: The hospitality industry is undergoing a digital shift, with online booking platforms playing an increasingly crucial role (Sriphaew & Katkaeo, 2017). This study examined the online reservation system (ORS) used at the Cebu Technological University (CTU) Danao Campus Facility Centrum. This comprehensive evaluation of the CTU Danao Campus Online Reservation System (ORS) combined both quantitative and qualitative methods, including performance metrics (such as response time and reliability) and user experience assessments (via surveys, interviews, and usability testing) (Wu, 2018). The study assessed key components of the system, with the following average ratings: user interface (4.52), performance (4.70), features (4.50), accessibility (4.59), enhanced security (4.39), and overall quality (4.51). The overall weighted mean for perceived quality was 4.54, with all ratings Strongly Agree. These results reflect high user satisfaction and system functionality, consistent with prior ORS evaluations in education and service contexts (Alkhaldi et al., 2018; Subiyakto et al., 2021). The findings indicate that the ORS is highly regarded by its users and performs reliably across a range of indicators. Its strengths lie in its user-friendly interface, fast performance, accessible features, and secure environment (IJRES, 2023). However, the evaluation also uncovered minor areas for improvement, particularly in strengthening security features and further optimizing the user experience. It is recommended that the development team continue refining the system based on user feedback, particularly by enhancing security protocols and conducting regular evaluations to maintain and improve system performance (Stakeholders’ Experiences, 2025). Continuous iteration will ensure that the ORS remains effective, efficient, and aligned with the evolving needs of the CTU Danao Campus community
With mobile communications becoming part of Nigerian daily life, SMS phishing attacks are taking a huge turn for the worse, targeting unsuspecting users with fraudulent messages that trick them into providing personal and financial information. This study proposes a user-centric SMS phishing detection system tailored to the Nigerian context to design and implement an intelligent and user-friendly solution capable of accurately detecting phishing SMS messages and to raise user awareness using multilingual messages and easy feedback channels. The four classification algorithms trained and tested were Support Vector Machine (SVM), Random Forest, Multinomial Naive Bayes and Logistic Regression, whose performance metrics were accuracy, precision, recall, F1-score and Area Under the ROC Curve (AUC-ROC). Cross-validation was employed to ensure reliability and generalisability of the model. The result obtained from the models showed that the Support Vector Machine model performed best with the highest overall accuracy and good generalisation error in the classification of high-dimensional text data; thus, it was selected as the final model to be deployed. To enhance access by multiple user groups, the system was further enhanced by the provision of a web-based interface, several Nigerian language supports, visual warnings, and audio prompts. In conclusion, the developed system correctly identifies phishing messages and provides an easy way to reduce fraud on mobile devices in Nigeria. Future research should focus on expanding the dataset with more native language samples, adding deep learning models to improve multilingual understanding and partnering with telecommunication firms to make real-time fraud prevention and extensive use
This rapid pace of digitization of financial services has greatly amplified the scale, complexity, and variety of financial transactions. Consequently, it has led to an increase in sophisticated fraudulent activities that challenge conventional detection methods. Manual and traditional solutions that are rule-based are unlikely to be effective in dealing with up-and-coming trends of fraud since they possess low levels of flexibility, high operation costs, and constrained capabilities of responding to fraud. As a result, AI has turned into a paradigm shift that can enhance the detection of fraud through automated learning and real-time analysis, and high-precision anomaly detection. The survey gives a detailed analysis of AI-based ML, deep learning (DL), unsupervised learning, and natural language processing (NLP) models applied to detect fraudulent behaviour in financial ecosystems. The paper has provided a detailed overview of what external and internal fraud is, examined traditional approaches and identified the downsides that can only be addressed with the help of AI. Besides that, it offers the most recent publications to analyze the progress of algorithms, performance metrics, the introduction of new tendencies, and the increased role of explainable and privacy-conscious AI. This work offers a comparative analysis of the latest models and points out how AI could make current fraud detection techniques more accurate, scalable, and flexible. The findings suggest that additional innovation needs to be maintained to address more intricate, data-driven financial fraud problems.
This study presents the design and implementation of an intelligent Human Resource Information System (HRIS) for the Department of Education (DepEd) Schools Division Office (SDO) of Laguna, focused on enhancing Recruitment, Selection, and Placement (RSP) processes through a Decision Support System (DSS). Anchored in Decision Theory, Human Resource Management (HRM) Theory, and Systems Theory, the study responds to challenges in traditional recruitment methods, such as inefficiencies, bias, and data handling limitations, by leveraging advanced technologies, including Natural Language Processing (NLP) and the Naive Bayes classifier. Employing a mixed methods approach, the research integrates structured applicant data, automates evaluations, and utilizes predictive analytics to ensure strategic, data-driven decision making. Results highlight the system's ability to streamline applicant screening, enhance transparency, and optimize placement accuracy. Ultimately, the developed HRIS-DSS model demonstrates a scalable, replicable framework for modernizing public education HR operations and significantly advances the digital transformation of government HR practices
Modern healthcare monitoring systems play a vital role in enhancing patients’ quality of life by facilitating continuous examination of their health conditions. In healthcare industry, diabetes mellitus is a key risk to public health, and early detection is essential for efficient care and issue prevention. Although numerous deep learning methods have explored smart healthcare systems for diabetes prediction, achieving high accuracy along with time-efficient prediction remains a significant challenge.In order to address these challenges, a novel method called Quadratic MapReduce domain adaptive Discriminative Artificial Intelligence (QMDADAI) is introduced. The main objective of QMDADAI model is to perform accurate diabetes disease prediction with high accuracy and minimal time consumption in healthcare monitoring system. The Discriminative AI model comprises of various processing steps, namely data acquisition, preprocessing, feature selection and classification to enhance the performance of diabetes disease prediction. In the data acquisition phase, the numerous patient data samples are collected from the dataset. Subsequently, the data pre-processing stage is carried out which includes two major processes namely missing data handling and outlier data removal from the input dataset. After data pre-processing, the more important feature selection process is carried out using MapReduce framework from the input dataset. Once the features selected, the classification process is carried out using sequential rank correlation for diabetes disease prediction with higher accuracy. The fine tuning process of Discriminative AI is done using Bats Echolocation optimization algorithm for minimizing the error and enhancing the accuracy of the diabetes disease prediction.Experimental assessmentof QMDADAI model is conducted with different evaluation metrics such as accuracy, precision, recall, F1 score, specificity, ROC-AUC, confusion matrix and training time. The results indicate that the proposed QMDADAI model achieved higher accuracy in diabetes disease prediction with minimal time consumption compared to existing deep learning methods.
The rapid integration of artificial intelligence (AI) into the U.S. labor market presents significant challenges for accurately forecasting employment trends, skill requirements, and workforce development needs. This paper examines how the U.S. Bureau of Labor Statistics (BLS) can enhance its employment projection methodologies to better capture AI’s impact on occupations, worker skills, and educational requirements. Drawing on 26 recent empirical studies and BLS’s existing frameworks, we summarize a comprehensive approach that combines task-based exposure modeling, real-time data analytics, causal inference methods, and improved gross flows estimation for tracking worker transitions. Key focus include a discussion on Dynamic Occupational AI Exposure Score (OAIES) that distinguishes between automation risk and augmentation potential at the task level, enhanced data collection strategies using job postings and administrative records, Bayesian inference methods for survey estimation, and refined methods for estimating how workers move between occupations as AI transforms job requirements. The paper integrates findings from multiple BLS methodological studies on productivity measurement, price indices, and employment projections. These enhancements would provide educators, policymakers, and workforce development professionals with more accurate, timely information to design training programs, allocate resources, and prepare students for an AI-driven economy. The paper concludes with a phased implementation strategy and recommendations for collaboration between BLS, educational institutions, and workforce agencies. This is a review paper and all ideas are from cited references.
The oil and gas pipeline industry is one of the sectors that moving towards using AI for enhancing equipment reliability, safety of operations and maintenance effectiveness in complex transmission networks' operation. The paper presents a complete summary of predictive and preventive maintenance measures in pipeline systems with the focus on AI-guided predictive analytics used in asset integrity management. The condition-based and risk-based maintenance models, time series prediction, classification, regression and anomaly detection algorithms that are used in failure forecasting and leak detection. Advanced sensing devices, such as smart sensors embedded in SCADA, and distributed monitoring systems, aid in real-time data capture and make it possible to make intelligent decisions. The mechanisms of corrosion and leakage in CO₂ pipelines are given special consideration as thermodynamic and environmental factors exacerbate operational risks in CO₂ pipelines. New trends like digital twins, edge computing, hybrid energy optimization, sustainability-driven metrics, etc. are analyzed as well. Although there has been a great improvement, the issues of the heterogeneity of data, model generalization, interpretability, and cybersecurity continue to be a major limitation to scalable AI implementation in high-stakes oil and gas environments.
Learning history as a simple memorization of events does not guarantee learning that allows one to use what has been assimilated for the benefit of oneself and society. Therefore, new learning methods must be developed that transcend passive memorization to achieve meaningful and lasting assimilation. This article presents the formal development of an Intelligent Tutoring System (ITS) for teaching agriculture in the Teotihuacan culture, implemented in an immersive Virtual Reality (VR) environment. The system integrates a virtual tutor based on Artificial Intelligence (AI) that guides the student through a gamified maize cultivation process. The main contribution of this work is the proposal of a mathematical formalism that models the gamified learning process. This model, based on Automata Theory and Markov Decision Processes (MDPs), defines the student's progress states, available actions, and transitions, which are influenced by both the user's decisions and the tutor's evaluation. Formalism allows for the systematic implementation and quantitative evaluation of the impact of gamification and AI tutoring on the acquisition of historical knowledge. The resulting virtual environment simulates the Valley of Teotihuacan, where the user interacts with pre-Hispanic tools (coa, uictli) and confronts natural phenomena, aiming to achieve a successful harvest. The final result is evaluated through an adaptive questionnaire. This approach demonstrates how the mathematical formalization of pedagogy in virtual environments can optimize the design of immersive, effective educational experiences.
: In the rapidly evolving digital healthcare ecosystem, individuals often face challenges in monitoring their emotional well-being and accessing time-ly medical support. Many existing healthcare platforms lack intelligent mechanisms to continuously analyze users’ mental health conditions and provide real-time assistance based on their emotional state. To address this issue, this project proposes the development of an intelligent health monitoring system called MindVerge, which integrates advanced sentiment analysis techniques powered by Large Language Models (LLMs) to enhance user well-being and accessibility to healthcare services. The proposed system follows a structured and multi-functional approach that com-bines natural language processing and real-time data integration. It analyses user-generated text inputs such as chat messages or health logs to detect emotional states, including stress, anxiety, positivity, or negativity. The system utilizes NLP techniques such as contextual text analysis and deep learning-based language understanding to accurately interpret user emotions and identify potential mental health risks. In addition to sentiment analysis, the system incorporates location-based healthcare support using Google Maps API to help users discover nearby hospitals and medical professionals. A distance calculation mechanism based on the Haversine formula ensures accurate recommendations of the nearest healthcare facilities. The platform also includes an appointment booking system that allows users to schedule consultations directly and re-ceive real-time updates on appointment status. Furthermore, the system provides secure authentication, data privacy, and role-based access for users and hospital staff. Emotional data and appointment records are safely stored and managed within the database to ensure reliability and confidentiali-ty. By integrating intelligent emotional analysis with real-time healthcare connectivity, Mind Verge offers a user-friendly, efficient, and proactive solution for continuous mental health monitoring and support.
The rapid integration of generative artificial intelligence (AI) tools into higher education has transformed how students seek information, complete tasks, and engage with their academic environments. This study investigates the relationship between AI tool usage and social connectedness among undergraduate Information Technology (IT) students in the Philippines. Grounded in social interdependence theory and academic belonging literature, the study examines whether increasing reliance on AI platforms such as ChatGPT, GitHub Copilot, and Google Gemini is associated with changes in students' help-seeking and peer interaction practicesin educational contexts. Using a descriptive quantitative design and cross-sectional survey method, data were collected from 45 IT students enrolled in a higher education institution in Cebu, Philippines. The research instrument, a researcher-developed questionnaire administered via Google Forms, measured four constructs: AI usage frequency, peer collaboration practices, learning preferences, and perceived impact of AI on social connectedness. Descriptive statistics including frequency counts, percentages, and weighted means were used to analyze the data. Results revealed that while students reported generally high levels of social connectedness (M = 5.46), academic interaction patterns showed a moderate composite mean (M = 3.85), with over 64% preferring to solve problems independently. Approximately 40% reported consulting AI instead of peers when seeking academic help. These findings suggest a tension between perceived social belonging and actual peer engagement behaviors. Although AI tools are viewed positively for academic performance and independent learning, their substitutive use may quietly erode the human connections integral to effective learning communities. The study recommends intentional pedagogical strategies that position AI as a complement to, rather than a replacement for, peer collaboration in IT education
The rapid growth in energy consumption by buildings and the limitations of conventional HVAC control systems have created an urgent need for more efficient and intelligent solutions. Conventional control methods do not keep up with dynamic occupancy and environmental variability, and complicated thermal interactions, and result in unreasonable energy consumption and decreased comfort. In the paper, the detailed overview of the approaches of HVAC system modeling and intelligent methodology of building energy optimization is described. The paper is a systematic review of the basis of HVAC, system components, centralized, decentralized, and hybrid systems. Various gray-box models are introduced and examined with regard to their utility in predicting thermal behavior and making decisions pertaining to controls. These models include physics-based, data-driven, and hybrid versions. Fuzzy logic, ML, DL, and RL are some of the smart control technologies that are compared to more traditional PID-based methods. Energy efficiency and thermal comfort can be achieved with smart HVAC control, according to recent literature. The paper also points out major challenges, limitations and possible research directions to building adaptive, data-driven and sustainable HVAC system in the modern buildings
Pest infestation and plant diseases significantly reduce agricultural productivity and pose a serious threat to food security and farmers livelihoods. Conventional disease identification techniques, which depend on expert manual examination, are often labour-intensive, time-consuming, and prone to human mistake, particularly in large-scale cultivation. Therefore, early and precise plant disease identification is essential for efficient crop management and better yields. This study suggests an intelligent deep learning-based system for autonomous disease detection and classification using leaf images. The main goal is to create a reliable and effective model that combines sophisticated computer vision techniques for accurate diseases localization and identification with convolutional neural networks (CNNs) for deep feature extraction. CNNs can learn discriminative and hierarchical features from pixel-level inputs, which makes them ideal for visual pattern identification. The proposed system incorporates the You Only Look Once (YOLO) technique to enhance real-time detection performance. Because it better balances accuracy and detection speed, the YOLOv3 model is specifically employed. The proposed Conv-YOLOv3 architecture enables simultaneous disease localization and classification in the leaves of ginger plants. The primary disease indicators that the system is designed to identify are soft rot disease, insect infestation patterns, and indications of nutrient deficits.. The proposed model's performance is evaluated in terms of inference time, computational complexity, detection efficiency, and classification accuracy. According to experimental results, the Conv-YOLOv3 model has a 93.16% overall accuracy, demonstrating its excellent capacity for accurate disease identification in practical settings. The suggested system helps farmers and agricultural specialists make timely decisions by providing a scalable, affordable, and automated precision agriculture solution. This methodology supports more productive agriculture and sustainable crop management by facilitating early disease identification and focused action
Deepfake technology has rapidly advanced in recent years, raising significant concerns about misinformation, identity manipulation, and the misuse of synthetic media. Although many deepfake detection systems reported in academic literature demonstrate strong benchmark accuracy, relatively little research focuses on the practical reliability and accessibility of publicly usable detection tools for ordinary users. This study addresses that gap through a practical evaluation of publicly accessible deepfake detection systems under benchmark and real-world conditions. A preliminary accessibility survey of publicly referenced deepfake detection platforms was first conducted. Several systems were found to be restricted, unstable, enterprise-focused, video-only, or unavailable for consistent public experimentation. As a result, Hive Moderation was selected as the primary detector for evaluation. The methodology included evaluating benchmark image samples from FaceForensics++ and Celeb-DF, comparing detector performance with human participants, and testing real-world deepfake images collected from online sources. The results demonstrated a noticeable performance decline on more realistic, real-world deepfakes. Human participants and the evaluated detector exhibited varying performance across benchmark datasets, while both showed reduced accuracy on Celeb-DF and real-world samples. In addition, several incorrect classifications occurred with extremely high confidence scores, suggesting inconsistency in prediction certainty and practical reliability
In modern era, agriculture sector has many challenges facing by farming communities. Numbers of services with ICT tools are being provided by government/ private leading departments to support agriculture activities and increase productivity but still needs to indentify information gaps that may help in effective decision making. There are need to address the efficient use of data mining techniques so that effective and accurate decision making can be done. This study presents a review of literature done by different authors. Experiments related to supervised learning (classification) techniques has also been done. Content which has been covered in literature review are application of data mining, machine learning, artificial intelligence in different sub fields of agriculture like disease detection, yield prediction, crop quality etc. Classification algorithms have been used to classify different types of crops. Five classification algorithms NB, SVM, NN and Decision Tree have been selected. Comparison of different parameters such as Accuracy, Kappa Statistic, RMSE and so on related to classification model have been done. It has been found that the decision tree (PART) algorithm is more suitable than other classifier on selected dataset.
Due to the increasing prevalence of muscular disorders linked to modern lifestyles, yoga has regained attention for its therapeutic benefits, yet accurate and efficient pose recognition remains a challenge. This paper presents CBYAR-Net, a novel content-based yoga asana retrieval framework designed to address key challenges, including variability in asana appearance, dataset scarcity, and high computational demands. The proposed system integrates Structural Detail Descriptor (SDD) and Spatial Color Distribution Descriptor (SCDD) to capture fine-grained structural features and spatial color patterns, while an unsupervised SVM (U-SVM) clusters similar feature vectors for efficient retrieval. Experimental results demonstrate that CBYAR-Net outperforms traditional methods like KNN and Cosine Similarity, achieving 96.15% retrieval accuracy. The framework provides a robust, scalable, and computationally efficient solution for automated yoga pose recognition and retrieval
This study assesses the performance and viability of the Modern Eco-Smart Egg Hatcher, an automated, solar-powered system designed to optimize the poultry incubation process. To evaluate the system, the researchers used a Key Performance Indicator (KPI) approach to collect data from a purposively selected group of 20 respondents: 10 IT experts, 5 game fowl breeders, and 5 poultry farmers. The system was rigorously evaluated based on four primary criteria: accuracy, effectiveness, functionality, and reliability. Quantitative findings demonstrated exceptional performance and high satisfaction across all metrics, with the system achieving outstanding grand mean scores in functionality (4.94), reliability (4.86), effectiveness (4.84), and accuracy (4.78). A comparative analysis across the evaluator groups further underscored the system's success, with IT experts and end users reporting weighted means of 4.86 and 4.88, respectively. The study concludes that the hatcher is highly reliable, maintains ideal incubation conditions with minimal user intervention, and successfully meets the practical demands of its target beneficiaries.
Attendance monitoring is an essential administrative task in educational and workplace environments, but traditional manual registers are time-consuming and prone to proxy attendance and recording errors. Earlier automated methods such as fingerprint and RFID systems reduce manual effort but require physical interaction and dedicated hardware. Recent advances in computer vision and artificial intelligence enable contactless attendance monitoring using face recognition, where deep learning models such as CNN, MTCNN, Face Net, lightweight networks, and ensemble approaches identify multiple individuals in real time through video surveillance. This survey analyzes multiple research works on face recognition-based smart attendance systems by comparing their methodologies, performance, advantages, and limitations. The study highlights major research gaps, including absence of continuous presence monitoring, lack of exit detection, sensitivity to illumination variation, occlusion challenges, and hardware constraints, and suggests future directions toward intelligent presence aware attendance systems suitable for real-world deployment.
This paper presents a systematic review comparing the effectiveness of two dominant usability evaluation methods: Heuristic Evaluation (HE) and Think-Aloud Protocols (TAP). Against the backdrop of rapid technological advancement between 2019 and 2025 including the proliferation of mobile health (mHealth) applications, AI-integrated systems, and the necessity of remote testing precipitated by the COVID-19 pandemic this research evaluates the quantity, severity, and typology of usability problems detected by each method. By aggregating data from 25 comparative studies published within this period, the analysis reveals a distinct dichotomy: while Heuristic Evaluation remains the superior method for identifying high volumes of surface-level and consistency issues at a low cost, Think-Aloud Protocols are indispensable for uncovering severe, task-oriented cognitive friction points that experts often overlook. Crucially, recent data indicate that the shift to remote moderated TAP has maintained data quality while reducing logistical overhead, narrowing the cost gap between the two methods. The study concludes that a hybrid methodology, sequenced specifically to leverage Heurustic Evaluation for cleaning and TAP for validating, yields the most comprehensive usability assurance in modern agile development cycles
Explainable artificial intelligence (XAI) methods such as SHapley Additive exPlanations (SHAP) and Local Interpretable Model agnostic Explanations (LIME) are increasingly deployed in health surveillance systems to provide transparency in machine learning predictions. However, concerns persist regarding the reliability of these explanations under imperfect data conditions commonly encountered in resource-limited settings. This study investigates the robustness of SHAP and LIME explanations for malaria test positivity rate prediction under systematic data perturbations. Three gradient boosting models (XGBoost, LightGBM, CatBoost) were trained on malaria surveillance data from Bayelsa State, Nigeria comprising 2,100 records across eight local government areas. Model explanations were evaluated under controlled perturbations including Gaussian noise (5 to 100 percent), missing value injection (5 to 50 percent), and feature corruption (5 to 50 percent). Stability was quantified using Spearman rank correlation and top k feature overlap metrics. Results demonstrate exceptional robustness of SHAP explanations, with mean Spearman correlation coefficients of 0.976 for XGBoost, 0.981 for LightGBM, and 0.982 for CatBoost. SHAP consistently outperformed LIME across all conditions. The top five most important features remained consistent across most perturbation scenarios with 100 percent overlap for XGBoost and CatBoost SHAP. These findings provide confidence for deploying XAI based decision support systems in malaria surveillance programs where data quality may be suboptimal
The Telecommunications industry is preparing to enter the post-5G phase, and the paper undertakes the task of outlining the pillars and visionary future that shape the course of the 6G wireless communication systems. As the industry recognizes the insatiable need to increase data rates and expand network coverage, this article traces the evolution of the current 5G infrastructure to the emerging 6G framework, which lays the groundwork for revolutionary strides in the 2030s. This review examines the transformation of telecom provisioning from outdated, tightly interlinked systems to contemporary, event-driven, cloud-native architectures that sustain the growth of intelligent, scalable communication networks. It highlights the disadvantages of SOA-based systems and illustrates how Event-Driven Architecture (EDA) offers greater modularity, faster responsiveness, and real-time processing. The paper discusses infrastructure auditing, the 5G transition, and security improvements as part of modernization processes, while considering integration challenges in mixed, multi-supplier environments. It further analyzes the use of Kafka and similar streaming technologies for the purpose of automated user provisioning, which in turn, leads to higher operational efficiency. The research forecasts 6G and identifies major architectural paradigms and enabling technologies such as AI, Integrated Sensing and Communication (ISAC), and blockchain, highlighting their role in creating resilient telecom systems ready for the future.
