Harnessing Online Learning with LLM Oriented Web-apps for Teaching, Research, Evaluation, and Engagement
This project aims to explore the use of large language models (LLMS, e.g. ChatGPT) in the higher-education classroom by providing students a way of checking their knowledge on specific topics by interacting with a chatbot that has been configured and scoped by course instructors to fit the course content.
Tip
For details on the research and methodology behing this project, we refer to our OSF Preprint
This repository holds the source code for two web-based Python Shiny apps that will help accomplish this goal:
Science Concept Understanding via Interactive Reinforcing Review by Educational Large-language-model
This is the student facing application where users will interact with the LLM (i.e. SCUIRREL) to explore the concepts of a specific topic by means of a guided conversation. In addition to the conversation, students can take instructor verified quiz-questions to test their knowledge
This application is the backend to SCUIRREL where instructors can set-up, manage and monitor the topics and concepts that SCUIRREL will be able to discuss with students.
Read more about this project based on your role and interests:
- I'm a student wanting to know how I can use SCUIRREL to help me in my learning
- I'm an instructor interested in exploring a use for an app like this my classroom and how to set it up
- I'm an IT admin assisting in hosting and deploying these apps
- I'm a developer interested in how these app were built and play around with the code
- I'm a researcher looking to analyse any of the data collected by the apps
The idea for this project came from the observation that in many higher education courses students come in with very diverse backgrounds and have different levels of understanding of the topics being discussed in the classroom. Instructors often need to spend valuable class time getting everyone on the same page, which can be frustrating for both the students who are already familiar with the material and the ones who are struggling to keep up.
The goal of this project is to provide a tool that can help students check their understanding of the material outside of class time, and to provide instructors with a way to monitor and guide this process. The hope is that this will help to level the playing field for students with different levels of background knowledge, and to free up more class time for discussion and activities that can help students deepen their understanding of the material.
SCUIRREL and ACCORNS are experimental applications built as part of an ongoing research project. They are not intended to be used as a replacement for traditional teaching methods, but rather as a supplement to explore the potential of LLMs in the classroom and collect data on the experience from both students and instructors.