Research Topic: Education
In this research work:
(1) We propose the checklist framework STRaWBERRY to guide students writing their first research papers. STRaWBERRY allows to manually assess paper drafts,
(2) We provide STRaWBERRY-LLM, leveraging large language models to automate the assessment of paper drafts based on the structure and criteria defined by the STRaWBERRY framework. We suggest to use this as a way to iteratively assess own paper drafts.
STRaWBERRY is used in lectures on scientific writing. As a student or lecturer, feel free to use the framework and test the LLM version. We recommend to read our paper first and we recommend not to blindly trust the LLM output of STRaWBERRY-LLM.
(1) The STRaWBERRY framework:
STRaWBERRY is a checklist of common criteria, structured by typical components of research papers (please refer to our paper to understand the full context):
STRaWBERRY checklist framework:
Engaging students in scientific writing: The STRaWBERRY checklist framework with LLM-based paper draft assessment. Andreas Theissler, Marco Klaiber, Felix Gerschner, Philip Ritzer, Jie Wang (2025). IEEE EDUCON 2025
(2) STRaWBERRY-LLM: Large language model-based paper assessment based on STRaWBERRY
STRaWBERRY-LLM:
Uses an LLM to assess a paper w.r.t. the STRaWBERRY framework.
(Requires an OpenAI API key!)
Acknowledgements:
- Thank you to the groups of students who participated the lectures this framework was initially developed for. These were students in undergraduate and Master courses at Aalen University, in a PhD seminar at Stanford University, and in a seminar in industry.
- Moreover, we would like to thank Christian Beissmann and Ertan Yilmaz who were involved at an early stage with an initial literature survey.