Publication Search

This study aims to develop a sentiment classification method for YouTube user comments related to the game Love and Deepspace using the Naïve Bayes algorithm, focusing on improving the text data processing and understanding user perceptions. Comment data were collected through scraping from YouTube videos, followed by preprocessing including text cleaning, normalization, stopword removal, stemming, and translation into English. Initial labeling was conducted using TextBlob, then the data were randomly sampled for training the Naïve Bayes model. Evaluation involved comparing sentiment distributions and visualization using Word Cloud and bar charts. The Naïve Bayes model achieved an accuracy of 77.36% in sentiment classification. The sentiment distribution shows differences between TextBlob (positive: 1,011, neutral: 1,312, negative: 575) and Naïve Bayes (positive: 901, neutral: 1,627, negative: 370), with Naïve Bayes being more conservative. The Word Cloud visualization identifies dominant words such as "bang," "game," and "main," while the bar chart shows the largest proportion of neutral sentiment. Naïve Bayes is effective for sentiment classification on informal comment data, with significant differences from rule-based methods like TextBlob. This research contributes to the development of text data processing techniques and user perception analysis, as well as opening up optimization opportunities with other algorithms like SVM for better accuracy.

Social media has evolved into a significant platform where consumers freely express their opinions, experiences, and levels of satisfaction regarding various products, including those offered by Micro, Small, and Medium Enterprises (MSMEs). The comments and reviews shared by customers on these platforms contain diverse sentiments that can serve as valuable indicators of how consumers perceive product quality. Understanding these sentiments is crucial for MSME owners, as it allows them to evaluate their products and adapt to market expectations more effectively. This study aims to analyze customer sentiment toward MSME products on social media by utilizing the Naïve Bayes algorithm, a widely used classification method in text mining. The data used in this research consist of customer comments collected from various social media platforms. The research process involves several stages, including data collection, manual labeling of sentiments, text preprocessing (such as tokenization, case folding, and stopword removal), and splitting the dataset into training and testing subsets. Subsequently, the classification process is carried out using the Naïve Bayes algorithm to categorize sentiments into positive, negative, and neutral classes. The results of this study demonstrate that the Naïve Bayes method is effective in classifying customer sentiments with a satisfactory level of accuracy. These findings provide a comprehensive overview of consumer perceptions regarding the quality of MSME products. Furthermore, this research is expected to assist MSME business owners in understanding customer feedback more systematically and using it as a basis for improving product quality and enhancing customer satisfaction in a competitive digital marketplace.

Islamic religious education at the Madrasah Diniyah level often relies on passive, one-directional teaching methods that reduce madrasah student engagement and long-term retention. This community service study aimed to improve the quality of Fiqh instruction at Madrasah Diniyah Al-Ishlah Kalirejo, Pasuruan, by implementing kinesthetic games, Small Group Discussion (SGD), and visual teaching aids within the framework of Kitab Mabadi’ Fiqhiyyah. A qualitative descriptive approach was employed through four participatory observation sessions, with visual documentation serving as primary data. The program was implemented in four thematic meetings covering funeral prayer procedures, tayammum, hajj simulation (tawaf), and najis classification. Findings indicate that kinesthetic games effectively reduced affective barriers and increased student focus, while the SGD model with a 1:8 mentor-to-student ratio enabled precise procedural correction for motor-based worship practices. The use of concrete teaching aids successfully transformed abstract Fiqh concepts into tangible, memorable knowledge. This study concludes that the integration of active learning methods, small-group mentoring, and visual media within traditional Islamic education settings can significantly enhance student engagement and comprehension. These findings offer a replicable pedagogical model for Madrasah Diniyah educators seeking to modernize instruction while preserving classical curriculum integrity.

Artificial intelligence has the potential to support radiology workflows by assisting in the identification of cases that may require additional clinical attention. However, alert-oriented medical AI systems should provide not only classification outputs but also interpretable evidence that can be reviewed and audited by clinicians. This study develops and evaluates an explainable multimodal framework for binary chest X-ray alert classification using paired radiology reports and chest X-ray images. The text branch employs TF-IDF n-gram features with a class-balanced Logistic Regression classifier, while the image branch fine-tunes a pretrained ResNet18 model. The two branches are integrated through probability-level late fusion using a validation-selected fusion weight. Explainability is implemented in a modality-specific manner: global coefficient analysis is used to identify influential textual cues, while Grad-CAM heatmaps are used to visualize salient image regions. Experiments were conducted on paired samples from the Open-i/IU X-Ray dataset using text-only, image-only, and fusion-based evaluation settings. Additional analyses include case-level complementarity analysis, bootstrap confidence intervals for ROC-AUC, shortcut-feature inspection, and qualitative Grad-CAM auditing. The results indicate that the text modality provides the dominant predictive signal under the current proxy-label setting. Late fusion produced a small descriptive improvement on the test set, increasing accuracy from 0.8533 to 0.8667, F1-score from 0.8817 to 0.8936, and ROC-AUC from 0.8936 to 0.9025 compared with the text-only baseline. However, the observed ROC-AUC improvement was not statistically conclusive based on bootstrap analysis. These findings suggest that the proposed framework is useful as a reproducible and auditable multimodal prototype, while also highlighting important limitations, including proxy-label ambiguity, potential label leakage from radiology reports, limited image-branch contribution, lack of external validation, and the need for stronger explanation and calibration assessment.

Penelitian ini bertujuan untuk menganalisis dan membandingkan kinerja model YOLOv11 dan MobileNetV3 dalam mengklasifikasikan varietas padi berdasarkan gambar digital. Kumpulan data terdiri dari 10.000 gambar yang mewakili lima varietas padi, yaitu Arborio, Basmati, Ipsala, Jasmine, dan Karacadag. Kumpulan data tersebut dibagi menjadi set pelatihan dan set pengujian, dengan proses pelatihan dilakukan selama 100 epoch. Hasil evaluasi menunjukkan bahwa MobileNetV3 mencapai akurasi klasifikasi sebesar 100%, sedangkan YOLOv11 memperoleh akurasi sebesar 99,8%. Meskipun akurasinya sedikit lebih rendah, YOLOv11 menunjukkan kinerja yang stabil dengan kesalahan klasifikasi yang minimal. Analisis matriks kebingungan menunjukkan bahwa sebagian besar prediksi masuk ke dalam kelas yang benar, dengan hanya sedikit kesalahan yang terjadi pada kategori yang secara visual mirip. Temuan ini menunjukkan bahwa kedua model tersebut sangat efektif untuk tugas klasifikasi varietas padi. Namun, evaluasi lebih lanjut menggunakan dataset yang lebih kompleks diperlukan untuk memastikan ketahanan dan generalisasi model dalam skenario dunia nyata.

The increasing use of mobile applications in the retail industry has generated a large volume of user reviews that contain valuable insights regarding customer experience and service quality. However, the unstructured nature of these reviews requires an automated approach to extract meaningful patterns efficiently. This study aims to perform sentiment analysis on user reviews of the Indomaret Poinku application by integrating lexicon-based labeling with machine learning classification. A total of 10,000 reviews were collected from Google Play Store and processed through a series of text preprocessing steps, including cleaning, case folding, normalization, tokenization, stopword removal, and stemming. Sentiment labeling was performed using the Indonesian Sentiment Lexicon (InSet), producing three sentiment classes: positive, negative, and neutral. The labeled data were vectorized using CountVectorizer and classified using two algorithms: K-Nearest Neighbors (KNN) and Random Forest (RF). Evaluation results show that Random Forest outperforms KNN, achieving an accuracy of 82.5%, compared to 69% for KNN. Random Forest demonstrates superior performance in handling high-dimensional sparse text features and yields more stable predictions across sentiment classes. This study contributes to the growing body of research on Indonesian sentiment analysis by demonstrating the effectiveness of combining lexicon-based labeling with ensemble learning methods, offering practical implications for developers seeking to improve the quality and user satisfaction of digital retail applications.

The Qur'an al-Karīm employs at least four principal terms in referring to human beings, namely al-basyar, al-insān, al-nās, and banī Ādam. These terms are not merely synonymous; rather, each represents distinct yet complementary dimensions of humanity in constructing a holistic concept of the human being. This study aims to: (1) analyze the semantic meanings of these four terms based on mufrodat studies, Makkiyah-Madaniyah classification, and asbābun nuzūl; (2) compare the interpretations of classical scholars - Al-Ṭabarī, Ibn Kathīr, Al-Qurṭubī, and Fakhr Al-Rāzī - with those of contemporary scholars - Sayyid Quṭb, Ibn ‘Āshūr, M. Quraish Shihab, and Buya Hamka; and (3) formulate their implications for Islamic education. This research employs a library research method using the tafsīr maudhū‘ī approach integrated with Izutsu’s semantic analysis model. The findings reveal that al-basyar represents the physical-biological dimension of human beings; al-insān represents the spiritual dimension in relation to ‘ubūdiyyah toward Allah; al-nās represents the social-collective dimension; and banī Ādam represents the intellectual-rational dimension inherited from Adam through the divine gift of teaching al-asmā’ (Qur'an 2:31). Collectively, these four dimensions provide fundamental implications for the development of objectives, curriculum, methodology, and evaluation within holistic and comprehensive Islamic education.

This study analyzes the obstacles in fulfilling the proof of deposit for Final Income Tax (PPh Final) as a requirement for land certificate name transfer at the Land Office of Bandar Lampung City. These obstacles impact legal uncertainty regarding land ownership and slow down public services. The research employs a descriptive analytical method with a qualitative approach. Data were collected through in-depth interviews with counter officers, verifiers, and the Land Rights Determination and Registration section, as well as direct observation and documentation at the local office. The findings identify three main classifications of obstacles. First, technical obstacles include payment data not being readable in the Land Office Computerization System (KKP) and barcodes on tax clearance certificates being unreadable. Second, administrative obstacles encompass data mismatches (nominal amounts, identities, Tax Object Numbers) and incomplete documents. Third, document-related obstacles include poor-quality photocopies and discrepancies in tax dates or years. Contributing factors are data entry errors (human error), lack of thoroughness, integration issues between the KKP system and the tax system, as well as taxpayers' low technical understanding. The consequences include process delays, re-verification with the tax office, and document returns. This study concludes that system-related obstacles, particularly the lack of optimal integration, are dominant. Recommended improvements include integrating the KKP system with the Directorate General of Taxes, enhancing automatic validation, simplifying validation procedures at the tax office, and increasing tax literacy among taxpayers.

The management of medicines and Medical Consumables (MCs) plays a crucial role in improving the quality of pharmaceutical services in hospitals. Common problems include stock imbalances, such as shortages and overstocking, which negatively affect service efficiency, increase operational costs, and raise the risk of product expiration. These issues are often caused by inadequate planning that is not based on consumption data, as well as limited knowledge of pharmacy personnel in applying appropriate inventory control methods.This community service activity aimed to improve the knowledge and skills of pharmacy staff in managing inventory using fast moving and slow moving methods based on real hospital data. The implementation method consisted of education, training, and hands-on mentoring conducted in several stages. The activity began with the analysis of medicine and medical consumables usage data from January to December 2025 at Labuang Baji Regional General Hospital, followed by training on pharmaceutical logistics management, and continued with practical exercises on classification and inventory control. Evaluation was carried out through discussions, case studies, and observation of participants’ ability to apply the methods.The results showed that 50.7% of medicines and 52.4% of medical consumables were categorized as fast moving, while the remaining items were classified as slow moving. After the intervention, there was a significant improvement in participants’ understanding of inventory classification, stock turnover analysis, and data-based planning. Participants were also able to identify items at risk of stock-outs and overstocking, enabling more appropriate control measures.This activity had a positive impact on the efficiency of pharmaceutical inventory management, reduced the risk of stock-outs and overstocking, and supported the improvement of healthcare service quality. Therefore, the fast moving and slow moving methods can be considered effective and applicable approaches for data-driven pharmaceutical inventory management in hospital settings

Drug availability at community health centers is an important indicator of health service quality. Drug shortages or excesses may affect service effectiveness and budget efficiency. This study aimed to determine the level of drug availability at the South City Community Health Center and the Piloloda'a Community Health Center in 2024 based on compliance with the formulary, demand, receipt, and drug availability categories. This study used a descriptive analytical method with a cross-sectional approach. Data were obtained retrospectively from the 2024 Drug Use Report and Request Sheet (LPLPO). Data analysis was conducted by calculating the percentage of compliance with the formulary, demand, and receipt, and by determining drug availability levels using the Indonesian Ministry of Health (2010) formula and the classification of Carolien et al. (2017). The results showed that formulary compliance was 82% at the South City Community Health Center and 67% at the Piloloda'a Community Health Center, both below the 95% standard. Drug demand compliance scores were 151% and 199%, exceeding the 100–120% standard, while drug receipt compliance scores were 71% and 56%, below the 100% standard. Drug availability categories varied from adequate and insufficient to excess stock, with most drug items classified as insufficient stock. In conclusion, drug management at both community health centers has not been fully optimal. Improved coordination between community health centers and pharmaceutical facilities is needed to maintain stable drug availability and support service needs.

Background: Recurrent Aphthous Stomatitis (RAS) is a common disorder characterized by recurrent ulcers limited to the oral mucosa. The etiology of RAS itself is not yet known for certain, but it is suspected that there are several predisposing factors, including hormonal changes, trauma, malnutrition, stress. Purpose: This study aims to determine the description of RAS in patients at the Oral Medicine Department of Dental Hospital University of Jember based on classification, general condition, suspected predisposing factors, and management. Method: This research is a descriptive observational study with a research population of 722 data from the Department of Oral Medicine, Dental Hospital of University of Jember. The number of samples that met the researchers' criteria was 171. Results: RAS patients is more common in women, 64.91% . RAS occurs in 69.60% of patients aged 21-30 years. RAS were mostly found in patients who did not experience symptoms of systemic factors, in 91.22%. The type of RAS that often occurs is the minor type in 78.37%. 68.42% RAS patients had no suspected predisposing factors. Based on RAS management, pharmacological therapy is divided into two parts, topical pharmacological therapy 59.07% and supportive pharmacological therapy 40.93%. Communication, information and education service was done for all 171 RAS patients. Conclusion: Based on the research conducted, that minor RAS is more prevalent in female 21-30 age group, with absence of underlying diseases, and the most frequently therapeutic modality is topical agents.

Background: Recurrent Aphthous Stomatitis (RAS) is a common disorder characterized by recurrent ulcers limited to the oral mucosa. The etiology of RAS itself is not yet known for certain, but it is suspected that there are several predisposing factors, including hormonal changes, trauma, malnutrition, stress. Purpose: This study aims to determine the description of RAS in patients at the Oral Medicine Department of Dental Hospital University of Jember based on classification, general condition, suspected predisposing factors, and management. Method: This research is a descriptive observational study with a research population of 722 data from the Department of Oral Medicine, Dental Hospital of University of Jember. The number of samples that met the researchers' criteria was 171. Results: RAS patients is more common in women, 64.91% . RAS occurs in 69.60% of patients aged 21-30 years. RAS were mostly found in patients who did not experience symptoms of systemic factors, in 91.22%. The type of RAS that often occurs is the minor type in 78.37%. 68.42% RAS patients had no suspected predisposing factors. Based on RAS management, pharmacological therapy is divided into two parts, topical pharmacological therapy 59.07% and supportive pharmacological therapy 40.93%. Communication, information and education service was done for all 171 RAS patients. Conclusion: Based on the research conducted, that minor RAS is more prevalent in female 21-30 age group, with absence of underlying diseases, and the most frequently therapeutic modality is topical agents.

Suspicious urban sound recognition is a critical component of intelligent public safety and urban monitoring systems, enabling the automated identification of anomalous acoustic events such as gunshots, sirens, and other security-sensitive sounds. However, existing deep learning approaches often struggle to simultaneously capture long-range temporal dependencies and global contextual relationships, particularly under noisy and acoustically complex urban conditions. This limitation can reduce reliability in safety-critical scenarios where missed detections carry significant risk. To address these challenges, this study proposes a Multi-Branch Bidirectional Long Short-Term Memory (BiLSTM) framework with Multi-Head Self-Attention (MHSA) for enhanced sequential and contextual feature modeling. Mel-frequency cepstral coefficients (MFCCs) are extracted from a curated subset of the UrbanSound8K dataset, comprising five suspicious sound classes, and used as input to the proposed architecture. The multi-branch design enables complementary temporal representations, while the self-attention mechanism provides lightweight contextual weighting of BiLSTM outputs. Experimental results demonstrate that the proposed model achieves a test accuracy of 95.59%, outperforming conventional Dense and LSTM-based baseline models under identical experimental settings. An ablation study further confirms the contribution of multi-branch integration and attention-based enhancement to overall performance. Class-wise evaluation reveals consistently high recall across all sound categories, particularly for safety-critical classes such as gunshots and sirens. These findings indicate that the proposed framework provides robust and reliable performance, making it suitable for real-time smart city surveillance and public safety applications.

Communication is an essential human right, yet a significant communication gap persists between individuals with sensory disabilities, specifically the deaf and speech-impaired, and the general public. While many technological solutions have been proposed to translate sign language, existing models primarily rely on heavy deep learning architectures such as Convolutional Neural Networks (CNN) or Recurrent Neural Networks (RNN/LSTM). These models often demand high computational power, leading to latency and limiting real-time application on standard devices. This study proposes a lightweight, fast, and highly responsive sign language translation system specifically designed to recognize static alphabets (A-Z) and single-character air writing. The system utilizes MediaPipe for hand tracking, where feature extraction is intelligently processed by calculating the relative spatial coordinates of fingertips to the wrist, reducing dependency on raw camera coordinates. Classification is performed using a Support Vector Machine (SVM) with a Radial Basis Function (RBF) kernel, prioritizing computational efficiency without sacrificing accuracy. To enhance user experience, the system introduces three key novelties: smart relative feature extraction, an anti-duplication hold system with a 1-second timer to prevent input spamming, and a non-blocking multithreaded audio execution (Daemon Thread) utilizing Google Text-to-Speech (gTTS), ensuring the webcam feed remains fluid during audio playback. Additionally, an alternative air-writing mode is integrated, utilizing geometric heuristics and PyTesseract OCR to read single drawn letters in the air. The results indicate that the proposed system operates swiftly and efficiently, bridging the communication barrier with a hardware-friendly approach.

Sentiment analysis for Indonesian regional languages faces two persistent challenges: labeled training data is extremely limited for most regional varieties, and transformer models pre-trained on Bahasa Indonesia do not generalize reliably to languages with substantially different morphological structures. Prior work on the NusaX benchmark has primarily relied on direct fine-tuning, treating each regional language independently and without exploiting linguistic proximity between related languages as a transfer signal. This paper proposes Language-Similarity-Guided Transfer (LSGT), a sequential fine-tuning strategy that first adapts a pre-trained model to a pivot language selected using character trigram similarity, followed by fine-tuning on the target language. Four transformer models are evaluated across all 12 NusaX languages using the official train/validation/test splits: IndoBERT, NusaBERT, mBERT, and XLM-R. Performance is evaluated using four metrics: accuracy, macro F1, macro precision, and macro recall. Experimental results show that LSGT improves macro F1 in 44 of 48 model-language combinations, demonstrating that the fine-tuning strategy itself is a major factor in low-resource cross-lingual sentiment classification. XLM-R benefits most strongly from LSGT, achieving an average improvement of +0.137 macro F1 and a peak gain of +0.298 on Madurese. SHAP-based token attribution analysis further reveals that predictions rely heavily on named entities and domain-specific nouns rather than sentiment-bearing vocabulary, indicating a dataset-level bias inherited from the original SmSA corpus and propagated through the NusaX translation pipeline.

Indonesian culinary comments on social media platforms such as Instagram are characterized by informal spelling, regional language mixing, slang expressions, and emojis, posing substantial challenges for automated sentiment classification. While IndoBERT has demonstrated strong performance across Indonesian natural language processing tasks, the contribution of individual preprocessing components to fine-tuning performance on informal text remains underexplored, particularly in the culinary domain. This study addresses this gap by conducting a systematic preprocessing ablation study on IndoBERT-Base fine-tuning for Indonesian culinary sentiment classification, accompanied by a comparative evaluation against Naive Bayes with TF-IDF, SVM with TF-IDF, and BiLSTM as representative baselines. A dataset of 3,500 manually labeled Instagram culinary comments across three sentiment classes was used, with a stratified 80/10/10 split. Six preprocessing variants were evaluated under identical experimental conditions to isolate the contribution of each component. The results show that slang normalization is the most impactful single preprocessing step, yielding a macro F1-score gain of +0.0609 over the no-preprocessing baseline, while the full pipeline achieves an accuracy of 0.8800 and a macro F1-score of 0.8465. IndoBERT-Base with the full pipeline outperforms all baselines across all evaluation metrics. Per-class analysis reveals that the negative class achieves the lowest F1-score of 0.7600, with sarcastic expressions and Banjar regional vocabulary identified as primary sources of misclassification. These findings indicate that preprocessing decisions have a measurable and non-uniform effect on IndoBERT fine-tuning performance. In this study, slang normalization provides the most substantial individual contribution in bridging the vocabulary gap between informal user-generated text and the model’s pre-training distribution.

This study analyzes visual signs in the Charlie Chaplin animations Safari at The Park and The King in The Ring using Charles Sanders Peirce’s semiotic framework, focusing on icons, indexes, and symbols. Film and animation communicate meaning through visual elements such as gestures, facial expressions, movements, and character interactions, making them rich for semiotic analysis. Using a descriptive qualitative method, the research identified and categorized visual signs in both animations. Results show that icons, which resemble real-world objects, dominate by establishing story settings natural safari environments in one animation and competitive boxing arenas in the other. Indexes reveal cause-and-effect relationships, demonstrating how gestures, expressions, and actions convey danger, fatigue, or emotional shifts. Symbols convey conventional or cultural meanings, such as Charlie Chaplin’s bowler hat and cane representing his comedic identity, a championship belt symbolizing victory, or a rose indicating affection. While both animations use the same types of signs, the intensity and focus vary with the narrative context: Safari at The Park emphasizes situational and natural elements, whereas The King in The Ring highlights competition and emotional reactions. This study confirms that Peirce’s triadic model effectively explains how meaning is constructed in animation through dynamic visual communication.

Group-based learning in the classroom plays a vital role in enhancing social interaction, individual responsibility, as well as students' critical thinking and collaborative skills. However, its implementation often faces challenges such as the dominance of certain members, social loafing, low participation, and interpersonal conflicts that hinder group effectiveness. This study aims to comprehensively examine the dynamics of learning groups by integrating four key aspects: the concept of group dynamics based on the Tuckman model, the characteristics of effective groups in cooperative learning, group formation techniques, and conflict management strategies. The research utilizes a qualitative approach with a literature study method, analyzing 25 sources including nationally accredited journals, academic books, and theses published between 2020 and 2024. Data analysis was conducted through reduction, thematic classification, content analysis, and conceptual synthesis. The results indicate that effective group dynamics can be achieved through the Tuckman stages, the application of the five elements of cooperative learning, the selection of appropriate group formation techniques with risk mitigation, and the implementation of the Thomas-Kilmann conflict management styles.The scientific contribution of this research is the development of an integrative model based on these four aspects, which serves as a conceptual framework to strengthen collaborative learning practices in the classroom. Practical implications include the formation of ideal groups consisting of 4–5 students, the establishment of initial group contracts, the use of dual assessment rubrics (individual and group), and peer evaluation mechanisms to enhance accountability and reflection.

The death of a Madrasah Tsanawiyah (MTs) student allegedly linked to police action raises significant legal issues concerning the limits of the use of force and the construction of criminal liability. This study aims to reformulate the elements of assault resulting in death by integrating the objective element (actus reus) and the subjective element (mens rea) within the framework of the doctrines of dolus and culpa. It also seeks to develop a model of criminal liability analysis that is more transparent, accountable, and oriented toward the protection of a child’s right to life. This research employs a normative juridical method using statutory, conceptual, and case approaches, supported by a literature review of legislation, court decisions, and criminal law scholarship. Data are analyzed qualitatively through grammatical, systematic, and teleological interpretation. The findings indicate that proving the act and the resulting death alone is insufficient without clearly establishing the form of fault. The distinction between dolus eventualis and culpa lata constitutes a decisive factor in determining the classification of the offense and the degree of criminal liability. Ambiguity in identifying the spectrum of fault may lead to sentencing disparities and weaken the principle of geen straf zonder schuld (no punishment without fault). Therefore, this study proposes a reconstruction of the elements of the offense that places proof of mens rea at the center of assessing police accountability while ensuring the protection of the child’s right to life.

Accurate classification of rainfall intensity patterns is important for early warning systems, hydrometeorological risk assessment, and water resource management. Surface rain gauges have limited spatial coverage, so this study uses NOAA NEXRAD Level II radar data from the KTLX station in 2023. K-Means clustering was applied to identify rainfall intensity patterns from 30 randomly selected days, with scans stratified into four daily time intervals. Seven features were extracted from each radar sweep, including reflectivity statistics, convective and stratiform ratios, and rainfall coverage. The data were normalized and balanced before clustering. The optimal cluster count was determined through a combined evaluation of the Elbow Method, Silhouette Score, and Davies-Bouldin Index, yielding K=5 as the most representative configuration. Evaluation results demonstrated a Silhouette Score of 0.3871 and a Davies-Bouldin Index of 0.8599, indicating moderate cluster cohesion that reflects the inherent overlapping nature of rainfall intensity transitions in radar reflectivity data. The clusters represent rainfall regimes from non-precipitating conditions to intense convective events. These results support the use of K-Means for automated rainfall pattern recognition and flood forecasting applications.