Speaker：Dr. Maiga Chang
Full Prof., School of Computing and Information Systems, Athabasca University, Canada
Topic：Chatbot Buildiig and Authorship Forensics
Abstract：
Online learning and teaching do not mean that putting course materials online and asking students to learn by themselves. It is important to provide students supports when they encounter questions about course content or materials. When students ask their question on a discussion forum in an online learning environment, sometimes there may have no one available at that time to help them due to time differences or study behaviors and needs – for instances some students may have family/children/baby and day job and they might not be able to do their study until late night or weekends. This leads to an obvious conclusion that if a system was in place to provide an automated summary, this could facilitate learning. Having an easily accessible system, which can quickly provide responses, allows students to get information that may have otherwise been difficult to find. In this talk, I will explain how chatbots can be developed and discuss three potential chatbot in education applications: (1) the Speaking-based Conversation Quests v1.0 (https://conversation.megaworld.game-server.ca) where teachers can create conversation quests for different topics and students would be able to practice their speaking skills with an NPC (Non-Player Character) as many times as they want; (2) ChatbotLLM (https://chatbot.vipresearch.ca), is IEEE Northern Canada Section’s Capstone Project Award winnder project, is a system to automatically train educational chatbots on the materials uploaded by teachers; and (3) Visualized Editing Environment (https://vp.vipresearch.ca) is a block-based, visual editing environment to alleviate the burden of knowledge imposed on users wishing to implement chatbots in their use of training and/or as an automated first-level of support as well as adopting existing service providers include ChatGPT, Ask4Summary, and ChatbotLLM. At the end, I will also talk about Authorship Forensics, is the winner of SLERD Student Design Contest 2024, presents a novel approach to authorship attribution, demonstrates a significant ability of the models to distinguish between human and AI-written text, with precision 0.9682 (F0.5 score 0.95) for the 2-class (human and ChatGPT) testing subset and precision 0.9806 (F0.5 score 0.96) in the 3-class (human, ChatGPT 3.5, and ChatGPT 4) testing subset.
Venu：Room A127, S.E. Building II
Remind：
1.In Class、English speech
2.including cross-domain autonomous learning Sign up(Link)