FACULTY OF COMPUTING

Hepatitis is a major global health issue, impacting millions worldwide. Early detection and precise diagnosis are essential for effective treatment and preventing complications. This study proposes a predictive model for hepatitis diagnosis using ensemble feature selection techniques and Adaptive Neuro-Fuzzy Inference Systems (ANFIS). The goal is to develop a robust and accurate model to assist healthcare professionals in promptly diagnosing hepatitis cases. The methodology follows the CRISP-DM framework, including data collection, preprocessing, ensemble feature selection, ANFIS model development, and evaluation using performance metrics like accuracy, precision, recall, and F1-score. Publicly available hepatitis datasets are utilized for experimentation and validation. The results demonstrate that ensemble feature selection effectively identifies informative hepatitis diagnosis features, enhancing the model's predictive performance. The ANFIS model, trained on the selected features, achieves high accuracy and balanced performance metrics, indicating its effectiveness in accurately diagnosing hepatitis cases. The implications include early hepatitis detection, healthcare resource optimization, and potential for personalized medicine. Future directions involve integrating additional data sources, continuous model refinement, and clinical validation studies. This study contributes to advancing predictive modeling for hepatitis diagnosis and highlights the importance of interdisciplinary collaboration between healthcare and data science disciplines in addressing global health challenges.

This research examined the extent of A Quasi-Experimental Study on the Effect of Informatics in Teaching and Learning in Universities in Edo State, with particular emphasis on differences across federal, state, and private institutions. The study adopted a descriptive survey research design and was guided by the Technology Acceptance Model (TAM) and the Unified Theory of Acceptance and Use of Technology (UTAUT). Data were collected from 345 lecturers and students using a structured questionnaire designed around key constructs such as ICT infrastructure availability, utilization, digital competency, and perceived challenges. Data analysis involved descriptive statistics, Analysis of Variance (ANOVA), correlation, and regression analysis using SPSS. Findings revealed that ICT infrastructure and utilization levels were generally high, with private universities demonstrating significantly higher adoption compared to public institutions. Gender did not significantly influence ICT utilization, while institutional type and academic role (lecturer or student) were significant determinants. Regression results further confirmed that ICT utilization positively and significantly enhances teaching and learning effectiveness. The study concludes that institutional readiness, infrastructural support, and digital competency are critical to effective ICT integration in higher education. It recommends increased investment in ICT infrastructure, continuous capacity building, and supportive institutional policies to enhance technology driven teaching and learning.

Fisheries and aquaculture are vital to global food security but face critical challenges, including security threats from unauthorized access, the limitations of reactive manual monitoring, and uncertainty in long-term production forecasting. Existing management tools often lack a unified approach to integrate real-time operational surveillance with predictive decision-support analytics. This study aims to address these gaps by designing and developing a hybrid methodological framework that integrates computer vision for security and advanced machine learning and time-series forecasting for production analysis and stock management. The study employed a dual-component methodology. For security, an OpenCV-based Histogram of Oriented Gradients (HOG) descriptor combined with a Support Vector Machine (SVM) was implemented to achieve automated, real-time human intrusion detection and alarm triggering. For predictive analytics, a hybrid approach was utilized, applying Seasonal Autoregressive Integrated Moving Average with Exogenous Regressors (SARIMAX) for long-term production forecasting and various ensemble classifiers (Random Forest, Gradient Boosting, K-Nearest Neighbors) to analyze fish survivability and production trends. Results demonstrate that this integrated framework is highly effective across operational and strategic levels. The computer vision system successfully provided continuous, automated surveillance with verifiable digital evidence. The ensemble classifiers achieved near-perfect performance in survivability prediction (e.g., Random Forest Accuracy = 0.98, F1-score = 0.99, AUC ≈ 0.999). Furthermore, the SARIMAX model effectively projected a steady increase in global fish production, forecasting growth from 232.19 million tonnes in 2024 to 272.12 million tonnes by 2043—a 17.19% increase—capturing long-term temporal dynamics that static regressors fail to model. Based on these findings, it is recommended that fisheries stakeholders prioritize the deployment of computer vision-based surveillance to reduce security risks and labor costs. Additionally, management authorities should integrate ensemble classifiers into stock-management systems to optimize survivability and adopt SARIMAX-based forecasting for robust, evidence-based policy planning. The implementation of these data-driven platforms will significantly enhance sustainable fisheries governance and support proactive resource allocation.

The procurement of motor-vehicle spare parts through traditional physical markets presents challenges such as difficulty in identifying compatible parts, lack of price transparency, geographic and time constraints, and absence of centralized vendor platforms. This study focused on the design and development of a web-based online marketplace for motor-vehicle spare parts to address these limitations by providing a centralized, secure, and user-friendly platform connecting vendors and customers. The research adopted the Agile methodology using the Scrum framework and implemented a layered system architecture consisting of presentation, application, data access, and data storage layers. The system was developed using PHP for server-side processing, MySQL for database management, and HTML5, CSS3, JavaScript, and the Bootstrap framework for responsive user interface design. Selected spare-parts markets in Benin City, Edo State—Evbareke Market, Uwelu Market, and Igun Market were used as case studies to understand vendor operations and customer needs. The study objectives included designing a marketplace that supports product listing and searching based on vehicle make and model, implementing secure user authentication for buyers and vendors, enabling price comparison and vendor rating features, and testing the system to ensure reliable operation. The implemented system successfully delivered a functional multi-vendor marketplace with advanced search and filtering, vehicle compatibility matching, secure shopping cart and checkout processes, order management, inventory control, and role-based access control. Results demonstrated that the system effectively addresses problems associated with traditional spare- parts procurement by improving accessibility, transparency, and efficiency. Challenges such as database optimization, security implementation, responsive design, and system integration were encountered and resolved. The study concludes that web-based marketplace technologies can effectively transform spare-parts trading and support digital transformation in the automotive industry.

Vehicle identification and verification play a critical role in ensuring the security of controlled environments such as university campuses. At the University of Benin’s Ugbowo Campus, the existing manual vehicle verification process suffers from several limitations, including delays, human error, poor record keeping, and the absence of real-time validation. This study addresses these challenges by designing and implementing an Automated Car Plate Number Recognition (ANPR) system capable of detecting, recognizing, and verifying vehicle license plates using computer vision and optical character recognition (OCR) techniques. The system was developed using the Flask web framework, OpenCV for image preprocessing and license plate localization, and the Tesseract OCR engine for extracting alphanumeric characters. A MySQL database was integrated for storing records of registered vehicles and maintaining logs of recognition events. The system allows security personnel to upload vehicle images through a web interface, after which the application automatically processes the image, recognizes the plate number, checks its validity against the stored records, and provides immediate feedback. Testing of the system using images captured under different conditions such as clear daytime lighting, low light, skewed angles, and motion blur revealed that the system performs accurately under favourable conditions and provides partial or usable information even under challenging scenarios. Although recognition performance decreases with poor image quality, the system still offers considerable improvements over manual verification by reducing reliance on human judgment and enhancing the accuracy, speed, and reliability of vehicle monitoring.

In the 21st century, digital technology has become deeply embedded in nearly every aspect of daily life from communication and commerce to healthcare and education. While younger generations have adapted rapidly to this digital transformation, many elderly individuals remain on the margins of the digital world. This generational divide has created a significant barrier to inclusion, autonomy and access to essential services for older adults. The inability to navigate digital platforms not only limits their participation in modern society but also exacerbates feelings of isolation, dependence and vulnerability. This project investigates the multifaceted challenges that elderly populations face in adopting and using digital technologies. These challenges include physical limitations such as impaired vision and motor skills, cognitive barriers like memory decline and unfamiliarity with digital interfaces and socio-economic factors such as lack of access to devices or internet connectivity. The study also explores the psychological impacts of digital exclusion, including reduced self-esteem, social withdrawal and increased mental health risks. The research further examines how emerging technologies such as voice activated assistants, simplified user interfaces and wearable health devices can be tailored to meet the needs of older adults. It also considers the role of caregivers, family members and community organizations in supporting digital adoption among seniors. Ultimately, this project argues that closing the digital generational gap is not merely a technical challenge but a social imperative. By fostering inclusive design, empathetic education and supportive ecosystems, society can empower elderly individuals to engage meaningfully with technology. Through inclusive design and community driven support, this project envisions a future where older adults actively participate in the digital world, fostering equity and stronger bonds across generations.

Efficient complaint management is essential for enhancing student satisfaction and institutional accountability in higher education. However, most existing systems lack intelligence, scalability, and analytical depth. This dissertation presents the design and implementation of an AI-driven Student Complaints Management System (SCMS) that integrates Natural Language Processing (NLP) and sentiment analysis to improve how student grievances are captured, categorized, and resolved. The system applies text-mining techniques to interpret complaint narratives, classify issues by department, and assess emotional tone to prioritize responses. A modular web-based architecture enables real-time reporting, trend visualization, and administrative insights. Using a design science research methodology, the study demonstrates that intelligent automation significantly reduces resolution time and enhances decision-making transparency. The proposed SCMS contributes to the field of educational technology by merging human-computer interaction with data analytics to foster responsive governance and continuous improvement in higher education service delivery

This project presents a detailed comparative study of six prominent AI-powered chatbots: DeepSeek, ChatGPT, Poe, DeepAI, Jasper, and Rasa. As organizations increasingly integrate chatbot technology to enhance customer service and user engagement, understanding the diverse capabilities of these tools becomes essential. This study aims to systematically evaluate the performance of each chatbot across multiple parameters, including functionality, accuracy, user satisfaction, task completion rates, and cost- effectiveness. To facilitate this comparison, we employ a weighted scoring model that assigns relative importance to each parameter based on organizational needs. This model allows for a nuanced analysis, accommodating the various strengths and weaknesses exhibited by each chatbot. For instance, while ChatGPT is recognized for its versatile conversational abilities and high user satisfaction, DeepSeek excels in specialized information retrieval. Similarly, Jasper focuses on content generation, whereas Rasa offers flexibility for developers in building contextual AI assistants. Data for this study is collected through user surveys, performance testing, and analysis of chatbot logs. The results are presented in a comprehensive format, providing both quantitative scores and qualitative insights. By the end of this project, we aim to deliver actionable recommendations for organizations, guiding them in selecting the most appropriate chatbot solution tailored to their specific requirements. Ultimately, this study contributes to the growing body of knowledge in AI chatbot technology, facilitating informed decision-making in an increasingly competitive digital landscape.

Polycystic Ovary Syndrome (PCOS) is a chronic endocrine disorder requiring continuous monitoring to manage its complex and irregular symptoms. Current digital health applications often fail to address this need, providing generalized tracking, inadequate visualization, and insufficient privacy protection. This project presents the design and development of a specialized, Python-based symptom tracking and visualization system to empower women with PCOS. Built using an Object- Oriented Analysis and Design (OOAD) methodology, the system ensures a modular and user- centered architecture. It leverages Matplotlib for generating insightful visualizations of symptoms, mood, and lifestyle factors. A Flask-based web interface offers an intuitive dashboard for data entry and review, with a commitment to user privacy through local data storage. The resulting system demonstrates that open-source Python technologies can effectively bridge gaps in women's digital health by providing an affordable, transparent, and adaptable solution. It facilitates improved health literacy, data-driven self-management, and enhances patient-clinician communication, highlighting the significant potential of Python in advancing personalized chronic disease management.

Credit risk assessment is a critical process in lending institutions aimed at evaluating the likelihood of loan default by applicants. Traditional manual assessment methods are often inconsistent and time-consuming, while many automated credit scoring systems lack transparency and explainability. This study presents the design and implementation of a web- based credit risk assessment system . The proposed system captures key applicant attributes and applies predefined decision rules to compute credit risk scores. Debt-to-income ratio and other affordability indicators are used to classify applicants into low, medium, or high risk categories. The system generates clear and interpretable credit decisions accompanied by explanatory reasons. A structured system analysis and design methodology was adopted to guide development. The system was implemented using web technologies and deployed as a browser- based prototype. Functional testing was conducted using representative test cases. Test results confirmed that the system produces accurate, consistent, and explainable credit decisions. The system enhances transparency, usability, and decision accountability. The study demonstrates that explainable rule-based models can effectively support automated credit risk assessment in small and medium-scale lending environments.