From an ML/LLM Project to a Research Paper at an A* Conference

Abstract

Building a good ML/LLM project does not automatically make it research. Even a strong model, a new benchmark, or an agent needs to be framed appropriately as a scientific contribution: identifying the research contribution, designing experiments, comparing the approach with existing methods, and presenting the results convincingly. In this workshop, we will walk through the entire process, from an initial idea and project development to a research paper submitted to a leading international conference. We will discuss how to assess whether your work offers scientific novelty, how to turn a project into a research question, how to choose the appropriate paper format and conference, how to structure a paper, and how to develop a strong experimental section. We will examine real-world examples and discuss how to strengthen your own academic, course, thesis, or research project to give it the potential for international publication.