Asian Journal of Research in Computer Science
Not a member yet
792 research outputs found
Sort by
An Enhanced Model for Intrusion Detection in a Cloud Computing Environment
No full textIntrusion is an important issue in computer networks especially in cloud computing where all the services are served using the internet. The fully distributed and open structure of cloud computing and services has made it an even more attractive target for potential intruders. The more sophisticated hackers and attackers get, the more there is work for the defense to prevent such attacks. A cloud computing system can be exposed to threats which include the integrity, confidentiality, and availability of its resources, its data, and the virtualized infrastructure can be vulnerable. The problem becomes bigger when an internal intruder misuses a cloud with massive computing power and storage capacity as a malicious party. This research developed an enhanced model for intrusion detection that monitors and analyzes data in a cloud environment and detects intrusion in the system or network. The model can detect intrusions from external and malicious internal (authorized and unauthorized) users by normalizing and classifying all data packets using machine learning techniques. The developed system is an enhanced model of Zhang by combining it with two machine learning techniques: Support vector machine and Bayesian network to aid in the classification of normal data and intrusion data to detect intrusions. The developed model is evaluated and found to be able to make strong predictions, detect attacks, and still maintain the efficiency of the network. The system, when implemented, can detect intruders by classification of data packets and also improve the existing system in terms of providing more accurate and more efficient intrusion detection. It also provides worthwhile information about malicious network traffic, helping to identify the source of the incoming probes or attacks, collecting forensic evidence that can be used to identify intruders, and alerting security personnel that a network invasion may be in progress
Packets’ Congestion Management with Fuzzy Admission Control Policy for Differentiated Service Networks
No full textThis study presents a priority-based admission control system for ensuring stability in a differentiated service network using a fuzzified admission policy. Packets of varying sizes of data were transmitted from N sources through a traffic conditioner which categorizes the packets into two independent sources, consequently classifying them into “high” and “low” priorities. While class A packets are not denied admission into the buffer, this is not the case with packets of class B. Arrivals from both sources follow a Poisson distribution process of(k) = ( /k!)e - . It is assumed that r > h / in order to avoid a situation in which a class B arrival is denied admission while the system is empty. The arrival rates 1 [02 ) and 2 [02 ) serve as fuzzy inputs with four linguistic values while the output is a decision, d. Simulation started with an initial state, zero and system performance for the first 300time units was monitored. Results indicate that the admission controller admits arriving class B packets provided that the value of y ≤ 4 while it denied admission to arriving class B packets when y > 4, thus giving a threshold policy of y=4 . Arrivals denied admission are dropped and transmitted to the “tree manager” via the bottleneck router. These packets are arranged as nodes in an AVL tree structure which adopts tree properties to manage and transmit nodes to the buffer based on node rotations
A 5 Year Bibliometric Review of Programming Language Research Dynamics in Southeast Asia (2018-2023)
No full textAims: To conduct a systematic examination and bibliometric analysis of Scopus-indexed literature focusing on emerging trends in programming languages research within Southeast Asia.
Study Design: This study employs a mixed method approach, incorporating both qualitative and bibliometric analysis.
Place and Duration of Study: Publication data for review was obtained from the Scopus database, covering the period from 2018 to 2023, with a specific focus on the progress in programming language and semantics research within ASEAN countries.
Methodology: We used the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) protocol to collect publication data. Bibliometric data was visualized through Biblioshiny and VOSviewer.
Results: From 2018 to 2023, the research production involving programming languages and semantics across ASEAN countries has been strong, yielding a total of 233 documents from 160 unique sources. However, the annual growth rate was at -10.87%. There was a total of 882 authors with only 10 sole authors in the field. 46.78% of the documents had international co-authorship with an average of 4.03 authors per document. The literature spanned across 764 unique author keywords and 7424 citations with an average of 11.12 citations per document.
Conclusion: Southeast Asia has a rich and collaborative research space in the field of Programming Languages but it faces several barriers such as the absence of a unified research agenda, the lack of adequate funding, and the relatively weak industrial base
Exploring the Security Challenges of an E-Voting System (EVS): A Case Study in the Bolgatanga Senior High School
No full textThe integration of electronic voting systems (EVS) into electoral processes promises increased efficiency, speed, and transparency. However, this technology also introduces a variety of security challenges that must be addressed to ensure the integrity of the voting process. This study explores the security vulnerabilities of an EVS, focusing on its implementation at Bolgatanga Senior High School (BIGBOSS).
Through a case study approach, this research examines the potential threats to the administration of EVS, including issues related to data privacy, system integrity, authentication protocols, and susceptibility to cyber-attacks. A mixed-methods strategy was employed, utilizing surveys, interviews, and system tests to assess both technical vulnerabilities and user experiences. Key findings revealed Potential Voter Manipulation where unauthorized access could allow individuals to cast votes fraudulently; Insider Threats where trusted users exploiting their access could manipulate election outcomes; Weak Authentication; Lack of Robust Verification; Potential Physical Attack where individuals could vandalize machines and Vulnerabilities in Encryption which may expose votes, compromising ballot secrecy. The study recommends the implementation a blockchain-based technology , such as cryptographic techniques, consensus mechanisms, use of biometric verification to ensure the identity of voters as well as the use of digital signatures to verify the authenticity of electronic ballots and ensure that they haven\u27t been altered during transmission, implementation of real-time monitoring of the e-voting system to detect and respond to security threats promptly” and also provide ongoing cybersecurity education and training to election officials, staff, and voters to raise awareness and reduce the risk of human error. These findings contribute to the ongoing discourse on EVS security, providing insights into practical measures for improving the reliability and safety of electronic voting systems in educational institutions and beyond
Assessing Performance of Estimation Techniques in Time Series Analysis when Trend-cycle Component is Linear
No full textAbstract: Two decomposition techniques are Buys-Ballot and least square techniques are presented in this study. The two important patterns that may be discussed are trend and seasonality and two competing models are additive and multiplicative models. The trend-cycle component is linear. The emphasis is to assess the performance of Buys-Ballot estimates and least square estimates using accuracy measures (Mean Error (ME), Mean Square Error (MSE), Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE). Results show that the two estimation techniques are very good in estimating the linear trend parameters and seasonal effects when the model for decomposition is additive. It differs for multiplicative model
Artificial Intelligence and Information Governance: Strengthening Global Security, through Compliance Frameworks, and Data Security
No full textThis study examines the dual role of artificial intelligence (AI) in advancing and challenging global information governance and data security. By leveraging methodologies such as Hierarchical Cluster Analysis (HCA), Principal Component Analysis (PCA), Structural Equation Modeling (SEM), and Multi-Criteria Decision Analysis (MCDA), the study investigates AI-specific vulnerabilities, governance gaps, and the effectiveness of compliance frameworks. Data from the MITRE ATT&CK Framework, AI Incident Database, Global Cybersecurity Index (GCI), and National Vulnerability Database (NVD) form the empirical foundation for this analysis. Key findings reveal that AI-driven data breaches exhibit the highest regulatory scores (0.72) and dependency levels (0.81), underscoring the critical need for robust compliance frameworks in high-risk AI environments. PCA identifies regulatory gaps (45.3% variance) and AI technology type (30.2% variance) as significant factors influencing security outcomes. SEM highlights governance strength as a primary determinant of security effectiveness (coefficient = 0.68, p < 0.001), while MCDA underscores the importance of adaptability in governance frameworks for addressing AI-specific threats. The study recommends adopting quantum-resistant encryption, enhancing international cooperation, and integrating AI automation with human oversight to fortify governance structures. These insights provide actionable strategies for policymakers, industry leaders, and researchers to navigate the complexities of AI governance and align technological advancements with ethical and security imperatives in a rapidly evolving digital landscape
Optimizing Energy Efficiency in Smart Home Automation through Reinforcement Learning and Iot
No full textThis thesis explores an innovative approach to optimizing energy efficiency in smart home environments by leveraging reinforcement learning (RL) and Internet of Things (IoT) technologies. As global energy demand rises and concerns over environmental sustainability intensify, smart homes offer a promising solution to reduce residential energy consumption while enhancing user comfort. The study presents a comprehensive architecture integrating IoT devices with RL algorithms, allowing for real-time monitoring and intelligent energy management. Through data collected from smart sensors, RL agents continuously learn and adapt to occupant behaviors and environmental changes, making optimal decisions to minimize energy usage without compromising user comfort. A real word-based analysis demonstrates that the proposed system achieves significant energy savings compared to traditional rule-based methods. The results underscore the effectiveness of combining RL and IoT for adaptive energy management, paving the way for scalable solutions that could extend to smart cities and renewable energy systems. This research provides valuable insights into how emerging technologies can contribute to sustainable energy practices in the residential sector
Rose Plant Leaf Disease Recognition Using Machine Learning Methodologies
No full textThe most popular flowers in the world are roses, not only cheer people up but also support livelihoods. Diseases, however, can harm these priceless flowers\u27 health and negatively affect both their quality and the growers\u27 livelihoods. The increased occurrence of ailments in rose plants poses a severe danger to the ornamental flower industry and agricultural productivity. In this paper, we describe a novel deep learning-based method for the automated diagnosis of leaf diseases in rose plants. A big dataset containing images of both healthy and damaged rose leaves was carefully picked to illustrate different disease types and stages. To analyze and identify the visual characteristics that correspond to various illnesses, we used a Convolutional Neural Network architecture, Support Vector Machine, and K-Nearest Neighbors architectures specifically intended for picture classification tasks. We address the interpretability and explainability of the model\u27s predictions in addition to performance indicators, offering insights into the decision-making process. This work addresses a fundamental requirement for effective and long-lasting disease management in rose cultivation by bridging the gap between deep learning and plant pathology. CNNs are often the preferred choice due to their ability to automatically learn relevant features from raw pixel values
A Bibliometric Analysis of Global Research Trends in Artificial Intelligence from 2019 to 2023
No full textThis bibliometric analysis examines global research trends in Artificial Intelligence (AI) from 2019 to 2023, using 7,030 Scopus indexed documents. The study found an annual growth rate of 25.93%, indicating a substantial increase in AI research effort. The majority of articles were created by collaborative teams, with an average of 4.28 authors per paper, with only 415 being single-authored. IEEE Access is the most prolific contributor, King Saud University is the leading institution, and China is the main publishing country, with 1,277 corresponding authors and the highest citation count (19,873). Thematic analysis highlights a strong emphasis on machine learning, deep learning, and neural networks as foundational topics, alongside growing interest in ethical AI and convolutional neural networks, signaling the field\u27s evolution toward addressing societal challenges and specialized applications. International collaboration plays a significant role, with 31.31% of publications involving authors from multiple countries. While the volume of AI research grows, newer articles have lower average citations due to their recent publication date. These findings highlight the interdisciplinary and worldwide nature of AI research, as well as its transformational potential for academia, industry, and policymakers. By mapping major trends and contributors, this report gives significant insights into the changing AI landscape, identifying potential for improving worldwide research collaboration and addressing growing difficulties in the field
Determining the Efficacy of Machine Learning Strategies in Quelling Cyber Security Threats: Evidence from Selected Literatures
No full textThe alarming security threats in the internet world continually raise critical concerns among individuals, organizations and governments alike. The sophistication of cyber-attacks makes it imperative for a paradigm shift from traditional approaches and measures for quelling the attacks to modern sophisticated, digital and strategic ones, such as those involving machine learning and other technologies of artificial intelligence (AI). This study is aimed at examining machine learning (ML) strategies for effective cyber security. ML involves using algorithms and statistical models to enable computers learn from and make decisions or predictions based on data. The study relied on secondary data, which were subjected to a systematic review. The results of its thematic and qualitative analyses prove that majority of the literatures allude to the fact that the maximal performance abilities and tactics of the ML constitute its strategies for quelling cyber security. These include its: early detection of threats that are tackled before they cause damages; ability to analyze huge quantity of data quickly and accurately; and processing of datasets in real-time. The study argues that the noted abilities and tactics constitute ML strategies for quelling cyber security, regardless of its challenges like data quality, security vulnerabilities and possible incidences of bias. The study concludes that ML can indeed be used to detect and respond to threats in real-time, ascertain patterns of malicious behavior, and improve on internet security, which thereby prove it to be a viable tool for quelling cyber security