Prompt-based Zero-shot Relation Classification with Semantic Knowledge Augmentation

Jiaying Gong, Hoda Eldardiry

Abstract

Recognizing unseen relations with no training instances is a challenging task in the real world. In this paper, we propose a prompt-based model with semantic knowledge augmentation (ZS-SKA) to recognize unseen relations under the zero-shot setting. We generate augmented instances with unseen relations from instances with seen relations following a new word-level sentence translation rule. We design prompts based on an external knowledge graph to integrate semantic knowledge information learned from seen relations. Instead of using the actual label sets in the prompt template, we construct weighted virtual label words. By generating the representations of both seen and unseen relations with augmented instances and prompts through prototypical networks, distance is calculated to predict unseen relations. Extensive experiments conducted on three public datasets show that ZS-SKA outperforms state-of-the-art methods under the zero-shot scenarios. Our experimental results also demonstrate the effectiveness and robustness of ZS-SKA.