International Journal of Education, Science, Technology, and Engineering (IJESTE)

Anomaly Detection Using Autoencoders for Household Electricity Meters

Nattaporn Wongsuwan — 2025-12-25

Household electricity consumption often exhibits sudden and unexplained spikes that typically go unnoticed until the monthly bill arrives. These anomalies may stem from equipment malfunction, inefficient appliance usage, or irregular electrical patterns that households cannot easily observe. This study proposes an unsupervised anomaly detection framework based on autoencoders to identify abnormal consumption behavior from high resolution household electricity meter data. The model learns normal consumption patterns through reconstruction and flags anomalies using a dynamic threshold derived from reconstruction error distribution. Experimental results demonstrate strong detection capability, particularly for sudden spikes, achieving a precision of 0.92, recall of 0.88, and F1 score of 0.90. The findings highlight the potential of deep learning–based unsupervised methods to support real time, edge deployable solutions for energy efficiency and early fault detection in residential environments.

Enhanced Techniques for Detecting Promiscuous Mode using Packet Fu and the Metasploit Framework

Partho Pandya — 2025-12-25

This article argues that Thailand’s public-sector digitalisation has so far failed to realise the principles of Digital Era Governance (DEG) because it remains institutionally and politically anchored in New Public Management (NPM) logic. Rather than enabling platform-based integration and citizen-centric services, digital initiatives have often reproduced audit-centric, siloed practices that prioritise measurable outputs and compliance. Using a policy-analytic approach, document review of national strategies and agency plans, and synthesis of recent literature and sectoral case examples; the article identifies three mechanisms by which NPM logic is perpetuated in Thailand’s digital transition: (1) proliferation of discrete applications driven by performance reporting and agency visibility; (2) digital tools as instruments of control and compliance rather than coordination; and (3) governance fragmentation and weak interoperability governance. The paper concludes with targeted policy recommendations to reorient Thailand’s digitalisation toward DEG: consolidate digital architecture around shared platforms and standards, redesign performance regimes to reward integration and outcomes, and strengthen cross-agency data governance.

An Exploratory Data Analysis Approach for Tax Revenue Systems

Ayomitope Isijola — 2025-12-25

Tax collation is an important part of any company's revenue system. Usually, over time, the process becomes more daunting, and the ability to monitor tax trends and revenue streams decreases. Not to mention gaining useful insights that can aid in decision-making and company transactions. That is why a tax data analysis system is able to continuously monitor tax information, pointing out anomalies, trends, and providing useful data visualizations. The tax analysis system would also enhance transparency and accountability in tax collection, improve efficiency, and reduce the need for audits, hence underlining its potential. Tax data is an important collection of information; however, many businesses fail to take advantage of this by not digging deeper into that collection. The aim of this research is to explore tax and sales data in an attempt to gain valuable insights and provide clearer information to the user. The methodology adopted is Exploratory Data Analysis (EDA) using Python as the main tool. The dataset used consists of 5,200 transactional tax records obtained from small and medium-scale enterprise (SME) sales reports spanning a 24-month period (Jan 2022 – Dec 2023). All data contained fields were pre-processed and stored in an SQLite database. Using Python libraries like Pandas, Matplotlib, Plotly, descriptive statistics, and visualization analyses showed that corporate tax contributions accounted for 47.8% of total tax revenue, while sales tax trends fluctuated seasonally, peaking in Q2 and Q4 of each fiscal year. The analysis demonstrated a 12% improvement in tax insight accuracy compared to manual spreadsheet tracking. The results show that with the approach, tax data can provide insights that can inform business decisions through charts and graphs. In conclusion, the platform can be a great tool in business decision-making and breaking down large datasets to give meaningful information.

The Persistence of New Public Management Logic in the Digital Government Transition

Anucha Vanchai Kittisak — 2025-12-25

This article argues that Thailand’s public-sector digitalisation has so far failed to realise the principles of Digital Era Governance (DEG) because it remains institutionally and politically anchored in New Public Management (NPM) logic. Rather than enabling platform-based integration and citizen-centric services, digital initiatives have often reproduced audit-centric, siloed practices that prioritise measurable outputs and compliance. Using a policy-analytic approach, document review of national strategies and agency plans, and synthesis of recent literature and sectoral case examples. the article identifies three mechanisms by which NPM logic is perpetuated in Thailand’s digital transition: (1) proliferation of discrete applications driven by performance reporting and agency visibility; (2) digital tools as instruments of control and compliance rather than coordination; and (3) governance fragmentation and weak interoperability governance. The paper concludes with targeted policy recommendations to reorient Thailand’s digitalisation toward DEG: consolidate digital architecture around shared platforms and standards, redesign performance regimes to reward integration and outcomes, and strengthen cross-agency data governance.

IoT-Based Monitoring System for Smart Agriculture to Enhance Crop Yield Efficiency

Pham Van Anh — 2025-12-25

This study investigates an IoT architecture for smart agriculture that combines event-triggered sensing with edge-level multi-sensor fusion to reduce energy consumption across distributed sensor networks. While prior research has largely focused on optimizing individual node efficiency, our findings reveal that the primary source of energy savings arises from systemic behavioral changes within the network’s communication ecology. Real-world experiments on a multi-node deployment show that edge fusion reduces redundant transmissions, stabilizes medium-access contention, and significantly extends sleep intervals. Collectively, these effects produce an average 30% reduction in wake-up frequency, even in relay nodes that do not perform fusion. The results indicate that the underlying mechanism is ecological rather than local: by lowering network-wide communication turbulence, the system achieves a more stable, low-activity state, allowing overlapping dormancy clusters to form naturally. This challenges the long-standing assumption that energy efficiency must be pursued primarily at the node level. Limitations include the controlled experimental environment, moderate network scale, and potential latency risks in time-critical scenarios. The study’s theoretical contribution lies in reframing energy-efficient IoT design as a complex adaptive systems problem, where efficiency emerges from interactions across the network rather than isolated node behavior. This ecosystem-centric perspective opens new directions for sustainable IoT architectures. Future research should focus on designing protocols, topology strategies, and fusion mechanisms that deliberately shape systemic behavior in IoT networks, aiming to achieve greater efficiency, resilience, and longevity than node-centric approaches alone.