Project: Transform

Contact: David Lamas (david.lamas@tlu.ee)

Position: Junior Research Fellow

Topic: Real-Time Data Pipelines to Enable Forecasting the Nature of Work, Labour-Market Needs and Required Competences

This refers to creating the mechanisms enabling the collection, processing, organisation and delivery of the data required for predicting trends related to job roles, tasks, and skills required in the workforce; understanding the demand for specific skills and competencies in the job market; identifying the essential skills, knowledge, and abilities needed for various roles.

Projec: Estonian Center of Excellence of Well-Being Science

Contact: Vladimir Tomberg (vladimir.tomberg@tlu.ee)

Position: Junior Research Fellow

Topic: Real-time Digital Behaviour Change Interventions

Digital behaviour change interventions are technology-based applications influencing human behaviour and promoting positive changes.

Computational interaction combines computer science, psychology, and design to create interactive systems influencing human behaviour.

When applied to digital behaviour change interventions on dynamic user profiles, computational interaction enables the development of personalised, adaptive, and effective interventions.

Contact: Linda Helene Sillat (sillat@tlu.ee)

Position: Doctoral Student

Topic: Enhancing Student Learning Experience Using Large Language Models

In today's educational landscape, lack of engagement and timely individualized feedback are common challenges faced by students. These challenges can significantly impact the learning process and hinder academic progress. Large Language Models (LLMs), offer a promising solution to address these challenges by providing advanced natural language processing capabilities.

The lack of engagement and timely feedback for students can lead to disinterest in learning and a decrease in academic performance, especially for High School and older students, who already have a whole stack of assignments, and a lot on the plate. Like other domains, AI has also given an effective and quick solution to this problem, in shape of Large Language Models. However, while using LLMs the prompt (input) provided to the model is not as easy as it seems. High quality inputs, or appropriately written prompts result in better output yet, poorly defined prompts lead to inaccurate responses or responses that may deviate or negatively impact the user.

The objective of this research is to develop a simple standalone prototype, that utilizes LLMs to aid 14 years and older students, in their learning process. The prototype will allow students to enter their prompts, which will then be processed (prompt engineering), hence generate tailored feedback, that addresses the specific needs and queries of students, which will further be given to the Language Model, for best possible outcome.

This research aims to enhance the learning experience of students, and providing them with timely as well as correct response to their respective queries. Therefore, the expected outcome of this research is a functional prototype that demonstrates the potential of LLMs in enhancing student learning experience. The prototype will provide students with timely and individualized feedback, addressing their queries and helping them stay engaged in the learning process. By leveraging LLMs, the prototype will have the potential to revolutionize the way students learn, providing them with personalized and effective feedback that caters to their individual needs.