The Problem
Students at the University of Batangas were relying on generic career guidance systems that did not capture individual interests, strengths, or goals. The goal was to create a more personal, data-informed, and supportive recommendation tool.
Objectives
- Analyze AI, machine learning, text-to-speech, database systems, and data analytics to create a comprehensive recommendation platform.
- Develop personalized program matching through validated questionnaires, academic insights, and faculty information profiles.
- Reduce program shifting by improving early academic decision-making and satisfaction.
- Evaluate the system using ISO/IEC 25010 standards across functionality, usability, reliability, efficiency, maintainability, and compatibility.
Concept & Process
I designed the BELONG framework, which categorized students into six personality types — Builder, Explorer, Leader, Organizer, Nurturer, and Generator — and used validated assessments across interest, strength, and career goal. This grounded the experience in personal insight instead of generic recommendations. I planned the development using the Agile methodology to guide the project from planning and requirements analysis through design, coding, testing, deployment, and maintenance.
System Architecture
I mapped system journeys with flowcharts, context diagrams, and top-down data-flow diagrams, then created low-fidelity and high-fidelity wireframes for the landing page and questionnaire interface using a cosmic design theme inspired by the human mind.
Results & Testing
The recommendation interface presented each student with their top three program matches, supported by AI-generated explanations and a validation accuracy of 89.78%. Students reported greater confidence in their decision-making and reduced anxiety when selecting a path.
Overall Rating
3.63/5.00
Usability
3.89/5
Reliability
3.76/5
Functionality
3.74/5