable, efficient, and user-centered assistive systems that can better support real-world deployment. Keywords: Electrooculography, Adaptive Machine Learning, Assistive Technology, Human Computer Interface, Deployment, Systematic Literature Review
Improving Indonesian Recipe Named Entity Recognition via Weighted Ensemble and Semi-Supervised Learning Dody Haryadi; Imas Sukaesih Sitanggang; Sony Hartono Wijaya; Budi Setiawan
Journal of Information Systems Engineering and Business Intelligence Vol. 12 No. 3 (2026): October
Publisher : Universitas Airlangga

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20473/jisebi.12.3.548-565

Abstract

Background: The rapid growth of online culinary platforms has produced a vast amount of unstructured Indonesian recipe texts. Extracting structured information such as ingredients, quantities, and measurement units002C is essential for downstream applications, including nutritional computation and diabetes-oriented dietary recommendation systems. However, Indonesian recipe texts present substantial challenges due to linguistic variability, informal writing styles, and limited availability of annotated corpora.  Objective: This study investigates effective modeling strategies for Indonesian recipe Named Entity Recognition (NER) in low-resource settings and to identify reliable approaches to improve the robustness and generalization of extraction.  Methods: A quantitative experimental approach was employed to evaluate the IndoBERT, BiLSTM-CRF, IndoBERT-CRF, and hybrid IndoBERT-BiLSTM-CRF architectures. A reliability-weighted voting ensemble combining heterogeneous models was proposed, and high-confidence ensemble predictions were used for pseudo-labeling in a semi-supervised setting. Performance was assessed using entity-level precision, recall, and F1-score.  Results: IndoBERT achieved a strong baseline performance (F1-score = 96.61%). IndoBERT-CRF and the hybrid IndoBERT-BiLSTM-CRF model achieved competitive F1-scores (92.99% and 92.82%, respectively) but did not outperform the Transformer-only baseline. The proposed weighted ensemble achieved the best supervised performance (F1-score = 96.79%). Semi-supervised training with ensemble-generated pseudo-labels maintained competitive performance (F1-score = 96.72%) while improving recall and generalization to unseen ingredient expressions.  Conclusion: Increased architectural complexity does not necessarily improve performance on the Indonesian recipe NER task. Reliability-weighted ensemble learning combined with pseudo-labeling provides a stable and effective approach for low-resource culinary NER and supports future nutrition-oriented recipe processing.  Keywords: Indonesian Recipe Text; Pseudo-Labeling; Named Entity Recognition; Semi-Supervised Learning; Weighted Ensemble 
Digital Ambidexterity in the Digital Transformation Era: A Systematic Literature Review of Dynamic Capabilities, Critical Enablers, and Sustainable Outcomes Endra Prasetya Rudiyanto; Munir Munir; Asep Wahyudin; Puspo Dewi Dirgantari
Journal of Information Systems Engineering and Business Intelligence Vol. 12 No. 3 (2026): October
Publisher : Universitas Airlangga

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20473/jisebi.12.3.704-725

Abstract

Background: Organizations undergoing digital transformation must simultaneously pursue innovation (exploration) and optimize existing capabilities (exploitation), a tension managed through digital ambidexterity. Although research on digital ambidexterity, dynamic capabilities, and digital transformation has grown rapidly since 2019, no prior systematic review has integrated the three constructs, leaving the information systems and strategic management literature fragmented.   Objective: This study maps the contexts and applications of digital ambidexterity, identifies how dynamic capabilities enable it, determines its critical enablers and performance outcomes, and synthesizes these findings into an integrated conceptual framework.   Methods: A systematic literature review following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 protocol was conducted using Scopus. Of 139 retrieved papers, title-abstract screening, full-text assessment, and an eight-item quality checklist (threshold ≥5/8) yielded 33 papers (mean score 7.35/8.0) published in 2019–2025. Bibliometric analysis of the full corpus used the bibliometrix package in R.  Results: Three principal findings emerged, corresponding to the first three objectives above. First, digital ambidexterity is most studied in manufacturing, financial services, and technology sectors, concentrated in China and Asia, with organizational size and digital maturity shaping the form adopted. Second, dynamic capabilities (sensing, seizing, and reconfiguring) evolved from structural approaches (2019–2021) toward ecosystem-level orchestration (2024–2025), with digital sensing operating recursively through real-time platform data. Third, organizational enablers, particularly digital leadership and innovation culture, act as preconditions, technological enablers amplify, and environmental enablers moderate ambidextrous outcomes, which bifurcate into short-term operational and long-term strategic results, with sustainability increasingly prominent.  Conclusion: Synthesizing these three findings, this study proposes the DA-DC-O (Digital Ambidexterity, Dynamic Capabilities, and Outcomes) Integrated Framework, the first synthesis unifying organizational context, critical enablers, dynamic capabilities mechanisms, and multi-tier performance outcomes, and offers digital leaders and Information Systems (IS) governance professionals a diagnostic taxonomy for prioritizing capability investments.  Keywords: Digital Ambidexterity, Dynamic Capabilities, Digital Transformation, Systematic Literature Review, Exploration-Exploitation, Sustainable Outcomes  
A Metadata-Based Framework for Automated Transformation of Digital Government Architecture Metadata into Enterprise Architecture Models Eva Hariyanti; Airlangga Dwi Satrio; Faried Effendy; Badrus Zaman; Taufik Taufik; Army Justitia; Shukor Sanim Mohd Fauzi; Vabela Dwi Seravin; Indra Kharisma Raharjana
Journal of Information Systems Engineering and Business Intelligence Vol. 12 No. 3 (2026): October
Publisher : Universitas Airlangga

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20473/jisebi.12.3.775-790

Abstract

Background: Enterprise architecture plays a critical role in aligning business processes, data, applications, and technology across government organizations. Metadata architecture provides the foundation for the development of enterprise architecture. It is generally maintained as a structured document, making it labor-intensive and difficult to maintain. Automating the transformation of such metadata represents an opportunity for process innovation in EAM. Objective: This study proposes a metadata-based framework for automatically converting metadata on digital government architecture into an enterprise architecture model. Methods: This framework defines the conceptual mapping between architectural metadata and ArchiMate elements based on the TOGAF architecture domain and applies rule-based transformations to generate enterprise architecture models. An automated prototype is developed and evaluated through expert review and quantitative performance assessment. Results: The proposed framework achieved a precision of 1.00, recall of 0.990, and F1-score of 0.995, indicating a high level of agreement between the generated and reference models. Expert evaluation indicated that the system is usable and that the resulting model represents the architecture well. However, diagram readability received the lowest rating among the evaluated aspects and was identified as an area for improvement. Conclusion: This research develops a conceptual model that maps metadata from each domain of a digital government architecture to ArchiMate elements based on the TOGAF domain structure. The evaluation of the system indicates that it is user-friendly, effectively represents architecture, and has the potential for widespread adoption within the Indonesian government sector. Future work should focus on generalizing the model for an international context and testing it within large-scale EA repositories. Keywords: Enterprise Architecture, Digital Government, Architecture Metadata, ArchiMate, Process Innovation, Model Transformation
Multiclass Classification and Spatial Clustering of Disaster Tweets for Humanitarian Awareness Mochammad Nafkhan Alzamzami; Nasa Zata Dina; Aji Akbar Firdaus; Yuita Arum Sari; Syamsul Bachri
Journal of Information Systems Engineering and Business Intelligence Vol. 12 No. 3 (2026): October
Publisher : Universitas Airlangga

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20473/jisebi.12.3.824-840

Abstract

Background: Social media is a critical channel of real-time information channel during natural disasters, but the informal and fragmented nature of tweets makes automatic classification difficult. Named entity recognition (NER) can capture contextual cues, such as affected locations. However, how entity information should be incorporated into classifiers and its value for disaster mapping remain underexplored. Objective: This study investigates whether NER can improve disaster-tweet classification and support clustering-based disaster mapping for humanitarian decision support. Methods: Eight configurations were evaluated using raw text: linear and BiLSTM classification heads combined with three input representations (Raw, Inline Tag, and Special Token), and two architecture-level injection models (BiLSTM+TagEmb and Two-Stream) using raw text. An ablation study compared eight pretrained language models across four architectures and input variants. In addition, six density-based clustering algorithms were applied to NER-extracted location entities to map the spatial distribution of disaster events. Results: The two-stream model achieved the highest performance (macro F1 = 0.947), followed by the BiLSTM classifier with raw input (0.914), indicating that the architecture-level entity integration outperforms the input-level entity marking. VDBSCAN, HDBSCAN, and OPTICS produced the most cohesive clusters (silhouette > 0.97). Conclusion: Integrating entity information at the architectural level is more effective than embedding entity markers in the input text. Density-based clustering is better suited to the exploratory discovery of spatial patterns than to reproducing predefined categories. These findings provide practical guidance for designing disaster-tweet classifiers and selecting unsupervised methods to analyze disaster-related social media data. Although the proposed approach may contribute to situational awareness, its utility for operational humanitarian decision-making remains to be empirically validated.   Keywords: Disaster tweet classification, Named entity recognition, natural language processing, social media analytics, disaster

Page 30 of 31 | Total Record : 301

 27   28   29   30   31 

Filter by Year

2015 2026


Reset
Filter By Issues
All Issue Vol. 12 No. 3 (2026): October Vol. 12 No. 2 (2026): June Vol. 12 No. 1 (2026): February Vol. 11 No. 3 (2025): October Vol. 11 No. 2 (2025): June Vol. 11 No. 1 (2025): February Vol. 10 No. 3 (2024): October Vol. 10 No. 2 (2024): June Vol. 10 No. 1 (2024): February Vol. 9 No. 2 (2023): October Vol. 9 No. 1 (2023): April Vol. 8 No. 2 (2022): October Vol. 8 No. 1 (2022): April Vol. 7 No. 2 (2021): October Vol. 7 No. 1 (2021): April Vol. 6 No. 2 (2020): October Vol. 6 No. 1 (2020): April Vol. 5 No. 2 (2019): October Vol. 5 No. 1 (2019): April Vol. 4 No. 2 (2018): October Vol. 4 No. 1 (2018): April Vol. 3 No. 2 (2017): October Vol. 3 No. 1 (2017): April Vol. 2 No. 2 (2016): October Vol 2, No 1 (2016): April Vol. 2 No. 1 (2016): April Vol 1, No 2 (2015): October Vol. 1 No. 2 (2015): October Vol. 1 No. 1 (2015): April Vol 1, No 1 (2015): April More Issue