A Few-Shot Approach for Relation Extraction Domain Adaptation using Large Language Models
Vanni Zavarella, Juan Carlos Gamero-Salinas, Sergio Consoli
TL;DR
This work tackles domain adaptation for relation extraction by leveraging in-context learning with large language models to generate in-domain, schema-constrained annotations for AECO scientific texts. A baseline SpERT model trained on SciERC data is compared against models trained on LLM-generated data, including combinations with SciERC labels. Findings indicate that LLM-generated data alone yields modest gains, but combining it with out-of-domain labels improves performance, particularly for named-entity recognition, suggesting a cost-effective path to adapt smaller RE models for new scientific domains. The study highlights both the potential and current limitations of LLM-based data augmentation for scalable knowledge-graph generation in domain-specific literature.
Abstract
Knowledge graphs (KGs) have been successfully applied to the analysis of complex scientific and technological domains, with automatic KG generation methods typically building upon relation extraction models capturing fine-grained relations between domain entities in text. While these relations are fully applicable across scientific areas, existing models are trained on few domain-specific datasets such as SciERC and do not perform well on new target domains. In this paper, we experiment with leveraging in-context learning capabilities of Large Language Models to perform schema-constrained data annotation, collecting in-domain training instances for a Transformer-based relation extraction model deployed on titles and abstracts of research papers in the Architecture, Construction, Engineering and Operations (AECO) domain. By assessing the performance gain with respect to a baseline Deep Learning architecture trained on off-domain data, we show that by using a few-shot learning strategy with structured prompts and only minimal expert annotation the presented approach can potentially support domain adaptation of a science KG generation model.