Articles

Access the latest knowledge in applied science, electrical engineering, computer science and information technology, education, and health.

Advanced Search
Filter Icon

Filters article

Years

FAQ Arrow
Single Year
Range Year
0
0

Source Title

FAQ Arrow

Authors

FAQ Arrow

28,428 Article Results

Generative Indonesian chatbot for university major selection using transformers embedding

10.11591/ijai.v14.i4.pp3474-3482
Mutiara Auliya Khadija , Bambang Harjito , Morteza Saberi , Astrid Noviana Paradhita , Wahyu Nurharjadmo
Selecting a university major is a crucial decision that impacts students' future career paths and personal fulfillment. Traditional guidance methods often lack the personalization and timeliness needed to support students effectively. This study explores the use of Indonesian generative artificial intelligence (AI) chatbots and transformer embeddings to enhance decision-making for university major selection. By leveraging advanced AI techniques, such as bidirectional encoder representations from transformers (BERT) and Gemini embeddings, the research aims to provide personalized, interactive, and contextually relevant guidance. Experiments showed that BERT embeddings achieved the highest accuracy, with recurrent neural network (RNN) and long short-term memory (LSTM) models also performing well but facing issues with overfitting. Gemini embeddings provided strong performance but slightly less effective than BERT. The findings suggest that BERT-based models with RNN are superior for developing decision-support systems in 92% accuracy. Future work should focus on further optimization and integration of user feedback to ensure the relevance and effectiveness of these AI tools in educational settings.
Volume: 14
Issue: 4
Page: 3474-3482
Publish at: 2025-08-01

Optimizing traffic lights at unbalanced intersections using deep reinforcement learning

10.11591/ijai.v14.i4.pp2991-3002
Duman Care Khrisne , Made Sudarma , Ida Ayu Dwi Giriantari , Dewa Made Wiharta
Unbalanced intersectional traffic flow increases vehicle delays, fuel consumption, and pollution. This study investigates the application of deep reinforcement learning (DRL) to optimize traffic signal timing at the Pamelisan intersection in Denpasar, Indonesia. Real-world traffic data were incorporated into a SUMO microsimulation environment to train DRL agents using the deep Q-network (DQN) algorithm. Experimental results show that DRL-based optimization reduced the average vehicle waiting time from 594.49 seconds (static control) to 169.44 seconds and 173.10 seconds for agents trained without and with noise, respectively. The average vehicle speed remained stable at 5.6–5.97 m/s across all scenarios, indicating enhanced traffic efficiency without adverse effects. The findings underscore the effectiveness and adaptability of DRL in addressing traffic inefficiencies, optimizing them, and offering a robust solution for dynamic traffic management at unbalanced traffic intersections in urban areas.
Volume: 14
Issue: 4
Page: 2991-3002
Publish at: 2025-08-01

Machine learning for global trade analysis: a hybrid clustering approach using DBSCAN, elbow, and SOM

10.11591/ijai.v14.i4.pp3033-3046
Musdalifa Thamrin , Ida Mulyadi , I Dewa Made Widia , Muhammad Faisal , Suardi Hi Baharuddin , Medy Wismu Prihatmono , Nurdiansyah Nurdiansyah , Nasir Usman
Global trade constitutes a highly complex and interdependent system influenced by diverse economic, geographic, and political factors. This study proposes a hybrid clustering framework that integrates density-based spatial clustering of applications with noise (DBSCAN), elbow, and self-organizing maps (SOM) methods to uncover latent structures in international trade patterns. Utilizing averaged trade data from 25 countries spanning the period from 2013 to 2023, the framework identifies distinct clusters based on export-import characteristics. The DBSCAN is employed to detect dense trade hubs and outlier behaviors, the elbow method determines the optimal number of clusters, and SOM facilitates the visualization of non-linear, high-dimensional trade relationships. The analysis reveals three prominent trade clusters: Global Trade Leaders, Emerging Trade Powers, and Niche Exporters, each reflecting varying degrees of trade diversification and dependency. These empirical findings align with established economic theories, including the Heckscher Ohlin model and dependency theory, and provide actionable insights for policymakers seeking to enhance trade competitiveness and regional integration strategies.
Volume: 14
Issue: 4
Page: 3033-3046
Publish at: 2025-08-01

Optimizing long short-term memory hyperparameter for cryptocurrency sentiment analysis with swarm intelligence algorithms

10.11591/ijai.v14.i4.pp2753-2764
Kristian Ekachandra , Dinar Ajeng Kristiyanti
This study investigates the application of swarm intelligence algorithms, specifically particle swarm optimization (PSO), ant colony optimization (ACO), and cat swarm optimization (CSO), to optimize long short-term memory (LSTM) networks for sentiment analysis in the context of cryptocurrency. By leveraging these optimization techniques, we aimed to enhance both the accuracy and computational efficiency of LSTM models by fine-tuning critical hyperparameters, notably the number of LSTM units. The study involved a comparative analysis of LSTM models optimized with each algorithm, evaluating performance metrics such as accuracy, loss, and execution time. Results indicate that the PSO-LSTM model achieved the highest accuracy at 86.08% and the lowest loss at 0.57, with a reduced execution time of 58.43 seconds, outperforming both ACO-LSTM and CSO-LSTM configurations. These findings underscore the effectiveness of PSO in tuning LSTM parameters and emphasize the potential of swarm intelligence for enhancing neural network performance in real-time sentiment analysis applications. This research contributes to advancing optimized deep learning techniques in high dimensional data environments, with implications for improving cryptocurrency sentiment predictions.
Volume: 14
Issue: 4
Page: 2753-2764
Publish at: 2025-08-01

Enhancing traditional machine learning methods using concatenation two transfer learning for classification desert regions

10.11591/ijai.v14.i4.pp2964-2978
Rafal Nazar Younes AL_Tahan , Ruba Talal Ibrahim
Deserts cover a significant portion of the earth and present environmental and economic difficulties owing to their harsh conditions. Satellite remote sensing images (SRSI) have evolved into an important tool for monitoring and studying these regions as technology has advanced. Machine learning (ML) is critical in evaluating these images and extracting valuable information from them, resulting in a better knowledge of hard settings and increasing efforts toward sustainable development in desert regions. As a result, in this study, four ML approaches were enhanced by hybridizing them with pre-training methods to achieve multi model learning. Two pre-training approaches (Xception and DeneseNet201) were used to extract features, which were concatenated and fed into ML algorithms light gradient boosting model (LGBM), decision tree (DT), k-nearest neighbors (KNN), and naïve Bayes (NB). In addition, an ensemble voting was used to improve the outcomes of ML algorithms (DT, NB, and KNN) and overcome their flaws. The models were tested on two datasets and hybrid LGBM outperformed other traditional ML methods by 99% in accuracy, precision, recall, and F1 score, and by 100% in area under the curve (AUC)-receiver operating characteristic (ROC).
Volume: 14
Issue: 4
Page: 2964-2978
Publish at: 2025-08-01

A fusion convolution neural network-local binary pattern histogram algorithm for emotion recognition in human

10.11591/ijai.v14.i4.pp2734-2740
Arpana G Katti , Chidananda Murthy M V
This paper proposes a fusion of algorithms namely convolution neural networks (CNN) and local binary pattern histogram (LBPH) techniques to comprehend the emotions in humans for greyscale images. In this work, the combined advantages of CNN for its ability to extract features, suitability for image processing and LBPH algorithm to identify the emotions of the human images are included. Though there are enhanced fused algorithms with CNN for image processing, the combination of LBPH with CNN is precise and simple in design. In this work, the secondary data sample is used to recognize the human emotions. The secondary data set consists of 160 samples with emotions of happy, anger, sad, and surprise is considered for making decisions. In comparison, the accuracy of the proposed method is high compared to the other algorithms.
Volume: 14
Issue: 4
Page: 2734-2740
Publish at: 2025-08-01

Machine learning application for particle accelerator optimization-a review

10.11591/ijai.v14.i4.pp3014-3021
Isti Dian Rachmawati , Nazrul Effendy , Taufik Taufik
Particle accelerators receive significant attention from researchers. This machine consists of various interdependent elements, so it is complex. Efficient system tuning and diagnostics are essential for utilizing accelerator technology. In addition, machine learning (ML) has been applied in several applications. ML methods such as artificial neural networks, random forest, reinforcement learning, genetic algorithm, and Bayesian optimization have been used for accelerator optimization. The optimization of particle accelerators covers their performance and efficiency. This paper reviews the application of ML techniques in optimizing particle accelerators, highlighting their importance in addressing the complexity inherent in accelerator systems and advancing accelerator science and technology.
Volume: 14
Issue: 4
Page: 3014-3021
Publish at: 2025-08-01

A reinforcement learning paradigm for Vietnamese aspect-based sentiment analysis

10.11591/ijai.v14.i4.pp3375-3385
Viet The Bui , Linh Thuy Ngo , Oanh Thi Tran
This paper presents the task of aspect-based sentiment analysis (ABSA) that recognizes the sentiment polarity associated with each aspect of entities discussed in customers’ reviews, focusing on a low-resourced language, Vietnamese. Unlike conventional classification approaches, we leverage reinforcement learning (RL) techniques by formulating the task as a Markov decision process. This approach allows an RL agent to handle the hierarchical nature of ABSA, sequentially predicting entities, aspects, and sentiments by exploiting review features and previously predicted labels. The agent seeks to discover optimal policies by maximizing cumulative long-term rewards through accurate entity, aspect, and sentiment predictions. The experimental results on public Vietnamese datasets showed that the proposed approach yielded new state of the art (SOTA) results in both hotel and restaurant domains. Using the best model, we achieved an improvement of 1% to 3% in the F1 scores for detecting aspects and the corresponding sentiment polarity.
Volume: 14
Issue: 4
Page: 3375-3385
Publish at: 2025-08-01

An optimal pheromone-based route discovery stage for 5G communication process in wireless sensor networks

10.11591/ijai.v14.i4.pp2788-2796
Sinduja Mysore Siddaramu , Kanathur Ramaswamy Rekha
The rapid advancement of 5G communication underscores the need for heightened efficiency within wireless sensor networks (WSNs), where challenges such as data loss, inefficiency, and jitter are exacerbated by complex operations. This paper presents the optimal pheromone-based route discovery stage (OpRDS) algorithm, inspired by the natural foraging behaviors of ants, as a novel solution designed to optimize routing processes in the dynamic and demanding 5G environments. The study conducts a comparative analysis of OpRDS against traditional routing protocols, including the ad hoc on-demand distance vector (AODV), destination-sequenced distance-vector (DSDV), dynamic source routing (DSR), and zone routing protocol (ZRP), focusing on key performance metrics such as packet delivery ratio (PDR), latency, throughput, routing overhead (RO), energy consumption (EC), network lifespan, route discovery speed, and scalability. Our results reveal that OpRDS significantly outperforms the conventional protocols, evidencing a 2% increase in PDR, a 5.5% decrease in latency, a 6.7% rise in throughput, an 8.3% reduction in RO, an 11.1% decrease in EC (resulting in an 11% extension of network lifespan), a 10% improvement in route discovery speed, and a 6.7% enhancement in scalability. These findings highlight the algorithm's superior efficiency and adaptability in addressing the robust demands of 5G networks.
Volume: 14
Issue: 4
Page: 2788-2796
Publish at: 2025-08-01

Estimating broiler heat stress using computer vision and machine learning

10.11591/ijai.v14.i4.pp2922-2934
Muhammad Iqbal Anggoro Agung , Eko Mursito Budi , Miranti Indar Mandasari
To optimize and enhance the efficiency of broiler chicken farming, it is essential to maintain the chicken’s welfare, as heat stress can decrease growth efficiency. The temperature-humidity index (THI) is a key indicator used to determine if chickens are experiencing heat stress. Precision livestock farming (PLF) based on computer vision is one method that can assist farmers in continuously and automatically monitoring the condition of their chickens. This research developed a computer vision-based PLF system to observe chickens with CP 707 strain in a commercial farm using the Mask region-based convolutional neural network (Mask R-CNN) method and object tracking algorithms to analyze features such as the cluster index, unrest index, and the distance traveled by broilers. The results indicated that all features tend to inversely correlate with the THI value, with the cluster index showing the most noticeable tendency. Additionally, it was found that external factors, such as the presence of farmers around the observation area, can affect the chickens' behavior, although the cluster index feature is relatively resilient to disturbances if the operator is not captured by the camera. It was concluded that there is a relationship between the features and the THI value; however, these features are not yet sufficient to distinguish the condition of chickens under high and low THI conditions in real-time.
Volume: 14
Issue: 4
Page: 2922-2934
Publish at: 2025-08-01

Improving firewall performance using hybrid of optimization algorithms and decision trees classifier

10.11591/ijai.v14.i4.pp2839-2848
Mosleh M. Abualhaj , Ahmad Adel Abu-Shareha , Sumaya Nabil Al-Khatib , Adeeb M. Alsaaidah , Mohammed Anbar
One of the primary concerns of governments, corporations, and even individual users is their level of online protection. This is because a large number of attacks target their primary assets. A firewall is a critical tool that almost every organization uses to protect its assets. However, firewalls become less reliable when they deal with large amounts of data. One method for reducing the amount of data and enhancing firewall performance is feature selection. The main aim of this study is to enhance the firewall's performance by proposing a new feature selection method. The proposed feature selection method combines the strengths of Harris Hawks optimization (HHO) and whale optimization algorithm (WOA). Experiments were performed utilizing the NSL-KDD dataset to measure the effectiveness of the proposed method. The experiments employed the decision trees (DTs) as a machine classifier. The experimental results show that the achieved accuracy is 98.46% when using HHO/WOA for feature selection and DT for classification, outperforming the HHO and WOA when used separately for feature selection. The study's findings offer insightful information for researchers and practitioners looking to improve firewall effectiveness and efficiency in defending internet connections against changing threats.
Volume: 14
Issue: 4
Page: 2839-2848
Publish at: 2025-08-01

Enhanced pre-broadcast video codec validation using hybrid CNN-LSTM with attention and autoencoder-based anomaly detection

10.11591/ijai.v14.i4.pp2864-2875
Khalid El Fayq , Said Tkatek , Lahcen Idouglid
This study presents a machine learning-based approach for proactive video codec error detection, ensuring uninterrupted television broadcasting for TV Laayoune, part of Morocco’s SNRT network. Building upon previous approaches, our method introduces autoencoders for improved anomaly detection and integrates data augmentation to enhance model resilience to rare codec configurations. By combining convolutional neural networks (CNNs) and long short-term memory (LSTM) networks with an attention mechanism, the system effectively captures spatial and temporal video features. This architecture emphasizes critical metadata attributes that influence video playback quality. Embedded within the broadcasting pipeline, the model enables real-time error detection and alerts, minimizing manual intervention and reducing transmission disruptions. Experimental results demonstrate a 97% accuracy in detecting codec errors, outperforming traditional machine learning models. This study highlights the transformative role of machine learning in broadcasting, enabling scalable deployment across diverse television networks.
Volume: 14
Issue: 4
Page: 2864-2875
Publish at: 2025-08-01

Driving agricultural evolution: implementing agriculture 4.0 with Raspberry Pi and internet of things in Morocco

10.11591/ijai.v14.i4.pp3462-3473
Raja Mouachi , Elbelghiti Youssef , Sanaa El mrini , Mustapha Ezzini , Mustapha Raoufi
The purpose of this project was to investigate the use of embedded system and smartphone technologies in conjunction with Raspberry Pi and NodeMCU to create an intelligent system for smart farming (SF). By means of experiments and comparative analysis carried out in several agricultural contexts, the research evaluated the efficacy of the intelligent system. Results showed that the system was able to handle pertinent agricultural activities and effectively monitor important environmental factors including temperature, humidity, soil moisture, and climatic quality. The system's remote accessibility helped farmers by allowing them to effectively oversee agricultural operations at any time and from any location. As a consequence, SF techniques produced more production, lower costs, and maintained assets.
Volume: 14
Issue: 4
Page: 3462-3473
Publish at: 2025-08-01

Power of blockchain technology for enhancing efficiency transparency and data provenance in supply chain management

10.11591/ijai.v14.i4.pp3452-3461
Kanimozhi Thirunavaukkarasu , Inbavalli Mani
Global supply chains face increasing challenges in improving efficiency, transparency, and compliance with regulatory requirements. Traditional supply chain systems often suffer from inefficiencies due to fragmented data and manual processes, which result in delays and higher costs. Blockchain technology has emerged as a potential solution by offering decentralization, data immutability, and automation through smart contracts. However, existing blockchain implementations struggle with issues like scalability and transaction speed, which limits their effectiveness in supply chain management. This study introduces a new framework based on distributed ledger technology (DLT) with enhanced smart contract functions and data provenance tracking. The framework aims to improve transaction throughput, reduce latency, and provide better data integrity, enabling more efficient and transparent supply chain operations. By incorporating mechanisms to track the origin and movement of goods, the framework ensures that stakeholders have real-time access to accurate information, improving decision-making and trust across the supply chain. We evaluate the performance of this framework using the AnyLogic simulation platform, comparing it to traditional blockchain systems. Metrics such as transaction throughput, latency, and efficiency are analyzed to demonstrate the improvements achieved by the proposed system. The results show significant enhancements in transaction speed and operational efficiency, offering a practical solution for optimizing supply chains in various industries.
Volume: 14
Issue: 4
Page: 3452-3461
Publish at: 2025-08-01

Optimizing signal conversion in uniform FBGs with InGaAs photodetectors for medical sensors

10.12928/telkomnika.v23i4.26164
Tengku; Universitas Riau Emrinaldi , Bambang; National Research and Innovation Agency, KST BJ HABIBIE Widiyatmoko , Bunga; Universitas Riau Meyzia , Sumiaty; Universiti Teknologi Malaysia Ambran , Saktioto; Universitas Riau Saktioto , Mohamad; National Research and Innovation Agency, KST BJ HABIBIE Syahadi , Agitta; National Research and Innovation Agency, KST BJ HABIBIE Rianaris , Dwi; National Research and Innovation Agency, KST BJ HABIBIE Hanto
This study experimentally interrogates the spectral response of uniform fiber Bragg gratings (FBGs) with varying reflectivity levels of 30%, 50%, 70%, and 90% under controlled environmental stimuli. The objective is to elucidate the influence of reflectivity on the wavelength shift behavior of FBGs and to inform the optimal interrogation of these elements with indium gallium arsenide (InGaAs) photodetectors in high-performance sensing systems. Utilizing high-precision measurement procedures and specialized instrumentation, the experiments revealed that the magnitude and pattern of wavelength shifts are significantly influenced by FBG reflectivity. Specifically, lower reflectivity enhances sensitivity, while higher reflectivity contributes to greater spectral stability. These findings highlight the critical role of reflectivity in shaping the spectral modulation characteristics of FBGs, establishing a critical theoretical framework for precision optical sensor systems. The outcomes give significant contributions to the design and calibration of FBG-based sensors, particularly biomedical applications where precision and responsiveness are paramount.
Volume: 23
Issue: 4
Page: 1058-1068
Publish at: 2025-08-01
1 2 3
...
1895 1896
Show 47 of 1896

Discover Our Library

Embark on a journey through our expansive collection of articles and let curiosity lead your path to innovation.

Explore Now
Library 3D Ilustration