Higher Education Category Submission

947 – IntelliTest

IntelliTest – Adaptive AI-Powered Assessment Generator is an innovative educational platform designed to transform how assessments are created, personalized, and used for learning reinforcement. It allows students and educators to upload study materials such as PDFs and PPTs, from which key topics and concepts are intelligently extracted. Based on these materials, users can customize assessments by selecting difficulty level, question type (MCQ, subjective, numerical), and targeted cognitive skills aligned with Bloom’s Taxonomy (Remember, Understand, Apply, Analyze, Evaluate, Create).

Technically, the platform processes the uploaded documents using text extraction and summarization techniques to build a structured knowledge base. Open-source large language models (LLMs) such as GPT-Neo are then guided through engineered prompts to generate context-specific, high-quality questions directly from the uploaded content. This ensures that the assessments are highly relevant and aligned to the user's actual learning resources.

After test completion, IntelliTest evaluates student performance across different cognitive skill levels and topics. It then automatically generates detailed solutions for each question and analyzes strengths and weaknesses. Based on the performance analysis, the platform provides a personalized study recommendation plan, suggesting targeted revision strategies and practice exercises to help the learner improve in specific areas.

The system is built using only free, accessible technologies to ensure feasibility and scalability without dependence on paid services. By offering automated, content-specific assessments along with adaptive learning guidance, IntelliTest supports deeper understanding, promotes AI-assisted self-learning, and makes personalized education accessible to a broader audience.

The educational landscape is evolving, yet the traditional methods of assessment remain largely unchanged and detached from the specific learning materials students need to focus on. Standardized tests and static question banks fail to provide personalized learning experiences that adapt to individual students' strengths and weaknesses. This led us to identify a crucial gap: the lack of dynamic, adaptive assessment tools that can cater to the diverse needs of modern learners.

Our idea for IntelliTest – Adaptive AI-Powered Assessment Generator was inspired by the desire to make assessments more responsive to the learner’s progress, not just as an evaluative tool, but as an integral part of the learning process itself. Many students struggle with learning material because the feedback and assessments they receive are not tailored to their specific needs or cognitive abilities. By utilizing advanced AI technologies and Bloom’s Taxonomy, a framework that categorizes cognitive skills into six levels: Remember, Understand, Apply, Analyze, Evaluate, and Create, we replace the one-size-fits-all assessments as we believe personalized learning paths can significantly enhance student engagement and outcomes.

The underlying hypothesis of this project is that by enabling students to engage with assessments that adapt to their abilities and provide immediate, personalized feedback, we can help them understand complex topics better and progress at their own pace. IntelliTest empowers students by offering dynamic assessments based on the material they study, and personalized study plans to address areas of weakness.

We believe this approach will succeed because it taps into the growing demand for personalized learning and AI-driven educational solutions. With advancements in AI models, such as GPT-Neo and similar tools, the ability to provide highly relevant, custom-tailored tests is now more feasible than ever. The system's capacity to generate immediate feedback and action plans is particularly relevant in today’s education, where personalized and adaptive learning is seen as essential to fostering both student success and engagement.

To implement IntelliTest – Adaptive AI-Powered Assessment Generator, we will leverage a combination of open-source technologies and platforms, ensuring both feasibility and scalability. At the core, the system will use GPT-Neo, a powerful open-source language model, to generate dynamic assessments based on uploaded study materials (PDFs, PPTs). For text extraction and content summarization from these study materials, we will utilize libraries such as PyMuPDF and Textract for PDF files and python-pptx for PowerPoint files. These tools will allow us to process diverse content types and make them ready for AI-driven analysis.

The LLM model will be hosted on Azure for cost-effective and scalable cloud computing, while MongoDB will be used to store user data and session histories. Additionally, the user interface will be built using Next.js and Tailwind CSS for an interactive and responsive frontend experience, with Flask (Python-based framework) to create an API that handles text processing, question generation, and response delivery.

To ensure market validation, we will first conduct pilot testing in educational institutions. Feedback from teachers and students will help refine the system. We will also use surveys and focus groups to understand the demand for personalized, AI-driven assessments, targeting both individual learners and educational institutions.

Core functionalities of IntelliTest include:
1. Personalized Assessment Generation: Dynamic creation of tests based on uploaded study materials and user settings (difficulty, question type, cognitive skill).
2. AI-Powered Feedback: Immediate solutions for test questions, with explanations that promote understanding.
3. Study Plan Recommendations: Tailored study plans based on test performance, helping students focus on areas that need improvement.

To ensure a positive user experience, the platform will offer an intuitive, easy-to-navigate interface and real-time feedback. Performance metrics to evaluate effectiveness will include user engagement (time spent, frequency of use), accuracy of question relevance, student performance improvement, and feedback ratings from users on the quality of assessments and recommendations.

The IntelliTest – Adaptive AI-Powered Assessment Generator is built on a modular architecture designed to provide personalized assessments based on uploaded study materials. The core components of the system include:

1. Frontend (User Interface): Developed using Next.js and Tailwind CSS, the frontend allows users to upload study materials (PDF, PPT), set preferences (difficulty, question type, cognitive skill), and view generated assessments. The interface is designed to be responsive and user-friendly, ensuring accessibility on both desktop and mobile devices.

2. Backend (Processing & APIs): The backend is powered by Flask for handling user requests, text processing, and integrating the AI model. It will manage the entire workflow, from receiving uploaded study materials to generating personalized assessments and providing feedback. The backend will also interface with the GPT-neo model to generate assessments and analyze user responses.

3. AI Model (Assessment Generation): The GPT-neo model will be used to generate personalized assessments based on the extracted and summarized content from the uploaded study materials. The model will create multiple types of questions—MCQs, subjective, numerical—based on the user settings. This AI model will be deployed on Azure Machine Learning for scalability and efficient cloud-based processing.

4. Database: User data, session histories, and test results will be stored in MongoDB, providing a flexible, scalable, and NoSQL solution.

Implementation Process for Innovative Features:
Dynamic Question Generation: The AI model will use the summarized content (via GPT-neo) to create contextually accurate and relevant questions.
Personalized Feedback & Study Plan: Post-assessment, the system will provide personalized study recommendations based on test performance, focusing on weaker areas.

Design and Development Timeline:
Day 1-3: Develop frontend interface using Next.js, design user flow
Day 4-5: Implement backend API with Flask, integrate text extraction and summarization features.
Day 6-10: Train and deploy T5 model for question generation, integrate with backend.
Day 11-13: Final testing, implement feedback and study plan generation features, deployment on Azure.

Performance Metrics:
Question Relevance: Accuracy of generated questions in relation to uploaded content.
User Engagement: Time spent on the platform, frequency of use, and completion rates.
Student Progress: Improvement in test scores over time, based on personalized study plans.
System Performance: Latency in generating assessments, scalability, and cloud deployment efficiency.

Conversion Work Plan:
Post-development, we will transition the project into scalable cloud-based deployment on Azure. The AI model will be continually updated to improve performance through user feedback and test data.

IntelliTest – Adaptive AI-Powered Assessment Generator represents a creative and innovative rethinking of how assessments can be made truly personalized and meaningful for learners. Unlike traditional models where assessments are generic and detached from what the student has specifically studied, IntelliTest allows users to upload their own study material and generates fully customized tests tailored to their personal learning context. This ensures that assessments are always directly relevant to the material the student is engaging with, creating a much more effective and motivating learning experience.

The system also innovates by integrating cognitive skill levels (based on Bloom’s Taxonomy) into the assessment generation process, allowing students to select whether they want to focus on remembering, understanding, applying, analyzing, evaluating, or creating. This choice empowers learners to engage at the level they are comfortable with or want to develop, respecting individual learning needs and supporting skill growth over time.

Another layer of creativity is introduced through personalized feedback and adaptive study plans. Instead of simply presenting right or wrong answers, IntelliTest analyzes test performance to generate targeted recommendations, helping students identify their weak areas and suggesting specific topics or skills to focus on. This turns assessments into learning tools rather than just evaluations, promoting continuous improvement.

By putting the learner at the center, offering dynamic, self-driven testing experiences, and providing actionable guidance afterward, IntelliTest creatively addresses the challenge of keeping assessments meaningful, motivating, and aligned with modern educational needs. It fosters autonomy, deeper engagement, and smarter learning journeys—all of which are essential for 21st-century education.

To ensure scalability, IntelliTest is designed with a modular architecture that separates frontend, backend, and AI processing layers. By deploying the backend APIs and AI model services on scalable cloud infrastructure, such as Azure, we can dynamically allocate resources based on real-time user demand. Load balancing, containerization using lightweight services like Docker, and database sharding for MongoDB will be used to handle large volumes of user uploads and concurrent assessment generations efficiently. Regular model optimization, caching frequently requested tasks, and asynchronous processing will help avoid performance bottlenecks.

To maintain long-term user engagement, IntelliTest goes beyond one-time assessments by offering continuous personalized learning journeys. After every assessment, students receive updated study plans, new challenges, and performance tracking insights, encouraging consistent learning and growth. Gamification elements, such as progress badges and streak rewards, will be introduced to enhance motivation over time.

For environmental sustainability, we will optimize the computational resources used by the AI models by selecting energy-efficient cloud regions and using inference-optimized instances when deploying models. Batch processing for non-urgent tasks and intelligent scaling policies will minimize unnecessary compute usage, thereby reducing the platform's carbon footprint.

As user needs evolve, IntelliTest is built to be adaptable. We will introduce modular updates, such as supporting additional file formats (e.g., Word docs, web pages), expanding cognitive skills beyond Bloom’s Taxonomy, and integrating multilingual support. Continuous feedback collection from users will guide feature updates and improvements.

By combining scalable infrastructure, resource-efficient design, continuous user engagement strategies, and adaptability to future educational trends, IntelliTest positions itself as a long-term, sustainable solution that grows with both its users and technological advancements.

IntelliTest addresses key social issues in education by promoting equity, inclusion, and personalized learning opportunities for a diverse range of students. By allowing learners to upload their own study materials and generate customized assessments, it removes the dependency on standardized textbooks and curricula, making quality assessment accessible to students across different socio-economic backgrounds, educational systems, and learning styles.

The integration of Bloom’s Taxonomy cognitive skills ensures that assessments cater to a variety of intellectual capabilities, recognizing that students have different strengths—whether in remembering, analyzing, or creating. This personalized approach not only respects individual learning diversity but also fosters deeper educational engagement for students who might otherwise feel marginalized by one-size-fits-all testing methods.

Moreover, IntelliTest is designed to be device-agnostic and lightweight, ensuring that students with limited access to high-end devices or fast internet connections can still benefit. By supporting inclusive education and providing adaptive learning paths, it helps bridge learning gaps and supports the broader social goals of equity, diversity, and lifelong learning.

To measure social impact, we will track metrics such as:
Accessibility usage (number of students from different backgrounds or regions),
Engagement rates (repeat users, test completion rates),
Learning improvement (pre- and post-assessment score differences),
Feedback analysis (qualitative data from users on relevance and inclusivity).

To ensure responsiveness to evolving community needs, IntelliTest will maintain an open feedback loop through surveys, user interviews, and collaboration with educational institutions. Continuous iteration based on real-world usage and community input will guide feature enhancements and maintain alignment with broader social objectives. Through its inclusive design and personalized support for all learners, IntelliTest aims to make high-quality, effective education accessible and empowering for everyone.

Personal Information Collection Statement (PICS):
1. The personal data collected in this form will be used for activity-organizing, record keeping and reporting only. The collected personal data will be purged within 6 years after the event.
2. Please note that it is obligatory to provide the personal data required.
3. Your personal data collected will be kept by the LTTC and will not be transferred to outside parties.
4. You have the right to request access to and correction of information held by us about you. If you wish to access or correct your personal data, please contact our staff at lttc@eduhk.hk.
5. The University’s Privacy Policy Statement can be access at https://www.eduhk.hk/en/privacy-policy.

• I have read and agree to the competition rules and privacy policy.