Annif at the GermEval-2025 LLMs4Subjects Task: Traditional XMTC Augmented by Efficient LLMs
This work addresses the challenge of efficient subject indexing for bibliographic data, but it is incremental as it builds on a previous system.
The paper tackled the problem of creating subject predictions for bibliographic records using large language models, with a focus on computational efficiency, by refining an existing system with small efficient models and LLM-based ranking, resulting in a 1st place ranking in both quantitative and qualitative evaluations of the GermEval-2025 shared task.
This paper presents the Annif system in the LLMs4Subjects shared task (Subtask 2) at GermEval-2025. The task required creating subject predictions for bibliographic records using large language models, with a special focus on computational efficiency. Our system, based on the Annif automated subject indexing toolkit, refines our previous system from the first LLMs4Subjects shared task, which produced excellent results. We further improved the system by using many small and efficient language models for translation and synthetic data generation and by using LLMs for ranking candidate subjects. Our system ranked 1st in the overall quantitative evaluation of and 1st in the qualitative evaluation of Subtask 2.