Skip to content

Large language models for social sciences (LLAMOSS)

 

Project no.: S-MIP-24-26

Project description:

The project seeks to take advantage of advances in data science, natural language processing and agent-based modelling to develop a system of novel research methods to conduct comprehensive literature reviews and to computationally model theoretical propositions obtained from such reviews. Researchers frequently have to review huge amounts of academic publications and other sources, including tabular data sets that may span different disciplines and fields within disciplines and the approach of systematic literature reviews is seldom a practical option due to its resource greedy nature. To speed up the process and make it transparent and repeatable by any interested part regardless of IT/programming skills we will develop a solution for researchers that automates the tasks by using large language models, agents, semantic search and network analysis. The novelty of the approach is in the unique configuration of processes that will be developed and automated and in open-source nature of the prospective product. Furthermore, once the knowledge from academic publications is acquired, and scholars struggle to develop (or test) a theory/a nomological network of constructs, we propose to make the use of the agent-based modelling and coupled with the large language models to approach the theory development and testing more systematically and transparently.

Project funding:

Projects funded by the Research Council of Lithuania (RCL), Projects carried out by researchers’ teams


Project results:

Planned project outcomes:

1. Software prototypes: code for automated literature reviews, agent-based models, and synthetic datasets.
2. Three research papers in peer-reviewed management journals ranked in the top two quartiles.
3. Participation in three international conferences on management and public administration.
4. Organization of three workshops for researchers and other stakeholders.
5. Dissemination of prototypes for automated literature reviews, theory generation, and testing on platforms like GitHub.

Period of project implementation: 2024-09-02 - 2027-08-31

Project coordinator: Kaunas University of Technology

Project partners: Vilnius University

Head:
Rimantas Rauleckas

Duration:
2024 - 2027

Department:
Academic Centre of Social Sciences, Arts and Humanities, Faculty of Social Sciences, Arts and Humanities