Journal of Information Technology, Cybersecurity, and Artificial Intelligence
Not a member yet
25 research outputs found
Sort by
Comprehensive Analysis and Detection of IoT Network Attacks Using Recon Host Discovery Traffic Dataset
Get PDFThe proliferation of Internet of Things (IoT) devices has introduced unparalleled interconnectivity and significant security challenges. Reconnaissance attacks, particularly Host Discovery, are often precursors to more severe cyber threats. In this study, we examine a labeled network traffic flow dataset to analyze patterns and identify key indicators of Recon Host Discovery attacks. Leveraging exploratory data analysis and feature correlation techniques, we uncover critical traffic behaviors, such as short flow durations and anomalous packet statistics, that distinguish benign from malicious activities. The findings lay the groundwork for developing robust detection mechanisms for IoT networks, emphasizing the importance of targeted feature selection and real-time analytics
Named Entity Recognition for Hindi Current Landscape and Emerging Trends
Get PDFNamed Entity Recognition (NER) plays a crucial role in Natural Language Processing (NLP) by automatically identifying and classifying entities such as names of people, places, organizations, dates, and numerical values within unstructured text. While NER has seen major advancements in resource-rich languages like English, building robust NER systems for Indian languages—particularly Hindi—remains a significant challenge. Hindi presents unique linguistic complexities such as rich morphology, free word order, the absence of capitalization cues, and widespread use of code-mixed text, which complicate the task further. Over the years, researchers have explored a wide range of approaches to address these challenges, starting with rule-based and statistical models and progressing to sophisticated deep learning and transformer-based techniques. Multilingual models like mBERT, IndicBERT, and MuRIL have shown promise in improving accuracy and generalizability for Hindi NER. This review offers an in-depth look at the current state of Hindi NER, including the available annotated datasets, computational models, and performance benchmarks. It highlights the gaps that persist, such as the scarcity of high-quality annotated data, difficulties in handling informal and domain-specific language, and limited adaptability across different text types. The paper also outlines future directions for research, emphasizing the need for low-resource learning strategies, domain adaptation, and better handling of noisy and code-mixed data. As Hindi continues to dominate communication in various digital spaces, advancing NER systems for this language is more relevant than ever
Legal regulation of industrial robots
Get PDFThis study examines the legal nature of industrial robots in light of the rapid progress in artificial intelligence technologies and the growing reliance on such systems in various aspects of life. It raises new legal and ethical questions about the extent to which industrial robots can be granted legal personality, and the forms and nature of such personality. The study discusses this concept as a legal framework that permits attributing certain rights or obligations to robots, depending on their use and degree of functional independence from humans. The importance of this discussion becomes evident with the emergence of advanced applications of industrial robots, such as self-driving cars, production-line robots, chatbots, and algorithms capable of making impactful decisions that may cause material or moral harm without direct human intervention. The study further investigates whether the legal system needs to adopt a special model of legal personality for these entities to enable accountability for their actions or the outcomes of their decisions.Using the descriptive-analytical approach, the study centered on the following question: What is the legal nature of industrial robots? The study concludes that granting robots a special or conditional legal personality may help fill the legislative gap concerning liability, particularly in cases where tracing the direct human actor proves difficult. The findings emphasize the necessity of establishing gradual and balanced legal frameworks through which industrial robots can be granted limited and well-regulated legal personality. Such frameworks must both account for technological advancement and ensure effective human oversight. The study also calls for international cooperation to formulate unified legal standards consistent with legal and human values
Hybrid Approach with a focus on Preprocessing Techniques for Detecting Phishing Websites
Get PDFPhishing is regarded as a significant cybersecurity problem in the digital era, utilizing the fabrication of fraudulent websites to deceive users and expropriate their sensitive information, including passwords and financial data. The growing dependence on the internet has led to a marked increase in the frequency of these attacks, resulting in considerable financial losses for individuals and businesses alike. This underscores the pressing necessity for efficient strategies to counteract such assaults.This research aims to create a hybrid model for identifying phishing websites via URL analysis. The suggested model combines Convolutional Neural Networks (CNN) with Long Short-Term Memory networks (LSTM) and an Attention Mechanism to make predictions more accurate and uncover hidden patterns in the data. The model was trained on the "Url_Detection_Dataset" from the Kaggle platform, and its performance was assessed using precision, recall, and F1-score measures. The results showed that the hybrid model is better than traditional methods at telling apart real and harmful URLs, making it a useful tool in cybersecurity. The results provide a framework for subsequent research and promote the creation of more resilient, flexible, and effective solutions
Improving Public Service Grievance Analysis: A Comparative Study of Topic Modelling Techniques with a Multi-Metric Data Cleaning Framework
Get PDFThe SewaSetu portal, a single window system for government services in Assam, India, processes thousands of applications and hundreds of grievances daily. And many such government grievance portals routinely receive a substantial volume of public complaints, each containing valuable information but often embedded in unstructured text. Extracting patterns from such data can enable public agencies to respond more efficiently and allocate resources more strategically. Manual classification of these grievances is a time-consuming bottleneck. This paper describes a methodology for reliable topic discovery in this noisy domain. Unlike standard studies that rely solely on stopword removal, this paper introduces a robust ”Multi-Metric Gibberish Filtering Pipeline to ensure that the subsequent dataset was free of incoherent noise. We then proceeded to perform a comprehensive coherence benchmarking of four primary topic modeling algorithms, Latent Semantic Analysis(LSA), Non-Negative Matrix Factorization (NMF), Latent Dirichlet Allocation(LDA), and BERTopic across varied topic counts (K=5 to 50) on the cleaned grievance data. Analysis through the Cv (coherence score) showed that NMF considerably outperformed the alternatives, reaching the highest overall score of 0.7898 at K=35. This work sets a benchmark for preparing data to handle noisy government feedback and asserts the NMF with 35-topic configuration as the most effective and coherent model in extracting interpretable themes from public service delivery grievances.
A Gray Image Quantum Encryption using GNEQR Representation
Get PDFIn the digital era, characterized by extensive online data exchange, information security has become a priority. While traditional encryption methods have proven effective in protecting data transfers, the advent of advanced quantum computing has increased susceptibility to security breaches. Quantum encryption provides a revolutionary solution to this problem by using quantum mechanics principles to establish algorithms that are impermeable to decryption. Using these quantum properties, cryptographic protocols are developed to provide superior security, unlike traditional encryption methods. The image plays an important role in transmitting information in all areas. Therefore, quantum image encryption methods are specifically designed to counter the potential risks posed by quantum computers, which can compromise conventional encryption protocols. This ensures the preservation of data security despite advances in quantum computing technology. In addition, quantum image encryption improves data transmission efficiency by establishing secure communication channels using quantum stats, thereby reducing the need for bandwidth and improving transmission speed. This paper proposes a new method of quantum encryption based on GNEQR representation and the modification of pixel values and positions in an image. After converting the image into a quantum form, we applied an algorithm to modify the values and positions of the pixels using a succession of quantum gates. We concluded this study with a statistical analysis showing the robustness of our quantum image encryption method
The Impact of Artificial Intelligence on Business Strategy: Redefining Competitive Advantage in the Digital Era
Get PDFArtificial Intelligence has become a disruptive force that essentially reinvents business strategy in all industries across the world. This paper will examine how AI technologies can be used to formulate corporate strategy, defining six new sources of competitive advantage and will critically examine the transformation that is necessary in the organization, the mechanisms of accelerating innovation, and how to improve customer experience. The study utilizes both qualitative and quantitative methods, which included bibliometric review of 1,039 articles and systematic review of 180 articles, financial performance data of Fortune 500 corporations, and 28 case studies in the industry. Based on the framework of analysis, the synthesis of the Resource-Based View, Dynamic Capabilities Framework, and Technology Acceptance Model are used to evaluate the strategic implications of AI. Findings indicate that three-quarters of organizations use AI in one or more business operations with a potential economic impact of US 4.4 trillion a year. The overall shareholder return premium of 10.7 percentage points was realized by firms that achieved competitive advantages in all six of the areas of AI capabilities identified. However, the percentage of firms that reached a mature AI capability was only 1 percent, and 42 percent give up because of difficulties in data preparedness, skills shortage, and integration issues. Organizational redesign is required to achieve successful AI adoption, and workflow reconfiguration is the best indicator of business impact. Although AI adoption is accelerating at a rapid pace, the realization of strategic value demands fundamental organizational transformation rather than superficial technological overlays
Enhancing Temperature and Rainfall Prediction Accuracy Through Deep Learning Frameworks
Get PDFThe information of accurate forecasting on temperature, rainfall is also very crucial for disaster preparedness, as well for climate management. Typical statistical and machine learning approaches have limited ability to capture nonlinear and spatiotemporally varying structure of climate fields. This research utilized recent state-of-the-art deep learning models to improve the prediction models for both temperature and rainfall. The hybrid Convolutional Neural Networks and Long Short-Term Memory (CNN–LSTM) method achieved the best results (R² = 0.98 for temperature, 0.91 for rainfall), outperforming those of Multiple Linear Regression (MLR) and Random Forest (RF) as a traditional model. The Physics-Informed Neural Network (PINN) model delivered physically consistent and stable predictions, especially under extreme weather such as heavy rainfall or heatwaves. Relative humidity, atmospheric pressure and sea surface temperature were found as most important predictors-base on feature importance analysis. The regional analysis remained that the coastal region performed best, whereas the hilly region with the high topographical complexity presented a relatively lower accuracy. In general, embedding deep learning into physical constraints ended up improving a lot both correctness and robustness of predictions. Further work should be carried out to improve interpretability, inclusiveness of data and transferability in space of such models with the ambition to build a more sustainable real-time weather forecasting system
The Impact of Artificial Intelligence on Enhancing Guest Experience of Hotel Industry in Kuala Lumpur, Malaysia
Get PDFUtilizing the Technology Acceptance Model (TAM) as its theoretical backbone, this research examines AI integration into hotel guest experiences of Kuala Lumpur. The study explores relationships between PU, PEOU, Attitude Toward Using AI, and Guest Satisfaction (GUSA), investigating AUAI as a mediator and examining the moderating role of guest age, previous AI experience and trip purpose. Through the empirical survey of 630 hotel customers and analyzing the data with structure equation model (SEM), this study finds that PUSE and PEUS have significant impacts on AUAI and thus indirectly affect GUSA; meanwhile, PUSE and PEUS are directly influencing customer satisfaction simultaneously so that there are parallel direct and indirect effects. Adding to the TAM-strain in hospitality research, this result places post-adoption satisfaction as a primary outcome. Disparities between demographic and experience segments imply specificity in implementing AI. The findings have broader implications in both theoretical and practical ways: theoretically, it furthers the understanding of AI adoption within services context by incorporating technology acceptance with satisfaction outcomes; practically, it provides hoteliers and developers of AI technologies with practical steps toward creating solutions that are useful, easy to use, and constructs favorable guest attitudes insights valuable for strategically fast growing urban destinations such as Kuala Lumpur
MACHINE LEARNING (ML) TO EVALUATE GOVERNANCE, RISK, AND COMPLIANCE (GRC) RISKS ASSOCIATED WITH LARGE LANGUAGE MODELS (LLMs)
Get PDFIn today’s AI-driven digital world, Governance, Risk, and Compliance (GRC) has become vital for organizations as they leverage AI technologies to drive business success and resilience. GRC represents a strategic approach that helps organization using Large Language Models (LLMs) automation tasks and enhances customer service, while maintaining the regulatory complexity across various industries and regions. This paper explores a machine learning approach to evaluate Governance, Risk, and Compliance (GRC) risks associated with Large Language Models (LLMs). It utilizes Azure OpenAI Service logs to construct a representative dataset, with key features including response_time_ms, model_type, temperature, tokens_used, is_logged, data_sensitivity, compliance_flag, bias_score, and toxicity_score. These features are used to train a model that predicts GRC risk levels in LLM interactions, enabling organizations to improve efficiency, foster innovation, and deliver customer value, while maintaining compliance and regulatory requirements