AI for Education Innovation Lab

AI for Education Innovation Lab

The AI for Education Innovation Lab was set up as a part of the research initiatives by the Centre for Education Research and Innovation (CERI). The objective of this lab is to facilitate the development and use of applications that leverage Artificial Intelligence (AI) to improve the efficiency of various practices on teaching and learning. Under this umbrella of AI for education, we have several projects including the following:

Project 1: AI-based teaching assistant

This project involves developing an AI-based teaching assistant with two main functionalities. First, it will assist students by answering their queries. By training the AI model on course content, the assistant will provide subject-specific answers and offer immediate support to learners. Second, it will aid faculty members in assessing student submissions. The AI model will deliver detailed evaluations of students’ work and provide in-depth, individualized feedback.

Project outcomes:

1. Increased engagement and understanding of course material: The AI teaching assistant will provide personalized support to students round the clock which can improve students' academic performance.

2. Streamlined Assessment and Feedback Process: The AI assistant will help faculty members save time in grading and provide more meaningful insights into student performance. 

Project 2: AI-based evaluation of traits of good instructors

This project involves using an AI model to analyse the performance of instructors. It is trained on various kinds of course data such as video recordings, transcripts, and course materials. It identifies a series of traits of effective teachers so that it can help with the design and improvement of faculty development programmes. It will also help in the process of recruitment of new instructors based on their teaching sample videos and career profiles.

Project outcomes:

1. Improved Faculty Development Programs: By identifying key traits of effective teachers, it will be able to provide actionable insights to faculty members about their performance. It can support continuous improvement among instructors by identifying specific areas that require attention while designing faculty development programs.

2. Enhanced Recruitment Processes: The AI model will provide a data-driven approach to streamline the recruitment process to select candidates who demonstrate the most effective teaching traits, ultimately improving the quality of instruction offered to students.

Project 3: AI-based generation of multiple variants of question papers

This project looks into the development of an AI model that can take a faculty-designed question paper as input and generate multiple variants of the question paper. These question papers generated will be such that the topic of focus and difficulty level will remain the same while the questions will be altered. These variants of questions will help us evaluate students fairly and uniformly while minimizing opportunities for academic dishonesty and enhance the integrity of the examination process. It can also aid in the creation and updation of subject-specific and topic-specific question banks.

Project outcomes:

1. Fair and Uniform Assessment: The AI-generated question paper variants will ensure that all students are evaluated on the same topic and difficulty level, promoting fairness in assessments. 

2. Reduction in Unfair Means: As the students are evaluated using multiple variations of the question papers in the examination, it reduces the possibility of students using unfair means in examinations.

3. Dynamic Question Bank Enhancement: By facilitating the creation and regular updating of  banks, the AI model will ensure that faculty have access to a diverse set of questions.

Project 4: AI-based Dissertation evaluation system 

This project aims to develop and deploy an AI-model that can assist faculty in the evaluation of student dissertation reports. It can also help students in the process of improving their dissertation prior to submission. By utilising an assessment rubric, the system can provide detailed evaluations of the dissertation and suggest areas for improvement.

Project outcome:

1. Time Efficiency in Evaluation: The AI model will significantly reduce the time faculty spend on assessing dissertation reports by automating initial evaluations based on a structured rubric. This allows instructors to focus on more complex aspects of feedback and mentoring, ultimately streamlining the evaluation process.

2. Comprehensive Feedback for Students: By delivering detailed evaluations and specific suggestions for improvement, the AI system will provide students with valuable insights into their work. This in-depth analysis can guide students in enhancing the quality of their dissertations, leading to higher-quality submissions and better learning outcomes.

Facilities

4 Desktop Systems

High-performance HPE DL380 Gen10 Server: Equipped with Intel Xeon Gold 6148 processors, NVIDIA A100 80GB GPU, 256 GB Ram, and ample storage designed to deliver seamless LLM training and inference for transformative educational applications.