Large Language Models for Chatbots

Project Overview

About the Current Project

Knowledge graph illustration for the ChatbotLLM project

Dr. Chang's research group proposed, developed, piloted, and published an Automatic Text Summarization service named Ask4Summary in 2021 based on Natural Language Processing techniques (Pal, Chang, & Iriarte, 2021). It answered user questions by identifying relevant content and generating summaries only based on the provided materials (Saleh, Chang, & Iriarte, 2022; Saleh, Iriarte, & Chang, 2022) and reached 82.69% success rate for providing quick responses of course relevant questions.

Following Ask4Summary, an Alberta Innovates/Athabasca University Summer Research Studentship project was proposed in 2024 to design and develop ChatbotLLM, an open-access system for training educational chatbots on uploaded course materials. ChatbotLLM's lead developer, AU BScCIS graduate Supun De Silva, won IEEE Northern Canada Section's Capstone Project Award on October 18, 2024.

Create and upload

Teachers create courses and upload PDF, DOC/DOCX, PPT/PPTX, XLS/XLSX, TXT, HTML/HTM, ZIP, MP3, MP4, and WAV materials.

Digest and train

The server reads the materials with NLP tools and prepares course-specific language models and retrieval data.

Respond with context

The chatbot dynamically loads the relevant course model and responds to student questions from course content.

Collect feedback

Anonymous interaction data and feedback help teachers and researchers evaluate chatbot response quality.

Access Policy

Terms of Use

The VIP Research Group is led by Prof. Maiga Chang at the School of Computing and Information Systems, Athabasca University. This large language models for chatbots service (chatbot.vipresearch.ca) is one of the research group's works, with follow-up research planned for future projects.

Almost all of Prof. Chang's works are open access or open source. This web service is open access and currently runs on a self-sponsored server, like other research projects listed among Prof. Chang's advanced projects. It is intended to remain online, improving, and accessible as long as operating costs are affordable and covered.

If access volume becomes too high, or if a business or commercial institution uses the service for profit, the terms may change to include donations, personal, academic, or business licensing, or subscription models.

Start Here

How to Access

Teachers Create courses, upload course materials, configure models, and manage training. Students Use the browser app to ask questions once a course chatbot is ready. Researchers Test model output, review responses, and compare chatbot behavior.

Research Direction

About Us

Our Mission

Our research trains course-grounded language models, builds educational chatbots, runs real-course experiments, and analyzes teacher and student feedback.

Our Supervisor

Dr. Maiga Chang is a Full Professor in the School of Computing and Information Systems at Athabasca University, Canada.

Research Goal

The project focuses on chatbots that answer student questions from large language models and retrieval context trained on specific course content.

Contributors

Our Team

Supun DE SILVA

2023 - Present

Supun De Silva is a Master of Science in Computing Information Systems student at Athabasca University, specializing in artificial intelligence and intelligent systems. He completed his Bachelor of Science in Computing Information Systems with Great Distinction and a 4.0 GPA, and received the Governor General's Silver Medal. As an NSERC research assistant and lead developer on ChatbotLLM, his work focuses on full-stack web development, natural language processing, and educational AI systems.

Zibusiso MAFAITI

2025 - Present

Zibusiso Mafaiti is a second-year student pursuing a BSc in Computing and Information Systems. He is passionate about research and its impact on modern learning.

Owen LI

2025 - Present

Owen is a software developer in the School of Applied Science at Queen's University, Canada.

Yi-Ting LEE (Tina)

2025

Tina is an undergraduate student majoring in Digital Multimedia Design at Cheng Shiu University, Taiwan. She contributes to flyer design and visual layout work.

Ting-Yu CHOU (Victoria)

2025

Victoria is a fourth-year university student majoring in Information Management at National Central University in Taiwan.

Demonstrations

Videos

ChatbotLLM Research

Introduces ChatbotLLM, the IEEE Northern Canada Section Capstone Project winner, and its support for course materials including documents, media, OCR, speech recognition, and ZIP packages.

Using the ChatbotLLM Web App

This guide shows how to use ChatbotLLM directly in a browser.

Using ChatbotLLM on Discord

This guide shows how to use ChatbotLLM on Discord.

Using ChatbotLLM on LINE

This guide shows how to use ChatbotLLM through the LINE messaging service.

Subprocess-based MASA vs. AskAU on March 30, 2026

Earlier version of subprocess-based MASA (Mobile-Accessible Seamless Agent), a browser-based chatbot app powered by ChatbotLLM research shows that MASA responded generally in the same speed like (sometimes had couple of seconds slower than) the AI-Powered Live Chat app but provided information for the users according that they may want to see while asking a question.

Worker-enabled MASA vs. AskAU on April 23, 2026

The improved Worker-enabled MASA (Mobile-Accessible Seamless Agent), a browser-based chatbot app powered by ChatbotLLM research shows that in the most of time worker-enabled MASA responded quicker than the AI-Powered Live Chat app and also provided information for the users according that they may want to see while asking a question.

Research Output

Publications

Fuzheng ZHAO, Zibusiso MAFAITI, Yongjin LI, and Maiga CHANG. (2026). Mobile-Accessible Seamless Agent (MASA): Using ChatbotLLM for Grounding Chatbots in Teacher-Provided Materials. In: APSCE 4th International Conference on Artificial Intelligence, Metaverse and Artificial Companions in Education and Society (AI-MetaACES 2026), Kumamoto, Japan, June 25-27, 2026 (accepted)
PDF
Supun DE SILVA and Maiga CHANG. (2025). ChatbotLLM - Training Educational Chatbots on the Materials Uploaded by Teachers. In: 13th International Conference on Information and Education Technology (ICIET 2025), Fukuyama, Japan, April 18-20, 2025. (IEEE)
DOI PDF

Details

Frequently Asked Questions

What is ChatbotLLM?

ChatbotLLM is an open-access educational chatbot service for training course-specific chatbots from materials uploaded by teachers. Teachers create a subject, choose configuration options, upload materials, and students can then ask questions through browser or connected client applications.
How is ChatbotLLM different from a general AI chatbot?

The published ChatbotLLM and MASA papers emphasize teacher-grounded retrieval. ChatbotLLM searches teacher-provided course materials with sentence embeddings and cosine similarity, returns the top matching course paragraphs, and can say it does not know when no relevant course content is found.
What file types can teachers upload?

Teachers can upload common teaching materials including PDF, PPT/PPTX, DOC/DOCX, XLS/XLSX, TXT, HTML/HTM files, MP3, MP4, WAV, and ZIP packages containing supported files. OCR, speech recognition, URL extraction, web crawling, enrichment, and data preparation are supported.
What are RAG and rephrase options used for?

Rephrase mode sends retrieved course paragraphs to a generative model to make the answer more readable while preserving original meaning and key terms. RAG mode sends the question, retrieved context, and prompt to a generative model so the final answer is generated from selected course context.
Which API endpoints are available for client apps?

Client apps can use endpoints in the includes folder. Common endpoints include get_trained_models.inc.php, get_course_details.inc.php, get-model-names-for-course.inc.php, get-model-output.inc.php, and update_feedback.inc.php.
Can ChatbotLLM be used outside the website?

Yes. There is a browser-based mobile-accessible client, and the MASA paper shows endpoint documentation on how external clients can request ChatbotLLM responses. The service can also be connected to Discord, LINE, and VEE.
How are private subjects protected?

A subject can be public or protected with an access key. Client apps can check and submit access keys through endpoints such as verify_access_key.inc.php and get_course_details.inc.php. Invalid private subject access is rejected and recorded.
How does feedback work?

Each chatbot response can be stored with a feedback ID. Client apps can call update_feedback.inc.php with the feedback type, rating, and wording to support research comparisons of learner perceptions.
Where can I read the research papers?

The Publications section links to the ChatbotLLM ICIET 2025 paper and the MASA AI-MetaACES 2026 paper.

Training educational chatbots on teacher-provided course materials