Practice Exam Generator (PEG)

• 

  Logo

• 

  Submission Page

• 

  Loading Page

• 

  Question Page

• 

  Feedback Page

Inspiration

The inspiration for this project is to develop a system that can automatically generate practice questions from uploaded class materials, allowing students to engage in self-paced learning and improve their understanding of concepts.

What it does

It helps students prepare for their exams by generating personalized practice questions based on class materials such as PDFs or slides.

How we built it

We built it using the Llama Large Language Model (LLM) to analyze and create problem sets. The AI model is linked to our backend service running on Node.js, with web pages written in React.js.

Challenges we ran into

Connecting a backend service with the AI model to receive proper responses took considerable time, not only to retrieve appropriate questions but also to validate JSON values. Another challenge we faced as a group was analyzing PDF files, as they are not built or programmed to be read by any programming languages. The result of which was using an OCR algorithm to detect the characters in a given PDF and convert it to text files.

Accomplishments that we're proud of

We successfully followed branch naming and commit naming conventions to reduce confusion in our codebase when pushing our code. We succeeded in merging every team member's work into one main branch in the repository.

What we learned

We learned how to collaborate with teammates to create an integrated project through GitHub and other services, such as Figma. We did our best to follow conventions and rules to avoid confusing other contributors when they examine the codebase. As a result, we improved our ability to work together effectively.

What's next for Practice Exam Generator (PEG)

We can add more text fields that can be used to request AI models to personalize questions, for example, 'Emphasize the materials related to the Master Method'. Additionally, it's possible to customize the style of the generated exam questions beyond simple multiple-choice format.

Built With

• figma
• git
• github
• groq
• javascript
• llama
• node.js
• python
• react