CORE: A Few-Shot Company Relation Classification Dataset for Robust Domain Adaptation

TL;DR

CORE is a new dataset for few-shot relation classification focusing on company relations, highlighting challenges in domain adaptation and demonstrating the importance of high-quality data for robust model performance.

Contribution

We introduce CORE, a novel dataset for few-shot company relation classification, and evaluate models' domain adaptation capabilities, revealing key insights into data quality and contextual understanding.

Findings

Models trained on CORE perform better out-of-domain.

High-quality data improves domain adaptation.

Contextual cues are crucial for relation classification.

Abstract

We introduce CORE, a dataset for few-shot relation classification (RC) focused on company relations and business entities. CORE includes 4,708 instances of 12 relation types with corresponding textual evidence extracted from company Wikipedia pages. Company names and business entities pose a challenge for few-shot RC models due to the rich and diverse information associated with them. For example, a company name may represent the legal entity, products, people, or business divisions depending on the context. Therefore, deriving the relation type between entities is highly dependent on textual context. To evaluate the performance of state-of-the-art RC models on the CORE dataset, we conduct experiments in the few-shot domain adaptation setting. Our results reveal substantial performance gaps, confirming that models trained on different domains struggle to adapt to CORE. Interestingly, we find that models trained on CORE showcase improved out-of-domain performance, which highlights the importance of high-quality data for robust domain adaptation. Specifically, the information richness embedded in business entities allows models to focus on contextual nuances, reducing their reliance on superficial clues such as relation-specific verbs. In addition to the dataset, we provide relevant code snippets to facilitate reproducibility and encourage further research in the field.