Jurnal Online Informatika
Not a member yet
276 research outputs found
Sort by
Plant Disease Detection Using Digital Image Processing: Opportunities and Challenges
Full text linkDiseases in plants affect the yield of the plant itself. Agriculture is essential in human life, and if plant conditions are left unchecked, it will result in crop failure, which can affect the economy. Many researchers have developed methods to detect plant diseases, ranging from expert systems to deep learning algorithms. Machine learning is particularly effective for this task as it relies on datasets composed of plant images, making image processing crucial for the identification process. This article reviews the current literature and identifies several research gaps, opportunities, and challenges that must be addressed. Specifically, the article outlines potential avenues for future research in detecting plant diseases using image processing techniques. A significant opportunity exists to develop more effective algorithmic models for detecting plant diseases
Hybrid Squeeze-and-Excitation Convolutional Neural Network with Elastic Weight Consolidation for Longitudinal Learning in High-Accuracy Waste Classification
Full text linkWaste management has become a global issue. Increased urbanization and per capita consumption have caused unprecedented garbage growth. Sustainability has always been about proper waste management within the ecological framework. Recently, numerous studies have been conducted on automating the identification of waste items. In this study, a Convolutional Neural Network (CNN) model equipped with Squeeze and Excitation (SE) module is proposed based on hybrid squeezing methods for waste item classification. The core aim of this research is to improve the accuracy of classification by highlighting intricate relations between various features encoded within the dataset. Based on extensive tests on a waste dataset, the CNN model with the SE module using hybrid squeezing outperforms all other models. The suggested method\u27s 99.63% accuracy proves its efficacy and robustness. Furthermore, we incorporate Elastic Weight Consolidation (EWC) to enable longitudinal learning, allowing the model to adapt to emerging waste types (e.g., e-waste, biodegradable materials) while retaining prior knowledge with minimal forgetting (<1%). Ablation studies validate the critical role of hybrid squeezing, showing a 1.5% accuracy drop when spatial-wise components are omitted. This revelation affects automated recycling, waste sorting, and intelligent waste management. The proposed technology\u27s accuracy shows its applicability and dependability, advancing sustainable waste management. By automating waste classification with unprecedented precision, the proposed framework can reduce landfill reliance, enhance recycling rates, and inform policy decisions for sustainable urban planning
Enhancing Weather Monitoring for Agriculture with Deep Learning: Anomaly Detection in East Java Using LSTM Autoencoder and OCSVM
Full text linkAgricultural productivity in East Java is under threat from unpredictable and harsh weather patterns, particularly rapid variations in sunlight length and rainfall intensity. These abnormalities can interrupt agricultural cycles, lower yields, and make farming communities more vulnerable to climatic calamities. However, current weather monitoring systems frequently fall short of detecting small anomalies in time series weather data that could serve as early warning signs of such disasters. This study seeks to close this gap by creating a robust anomaly detection methodology adapted to time-dependent weather variables important to agriculture. In this study, a hybrid model combining Long Short-Term Memory (LSTM) autoencoder and One-Class Support Vector Machine (OCSVM) is proposed. The LSTM autoencoder\u27s structure reconstructs time series data and signifies anomalies through reconstruction errors (MSE), while OCSVM validates these anomalies to reduce false positives. The model was applied to daily weather data from East Java spanning 2015–2024. The results showed that the model effectively detected 11 anomalies in sunlight duration and 7 in rainfall, with F1-scores of 0.71 and 0.82, respectively. Several of these anomalies corresponded to actual disaster events such as floods, landslides, and droughts. This research contributed to the field by demonstrating the effectiveness of combining deep learning and machine learning for weather anomaly detection. The proposed framework offers valuable insights for early warning systems and can support local governments and farmers in improving disaster preparedness and enhancing agricultural resilience in East Java
Land Cover Classification in Mountainous Regions Using Multi-Scale Fusion and Convolutional Neural Networks: A Case Study on Mount Slamet
Full text linkMount Slamet, located in Central Java, Indonesia, is a high-risk volcanic region where accurate land cover classification is essential for disaster mitigation and sustainable land management. However, satellite imagery in this area often suffers from haze and cloud cover, posing challenges to reliable classification. This study aims to develop an effective land cover classification model using Sentinel-2 imagery by addressing these visual distortions. The specific goal is to classify land cover into five classes—Forest, Settlements, Summit, RiceField, and River—using enhanced satellite images. A total of 1101 labeled images were processed through dehazing with Multi-Scale Fusion (MSF) and smoothing using a Guided Filter to improve image quality. The classification was performed using three Convolutional Neural Network (CNN) architectures: VGG-16, MobileNetV2, and DenseNet121. The main contribution of this study is the integration of a tailored preprocessing pipeline with CNN-based modeling for haze-affected mountainous satellite imagery. Among the models tested, MobileNetV2 achieved the highest accuracy of 85.4%, outperforming DenseNet121 (83.8%) and VGG-16 (82.3%). The results demonstrate the effectiveness of combining image enhancement techniques with lightweight CNN architectures for land cover classification in challenging environments with limited and imbalanced dataset
The Application of AI Technology in Vocational High School Curriculum Design Based on Individual Student Skills in Facing the Challenges of the 21st Century Industry
Full text linkVocational high schools (SMK) are confronted with the challenge of adapting their curricula to align with the demands of industrial development in the 21st century. An inappropriate curriculum may result in students being inadequately prepared to navigate the demands of the professional world. Consequently, the objective of this research is to optimize the SMK curriculum through the utilization of an AI-based system, thereby enabling students the curriculum to be tailored to the specific skill requirements of individual. The methodology employed is Design-Based Research (DBR), which entails the analysis of student skill data, the design of an adaptive curriculum, and the evaluation of said curriculum through trials. The research comprised several phases, beginning with data collection and student skills analysis and concluding with an evaluation of student satisfaction with the implemented curriculum. The findings indicated that the introduction of an AI-assisted personalized curriculum resulted in an average improvement of 15% in students\u27 practical skills over a six-month period. Furthermore, student satisfaction with the implemented curriculum increased by 25%, from 70% at the outset of implementation to 95% following the introduction of the AI-based system. This research can serve as a reference point for the development of more adaptive and responsive SMK curricula
Modified Hash to Obtain Random Subset-Tree (MHORST) Using Merkle Tree and Mersenne Twister
Full text linkThe development of quantum computing triggers new challenges in data security, particularly in addressing attacks that can solve complex mathematical problems on the fly. Several hash-based data security methods have been proposed to deal with this threat, one of them being Hash to Obtain Random Subset-Tree (HORST). However, HORST has drawbacks, such as low security, because it only uses one hash round. The security of HORST is already improved by Hash to Obtain Random Subset and Integer Composition (HORSIC). However, HORSIC’s execution time is significantly increased. The problem of this research is the low-security HORST and the high execution time of HORSIC. This research proposes a new method, Modified Hash to Obtain Random Subset-Tree (MHORST), which aims to improve the security of HORST and reduce the execution time to less than HORSI’s. MHORST uses Merkle tree, SHA-256 hashes, and Mersenne Twister to build public keys and digital signatures. Based on the experiment results, MHORST reduces the signing time by more than 3.3 times compared to HORST. MHORST reduces the verification time by more than 1.1 times HORST and 17 times HORSIC. Although the security level of MHORST decreases slightly compared to HORSIC, this method is still more secure than HORST against signature forgery
K-Means-Based Pseudo-Labeling Technique in Supervised Learning Models for Regional Classification Based on Types of Non-Communicable Diseases
Full text linkNon-Communicable Diseases (NCDs) pose a critical threat to global public health, with Indonesia experiencing significant challenges due to high mortality rates and uneven regional distribution. In Banten Province, limited access to labeled health data hampers effective, data-driven intervention strategies. This study proposes a semi-supervised learning approach to develop a regional classification model for NCDs. The methodology begins with K-Means clustering applied to data from 254 community health centers (Puskesmas) to generate pseudo-labels. Various cluster configurations (k=2 to 8) were evaluated, with the optimal result being two clusters based on a silhouette score of 0.735. These clusters were then used to create a semi-labeled dataset for supervised learning. Eight classification algorithms—CN2 Rule Inducer, k-Nearest Neighbor (kNN), Logistic Regression, Naïve Bayes, Neural Network, Random Forest, Support Vector Machine (SVM), and Decision Tree—were trained and compared. Among them, the Neural Network model achieved the highest performance, with an AUC of 0.999 and an MCC of 0.976, indicating excellent stability and predictive accuracy. The findings validate the effectiveness of semi-supervised learning for health classification tasks when labeled data is scarce. This approach can serve as a valuable decision-support tool for regional health planning and targeted interventions, enhancing the precision and efficiency of public health responses
Reviewing the Framework of Blockchain in Fake News Detection
Full text linkIn the social media environment, fake news is a significant issue. It might be online or offline, depending on the field of journalism. Concerns have been expressed by media and publishing houses, who are looking for solutions to the problem. One of the solutions the industry has to offer in this area is Blockchain. It could be digital security trading, source or identity verification, or quotes following a certain news piece, photo, or video. It\u27s miles of shared document generation to deliver timely files, and it\u27s done with the help of a specific article, video, or image that has been addressed. This will no longer assist the fact abuser in verifying the details. This will help the fact abuser confirm the details, but it will also offer documentation of metadata generated at all phases. It allows you to cut the expense of disseminating false information by forwarding and explicit disclosure to persons who have first-hand knowledge of the subject. The proposed structure for acquiring fake news is supported by the blockchain age, which allows news organizations to deliver their content to their subscribers transparently. This framework was created for journalists and can be integrated into any current platform to publish a news piece and include asset statistics
Adoption of Artificial Intelligence and Digital Resources among Academicians of Islamic Higher Education Institutions in Indonesia
Full text linkThis study aimed to assess the readiness, attitudes, knowledge, and skills of lecturers in using artificial intelligence (AI) and electronic resources (ER) to enhance academic capacity. Understanding this adoption level is crucial for effectively integrating AI and ER into educational practices. In addition, this study contributes both theoretically and practically to digital scholarship by enhancing digital adoption and competence in education. This mixed-method study captured individual experiences and statistical trends related to digital scholarship in higher education. The qualitative method includes interviews, while the quantitative method involves survey questionnaires. The study focuses on lecturers from Islamic higher education institutions (IHEIs) in Indonesia. The results indicate that while lecturers rarely use AI and ES, they recognize the potential of digital technology in academic tasks. Despite limited exposure to AI and ER, IHEI lecturers in Indonesia can define these technologies accurately. Most lecturers actively update their knowledge and consider bias and ethical aspects in AI and ES usage. Regarding skills, over 60% of respondents reported proficiency in using AI and ES, suggesting a growing level of digital competence. These findings suggest that while many IHEI lecturers in Indonesia are prepared to adopt AI and ER, further support may be needed to ensure widespread acceptance
4 Levels of IoT Architecture for Smart Irrigation Rice Fields
Full text linkWater is one of the main components in the agricultural sector. Traditional irrigation systems are often inefficient and ineffective, which can lead to water wastage and require huge resources. Intelligent irrigation systems based on the Internet of Things (IoT) offer a solution to overcome these problems. The purpose of this research is to create a 4 Layer IoT architecture for smart irrigation in Gadon village. The method used in this research uses research and development methods, starting from literature study, field survey, design, assembly, and testing. Design of Internet of Things (IoT) architecture using ESP8266 for irrigation of rice fields in Gadon village Dlingo, Bantul. The design of this system aims to facilitate irrigation. This system utilizes IoT technology in its implementation. This system consists of four IoT layers, namely the Smart Things layer which consists of a water level sensor, water ph sensor, with control using ESP8266. Networks and Gateways layer, which consists of a router to connect smart things with the internet, Middleware layer, and Application layer which consists of an android application for the system interface. This system contributes directly to the form of convenience for farmers to manage irrigation of rice fields using ESP8266-based IoT applications. In addition, this system also provides water level information to facilitate farmers in the irrigation process