Preprint
Enhancing Student Engagement and Support in Hybrid IT Teaching through an AI Chatbot [Practice Report]
SSRN Electronic Journal, Vol.Series Paper No.30
Southern Cross University Scholarship of Learning and Teaching Research Paper Series, Elsevier
19/08/2025
Metrics
21 Record Views
Abstract
This practice-based paper presents the design and implementation of an AI-powered chatbot to support students in an Information Technology unit at Southern Cross University (SCU), delivered in a six-week focused block teaching model. The initiative was introduced to address challenges faced by diverse cohorts, including limited time for concept mastery and the need for timely, out-of-hours academic support. Leveraging GPT-4 as the language model, the chatbot was developed using Langchain and Streamlit, integrated with a MongoDB vector database and embedded into Blackboard. The chatbot was trained on a curated knowledge base comprising previous FAQs, unit content, and key weekly concepts. It provided tailored assistance with assessments, contents, and general study strategies. Evaluation showed promising results: content engagement reached 45.3% - above the average of 38.3% for comparable first-year units - with over 500 chatbot interactions, 53% of which occurred after hours or on weekends. Most student queries (72%) were simple conceptual questions, with a strong focus on assessments (54%). Unit satisfaction increased to 4.67/5, and the success rate reached 90%. The findings underscore the potential of customised AI chatbots to enhance engagement and learning, offering advantages over generic platforms (e.g., ChatGPT) by ensuring content relevance, privacy, and secure integration within institutional learning ecosystems.
Details
- Title
- Enhancing Student Engagement and Support in Hybrid IT Teaching through an AI Chatbot [Practice Report]
- Creators
- Vinh Bui - Southern Cross University
- Publication Details
- SSRN Electronic Journal, Vol.Series Paper No.30
- Series
- Southern Cross University Scholarship of Learning and Teaching Research Paper Series
- Publisher
- Elsevier
- Identifiers
- 991013310928502368
- Academic Unit
- Information Technology; Faculty of Science and Engineering
- Language
- English
- Resource Type
- Preprint
- Local Fields
- Original Research - SoLT