Implementing a Chatbot Solution for Learning Management System
Project Overview
The document explores the role of generative AI, particularly chatbots, in the educational sector, emphasizing their integration within the Blackboard learning management system to enhance student learning experiences. By utilizing natural language processing, these chatbots effectively engage with students, offering timely responses to course-related inquiries and fostering interactive learning environments. While the implementation of such technology presents challenges, including the necessity for comprehensive training and the intricacies of language modeling, the potential benefits are significant. The use of chatbots supports various learning styles, helping to accommodate diverse student needs and promoting a more personalized educational experience. Overall, the findings suggest that generative AI can play a transformative role in education by facilitating student engagement and improving access to information.
Key Applications
Educational Chatbot
Context: Used in various educational settings including higher education for students, group chats during the pandemic, and support for individuals with intellectual disabilities. The chatbots facilitate interactive learning, group engagement, and social skills practice.
Implementation: The chatbots were developed using technologies such as Python with frameworks like ChatterBot. They were integrated into platforms like Blackboard and used in group messaging applications like WhatsApp to maintain engagement. Additionally, specialized bots like CapacitaBOT were designed to improve social interactions and daily living skills for users with intellectual disabilities.
Outcomes: Increased student engagement, improved educational outcomes over traditional methods, and enhanced social skills practice for users. Reported improvements in student performance by over 20% in sign language learning contexts.
Challenges: Initial training issues, difficulties in mimicking human conversation, integration challenges with existing platforms, and limitations in accessibility and customization for specific user needs.
Gamified Educational Chatbot
Context: Targeted at children and students learning American Sign Language, focusing on lifestyle choices and sign language movements.
Implementation: A gamified chatbot designed to educate children on lifestyle choices while also providing feedback on sign language movements, improving learning outcomes through interactive engagement.
Outcomes: Promising results in educational improvement over traditional tutoring methods and a reported increase in student performance in sign language courses by over 20%.
Challenges: Need for targeted content and specific audience adaptation, as well as dependence on accurate movement recognition and feedback mechanisms.
Virtual Teaching Assistant
Context: Used in online courses to assist students by answering frequently asked questions and providing support.
Implementation: The virtual teaching assistant was designed to interact with students by addressing their inquiries, enhancing student interaction and support during online learning.
Outcomes: Enhanced student interaction and support for course-related inquiries, leading to a more engaging online learning environment.
Challenges: Limited by the virtual assistant's understanding of complex questions and the variability of student inquiries.
Implementation Barriers
Technical Barrier
Integration issues with Blackboard’s security system hindered effective web scraping.
Proposed Solutions: Developing alternative sources for data gathering from low-security websites or trusted external resources.
Training Barrier
Chatbot needed extensive training to understand context and provide accurate answers.
Proposed Solutions: Use of virtual machines for training, continuous data feeding for learning, and employing smaller, targeted datasets.
Design Barrier
The initial design of the chatbot was not user-friendly and lacked aesthetic integration with the Blackboard system.
Proposed Solutions: Collaboration with design experts to improve the interface and user experience.
Security Barrier
Risks associated with open-source package management led to vulnerabilities.
Proposed Solutions: Implementing secure coding practices and careful package management.
Project Team
Dimitrios Chaskopoulos
Researcher
Jonas Eilertsen Hægdahl
Researcher
Petter Sagvold
Researcher
Claire Trinquet
Researcher
Maryam Edalati
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Dimitrios Chaskopoulos, Jonas Eilertsen Hægdahl, Petter Sagvold, Claire Trinquet, Maryam Edalati
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