Skip to main content Skip to navigation

Build-a-Bot: Teaching Conversational AI Using a Transformer-Based Intent Recognition and Question Answering Architecture

Project Overview

The document outlines the 'Build-a-Bot' initiative, designed to enhance AI literacy among students by enabling them to create transformer-based chatbots through hands-on experience. This initiative underscores the importance of AI education for all learners, regardless of their future career paths, and provides an open-source tool that supports educators with diverse technical backgrounds in teaching AI concepts. By allowing students to gather data, identify intents, and generate responses aligned with their curriculum, the tool fosters a constructivist approach to learning, encouraging active engagement and deeper understanding of AI technologies. The findings suggest that such initiatives can significantly enhance students' comprehension of AI and its applications, equipping them with essential skills for the future. Overall, the document highlights the potential of generative AI in education to democratize knowledge, promote critical thinking, and prepare students for an increasingly AI-driven world.

Key Applications

Build-a-Bot

Context: Middle school education, focusing on AI literacy and hands-on learning.

Implementation: An open-source tool that allows students to create their own chatbots based on course material using a natural language processing pipeline.

Outcomes: Students learn AI principles, develop problem-solving skills, and engage in constructivist learning by customizing their chatbots.

Challenges: Ensuring the tool is accessible and intuitive for middle school students, and addressing the varying levels of technical expertise among educators.

Implementation Barriers

Technical

Difficulty for educators with low technical expertise to implement AI tools in the classroom.

Proposed Solutions: Developing an intuitive user interface and providing comprehensive documentation to facilitate use.

Resource Availability

Limited access to computational resources for running AI models in schools.

Proposed Solutions: Creating lightweight models that can run efficiently on standard classroom hardware.

Project Team

Kate Pearce

Researcher

Sharifa Alghowinem

Researcher

Cynthia Breazeal

Researcher

Contact Information

For information about the paper, please contact the authors.

Authors: Kate Pearce, Sharifa Alghowinem, Cynthia Breazeal

Source Publication: View Original PaperLink opens in a new window

Project Contact: Dr. Jianhua Yang

LLM Model Version: gpt-4o-mini-2024-07-18

Analysis Provider: Openai

Let us know you agree to cookies