Thesis Title: Student-Centered ICT Technical Support and Infrastructure Provision Architecture based on AI-Powered Virtual Assistant

Student’s Name: Paul Oduor Oyile

Supervisors Names:

1. Dr. Anselemo Peters Ikoha
2. Dr. Roselida Maroko Ongare

Abstract:

The essence of a learning institution is to deliver its core services to clients in the best way possible which is done conventionally in most countries. Some of the conventional approaches used in teaching and learning failed amid COVID-19, making most of the learning institutions to suspend physical learning and resort to online platforms which required intensive use of ICT. The use of ICT to deliver services to clients is accompanied with various technical challenges that require frequent attention some of which are beyond the student’s technical know-how. The challenge therefore was difficulty in assuring the consistent availability and efficiency of ICT technical support to address these technical challenges experienced by students owing the ICT staff-to-students requests ratio as well as in the absentia of ICT staff. The purpose of this study was to investigate the ICT Technical support services requested by students, ICT infrastructure requirements, and develop a student-centered technical support and infrastructure provision architecture based on AI-powered virtual assistant for students in public universities in Kenya. This was guided by the following objectives: establish the requested ICT technical support services by students, analyze the ICT infrastructure requirement for the implementation of ICT technical support virtual assistant in public universities in Kenya, and design a student-centered ICT technical support and infrastructure provision architecture based on AI-powered virtual assistant. The study used design science research design. The sample unit of the study included public university where the students and ICT staff were targeted. Kibabii University was selected as a test environment. content analysis, questionnaires, focus groups, and guided interviews were used for data collection from which the architecture was developed. The sample size for this study was established using Yamane’s formula to be 389 students. 16 staff were purposively sampled to provide data that was to complement the data from content analysis. Quality control was ascertained by ensuring the validity (Content and Face Validity) and reliability (pilot study at MMUST) of the research instruments. Ethical standards were adhered to. The data collected was subjected to both descriptive (mean, frequency and Percentages) and inferential statistics (Factor Analysis). The major findings were as follows: ELearning Services, Online Application and Clearance Services and Email Services Technical Support were the commonly requested ICT technical support services by students in public universities in Kenya. Smartphone, Personal Computer, server and internet routers were the key hardware requirements for the implementation of a student-centered ICT technical support provision virtual assistant. Application Programming Interface (API), Database Management System (DBMS), Operating System (OS) and Integrated Development Environment (IDE) were the key software requirements; JavaScript and Python were viewed by this study to be most appropriate programming languages for developing AI-powered virtual assistants; Google’s Dialogflow and IBM’s Watson were found to be most appropriate natural language processing platforms for chatbot development and  WhatsApp was the most preferred messaging platform for testing and deployment of the AI-powered virtual assistants (chatbot). The study used the requirements gathered in objective 1 and 2 to develop a Student-Centered ICT Technical Support Provision Architecture (SCITSPA). The architecture was presented to 12 experts and 10 students for domain expert evaluation and prototype evaluation respectively. SCITSPA architecture shall form: a basis for implementing a student-centered ICT technical support service provision in public universities in Kenya; a basis for further research in student-centered ICT technical support service provision, and a blueprint for automation of ICT technical support provision in public universities through virtual assistants by software development firms.