Browsing by Author "Koroma, Mohamed"

Now showing 1 - 9 of 9

AI-ENHANCED INTRUSION DETECTION FOR INDUSTRY 4.0: A CROSS-REGIONAL STUDY ON MITIGATING ADVANCED PERSISTENT THREATS IN CYBER-PHYSICAL SYSTEMS
(2025-08-30) Koroma, Mohamed
This study proposes an AI-enhanced Intrusion Detection System (IDS) framework to combat Advanced Persistent Threats (APTs) in Cyber-Physical Systems(CPS)across diverse regional infrastructures. Traditional IDS struggle in resource-constrained environments, with high false positives(72% in Nigeria)and poor adaptability.TheCOVID 19 pandemicworsened vulnerabilities, leaving 68% of manufacturers without real-time OT monitoring. Our solution integrates federated learning (FL) for decentralized training, explainable AI (XAI) for interpretable alerts, and quantum-resistant cryptography for long-term security. This study tackles four challenges namely the 52% energy savings in Africa via 8-bit models, why FL maintains >90% accuracy in low-bandwidth networks, XAI boosts operator trust by 21% in Kenya, and the 96% quantum resilience. Validated across Africa (Kenya), Asia (India), and the West (USA) using real-world datasets (SWaT) and synthetic APTs, the framework achieves 93.2% detection accuracy with a 4.1% false positive rate, outperforming traditional IDS by 27% while reducing bandwidth by 62% and energy use by 42.9%. Field tests in Kenya showed a 35% increase in operator trust due to XAI transparency. Ethical safeguards include differential privacy in FL to protect sensitive data and adherence to ITU-D Ethical AI Guidelines for operator consent in field trials.
An Integrated PMT-TAM Approach to Cybersecurity Awareness: Evidence from Sierra Leone's Universities and Policy Lessons for Developing Nations
(2025-08-27) Koroma, Mohamed
This study pioneers the first empirical integration of Protection Motivation Theory (PMT) and Technology Acceptance Model (TAM) in West African academia to analyze cybersecurity awareness among 1,000 students across 10 Sierra Leonean universities (5 government, 5 private) using PLS-SEM. The novel PMT-TAM framework addresses theoretical fragmentation in behavioral cybersecurity literature by demonstrating how institutional factors (TAM) compensate for cognitive limitations (PMT) in resource-constrained settings a dynamic previously unexplored in African contexts. Results reveal moderate overall awareness (M = 3.42/5) but critical gaps, with only 12% accurately identifying phishing attempts. Key findings show self-efficacy (β = 0.58) and institutional support (β = 0.49) as the strongest predictors of protective actions, while perceived severity had weaker impact (β = 0.35). A significant "usage-awareness gap" emerged, with 89% of students using the internet daily but lacking fundamental security knowledge. Private institutions outperformed government counterparts in training quality (4.12 vs. 3.56, p < 0.01) and tool adoption (68% vs. 41% VPN usage). Theoretically, the findings advance hybrid behavioral models by revealing context-specific mediation effects (63% of perceived severity’s impact mediated by self-efficacy). Practically, the study provides actionable policy benchmarks, including allocating 3% of university IT budgets to cybersecurity and establishing national training standards. These insights are critical for developing nations undergoing rapid digitalization with limited security education infrastructure.
Determinants of E-Learning Adoption among Students in Sierra Leone's Public Tertiary Institutions: A Cross-Institutional Survey of Digital Readiness, Communal Learning Preferences, and Platform Fit
(Journal of Advanced Multidisciplinary Studies, 2026-10-07) Koroma, Mohamed
The global COVID-19 pandemic has forced Higher Education (HE) institutions to undergo a digital transformation that has reshaped their operations and made online learning platforms a necessity instead of a luxury. The adoption of e-learning is not uniform in Sierra Leone where seven public tertiary institutions have more than 40,000 students. Learners often use the resources provided for the course, but do not use interactive and collaborative resources very often. This partial adoption brings up important questions relating to how certain student behaviors are connected to the use of e-learning platforms, and why adoption levels may differ in various institutional contexts. This study investigates adoption factors of e-learning from students' perspectives in public tertiary institutions in Sierra Leone, which includes digital readiness, communal learning orientation and platform fit. The study combines existing technology acceptance factors with culturally-based constructs and context factors. The design of the study used was cross-sectional with 1100 undergraduate students from seven institutions. Hypothesized relations were tested by using partial least squares structural equation modelling. The results indicate that communal learning orientation had a significant total direct effect on e-learning adoption, and this effect was partially mediated by platform fit which accounted for 51.9% of the total effect. The results indicated that there were significant differences between the groups, with perceived usefulness being the most important at the institutions that had partial implementation of LMS, the facilitating conditions at the institutions where infrastructure was limited, and the digital readiness at the polytechnic. The study has both theoretical and practical implications, as it offers a possible mediating mechanism (platform fit), and evidence for differentiated intervention frameworks, and methodological implications by operationalizing communal learning orientation as a measurable construct.
Digital Trade Barriers: Assessing Infrastructure Deficiencies, Security Risks, and Regulatory Challenges in West Africa’s Emerging E-Commerce Markets (Ghana, Nigeria, Senegal, Liberia, Sierra Leone, Guinea)
(2025-07-11) Koroma, Mohamed
West Africa's e-commerce sector is expanding rapidly due to growing internet access, mobile adoption, and AfCFTA, yet remains constrained by infrastructure gaps, cybersecurity threats, and regulatory fragmentation. This pioneering mixed-methods study compares these barriers across six nations (Ghana, Nigeria, Senegal, Liberia, Sierra Leone, Guinea), integrating World Bank/ITU datasets with regulator interviews (n=12) and SME surveys (n=150). The analysis reveals severe urban-rural ICT disparities (>39 percentage points) limiting SME participation, Nigeria's disproportionate cybercrime burden (58% of regional losses, \$500M annually), and costly regulatory misalignment (15-30% of revenues). These findings empirically validate van Dijk's (2020) digital divide theory and Williamson's (1981) transaction cost economics in West Africa's digital markets. Urgent interventions include targeted ICT investments to achieve 52% penetration thresholds, an ECOWAS Cybercrime Task Force (projected 22-28% fraud reduction), and AfCFTA protocol harmonization adopting Ghana's hybrid governance model (\$140M annual compliance savings). These measures directly address implementation gaps in regional digital trade. The study calls for longitudinal tracking via a West African Digital Trade Index and AI-driven customs solutions to sustain progress. These evidence-based solutions balance immediate action with long-term innovation for inclusive digital growth.
Evaluating the Adoption of Virtualization and Cloud Computing as Strategic IT Outsourcing Solutions for Enhancing Educational Efficiency at Njala University
(2025) Koroma, Mohamed
The rapid evolution of virtualization and cloud computing presents transformative potential for higher education institutions (HEIs), particularly in resourceconstrained settings like Sierra Leone. Despite their proven benefits in developed nations, significant gaps remain in understanding their adoption challenges and impacts in African HEIs. This study evaluates the implementation of these technologies at Njala University, addressing critical research gaps in IT outsourcing strategies for developing regions. Employing a mixedmethods approach, the research combines quantitative surveys (N = 1,000) and qualitative interviews (n = 15) to assess three key hypotheses: the impact on IT resource management (H₁), operational efficiency (H₂), and educational outcomes (H₃). Findings reveal strong correlations (r = 0.805–0.939, p < 0.001), confirming significant improvements in all areas, with 87.6% of respondents reporting enhanced IT management and 76.7% observing better educational efficiency. However, infrastructural deficits (e.g., unreliable internet) and skill gaps emerged as critical moderators. The study validates the Technology-Organization-Environment (TOE) framework while proposing "infrastructure viability" as a new construct for developing contexts. Practical implications include prioritizing bandwidth upgrades, capacity building, and localized cloud policies. Despite its contributions, the single-institution focus limits generalizability, suggesting the need for cross-national studies. Future research should explore longitudinal costbenefit analyses and policy frameworks to support scalable adoption. This work provides actionable insights for HEIs and policymakers, bridging the gap between theoretical models and real-world implementation in resourceconstrained environments.
Optimizing IoT and IIoT Intrusion Detection: A Comparative Analysis of Ensemble and Deep Learning Techniques
(European Organization for Nuclear Research, 2026-01-13) Koroma, Mohamed
The use of Internet of Things (IoT) and Industrial Internet of Things (IIoT) has greatly enlarged the attack surface of the cyberspace, exposing the key systems to risks. This is a threat environment that requires smart intrusion detection systems (IDS) that overcome the shortcomings of traditional and signature security programmes. Despite the potential of machine learning (ML) to be the best option of IDS, the comparison of the modern algorithms to the problem of IoT specifics is not yet conclusive. Neither a resource-limited hardware nor a high-accuracy ideal model has been uncovered to provide both high-accuracy and operational efficiency. The proposed research is aimed at identifying the best ML model to use in the detection of IoT/IIoT intrusions. It aims to: i) evaluate seven different algorithms including tree-based ensemble and deep learning networks; iii) preprocess the Edge-IIoTset dataset with SMOTE to address the problem of class imbalance; iii) benchmark the performance involving standard metrics; and iii) discover the most important features in detection of threats. The research is conducted in a systematic experimental manner with the use of the Edge-IIoTset dataset. Using SMOTE to preprocess, seven models are trained and assessed based on accuracy, precision, and F1-score: Logistic Regression, Random Forest, XGBoost, CatBoost, LightGBM, ANN and CNN. An analysis of the most performing features is then carried out. Gradient boosting models were the best models to use, with XGBoost having the highest accuracy (95.07%), precision (95.46%), and F1-score (95.14%). The TCP flags and application-layer data were found to be the major intrusion indicators analysed. XGBoost and LightGBM offered the most desirable compromise between the strength and power of detection and computational efficiency. The research establishes the use of gradient boosting algorithms namely XGBoost and LightGBM are very effective in addressing IDS in IoT/IIoT setting, best combining high-performance, efficiency and interpretability. The findings provide practitioners with a clear evidence-based guideline on how to choose IDS models to provide them with effective cybersecurity without having to overstretch their limited processing resources. The next work should confirm these models in test-beds in real-time and investigate the construction of lightweight hybrid systems in new areas such as 5G and vehicular networks.
Seroprevalence of HBsAg among Female Patients Seeking Healthcare in Bo Government Hospital, Southern Sierra Leone: 14-Months Database Study
(Sciencedomain International, 2014-01-10) Koroma, Mohamed
**Background:** Chronic hepatitis B virus (HBV) seroprevalence varies globally but is positively correlated with the prevalence of HIV since both have the same route of transmission-contact with infected body fluids. Sierra Leone has a nascent hepatitis B epidemic and there is a paucity of research data and public health awareness about the infection. Bo where this research is done is the second largest city of Sierra Leone and is the regional headquarters of southern Sierra Leone. **Aim:** To determine the seroprevalence of hepatitis B surface antigen (HBsAg) among female healthcare seekers who used the main government referral hospital in Bo, southern Sierra Leone. **Methodology:** We retrospectively analysed the serological diagnostic results of HBsAg for female healthcare seekers at the Bo Government Hospital. We later determined the HBsAg seroprevalence for various categories of the study subjects; pregnant women, lactating mothers, girls 5-18years and adults >18years for the period under review. Place and Duration of Study: This study is a multicenter investigation conducted at the Bo Government Hospital laboratory, the Departments of Community Health and Clinical Sciences and Environmental Health Sciences at Njala University in Bo southern Sierra Leone. All laboratory diagnostic tests for HBsAg diagnosis were conducted at the Bo Government Hospital laboratories from May 2012 to June 2013. Literature review and data analysis were done at Departments of Community Health and Clinical Sciences and Environmental Health Sciences at Njala University Study Subjects: A total of 2,218 female subjects took part in this investigation; girls 5-18yrs (15.2%), adults females >18yrs (48.9%), pregnant women (17.4%), and lactating women (18.3%). Study subjects were between 5-45yrs; mean age 26yrs; HBsAg+ mean age was 24yrs; HBsAg- mean age was 29yrs. **Results:** One thousand and fifty four (47.5%) of all females who sought healthcare service at the Bo Government Hospital laboratory during the period under review were tested positive for HBsAg; 159(7.1%) pregnant women, 168 (7.5%) lactating mothers, 154 (6.9%) were girls 5-18yrs and 573 (25.8%) were >18yrs. Forty-five percent (154/339) of all girls between 5-18years who took part in this study were HBV seropositive. Discussions: Few independent investigations in Sierra Leone have estimated the national HBsAg seroprevalence at 2.6%-5.1%. One import of this study is the high 154/2218(6.9%) HBsAg seroprevalence rate among girls 5-18 years which can be attributed to multiple sex partners, unsafe sex practice and other socioeconomic and behavioral factors that might have been prevailing during the period under review. The generally high HBsAg seroprevalence rate in our study can also be as a result of our large cohort size and the low sensitivity and specificity of the Rapid Determine diagnostic test kits used in our study compared to previous similar studies conducted in Sierra Leone. **Conclusions:** The high seroprevalence rate of HBsAg among girl between 5-18years in this study emphasizes the need for a robust public health preventive and control measures for HBsAg including HBsAg vaccination, routine blood screening, and safe sex practice as a means of reducing the spread and prevalence of HBV infection in Bo.
Technological Advancements of Energy Storage Systems Technologies in Africa: A Review
(Scientific Research Publishing, 2025) Koroma, Mohamed
The global energy landscape is transforming to tackle climate change, improve energy security, and meet growing energy demands. Energy storage technologies are vital for incorporating “renewable energy”, stabilizing electrical network, and advancing electrification. This review paper provides a comprehensive analysis of the technological advancements in energy storage systems (ESS) and their applicability in Africa. The study highlights the potential of ESS to address Africa’s energy challenges, including grid instability, rural electrification, and renewable energy integration. The paper critically evaluates various ESS technologies, such as lithium-ion batteries, pumped hydro storage, and flywheels, and assesses their economic, environmental, and technical feasibility in different African regions. The findings suggest that tailored ESS solutions, supported by innovative financing models and regional-specific policies, are essential for sustainable energy transitions in Africa. The paper concludes with recommendations for policymakers, researchers, and industry stakeholders to accelerate the adoption of ESS technologies across the continent.
User-Centric Evaluation and Optimization of Resource Allocation in Edge Computing Environments
(2025) Koroma, Mohamed
The proliferation of mobile applications with intensive computational demands has necessitated the adoption of edge computing to reduce latency and energy consumption. However, edge servers face challenges such as limited resources, dynamic wireless conditions, and inefficient task offloading strategies, particularly in multiuser environments.