Tailoring Chatbots in Education Through Large Language Models and Prompt Engineering

JTELSS logo large
Tailoring Chatbots in Education Through Large Language Models and Prompt Engineering Thursday 16/05 16:00-19:00h Workshop Space A Needs Analysis This workshop introduces how to use large language models (LLMs) for creating educational chatbots that can assist learners, educators, and researchers. We will teach advanced generative AI prompting techniques, a skill

Speakers

Erfan Jalili Jalal
Erfan Jalili Jalal
Open University of the Netherlands
Alla Krasulia
Alla Krasulia
Sumy State University, Ukraine & Riga Technical University, Latvia
Roland Klemke
Open University of The Netherlands, The Netherlands / TH Köln, Germany
Krist Shingjergji
Krist Shingjergji
Open University of the Netherlands

Start

16/05/2024 - 16:00

End

16/05/2024 - 19:00

Tailoring Chatbots in Education Through Large Language Models and Prompt Engineering

Thursday 16/05 16:00-19:00h
Workshop Space A
Needs Analysis

This workshop introduces how to use large language models (LLMs) for creating educational chatbots that can assist learners, educators, and researchers. We will teach advanced generative AI prompting techniques, a skill for using LLMs effectively, and show LLMs interfaces for chatbot design. The workshop is relevant and timely for TEL and EdTech, as chatbots can improve learning outcomes, engagement, motivation, and well-being. Following the instructions and presentation, we will showcase a chatbot designed for online group learning environments, ready for students to start a conversation and help them cope with their emotional challenges during group collaboration, such as conflicts and frustration. The workshop will have a hands-on group activity, where participants will create prompts for a chatbot based on an emotional conflict scenario provided to them. This will give them practical experience and knowledge and also provide feedback for our study and publication.

 

Learning Objectives
  1. Understanding Large Language Models (LLMs): Acquire foundational knowledge of LLM functions, development, and application in education.
  2. Mastering prompt engineering: Learn 12 advanced AI prompting techniques for TEL, with guidelines for customizing LLMs for research and education.
  3. Developing custom chatbots: Build custom chatbots using advanced AI prompts, judged by their conversational accuracy and adherence to instructions in educational simulations.
  4. Utilizing LLM interfaces for advanced applications: Gain proficiency in LLM interfaces like ChatGPT API for project-specific training, usage, and customization.
  5. Exploring chatbots in emotion regulation: Explore chatbots’ impact on emotion regulation in online group learning using case studies and research.
  6. Applying skills to research and educational projects: Implement knowledge and skills in LLMs or chatbots for innovative TEL projects, focusing on innovation, feasibility, and LLM integration.

 

Pre-activities

Please bring your laptops and ensure you have an internet connection. Come prepared with an open mindset, ready to learn and apply new technological skills.

 

Session Description
  • Introduction and icebreaker (5 mins): A quick meet and greet to foster a welcoming workshop environment.
  • LLM Basics (30 mins): A comprehensive overview of Large Language Models (LLMs), including the intricacies of prompt engineering and practical tips for effective utilization. Participants will also be introduced to the ChatGPT API interface as a crucial tool for leveraging LLMs.
  • Learning 12 sophisticated AI prompting techniques (30 mins): Participants will delve into 12 advanced AI prompting techniques, exploring their applications and impact on enhancing interactions with AI models.
  • Chatbot Demo (15 mins): A showcase of a chatbot specifically designed for emotion regulation, demonstrating its capabilities and the technology behind it.
  • Break (10 min): A short intermission to relax and network.
  • Group Activity (35 mins): A hands-on session where participants, working in groups, apply their newly acquired knowledge to create prompts for an assigned emotion regulation scenario.
  • Group Presentations (35 mins): Time for groups to share and discuss their designed chatbots, reflecting on the creative process, the outcomes, and the lessons learned from the activity.
  • Wrap-Up (20 mins): A concluding session to summarize key takeaways, discuss potential applications of the concepts learned, and provide guidance for further exploration and learning in the field of LLMs and AI prompting techniques. This segment will also include a Q&A, allowing participants to address any remaining questions and engage in deeper discussions on the topics covered.