Abstract: Prompt tuning shows great potential to support relation extraction because it is effective to take full use of rich knowledge in pretrained language models (PLMs). However, current prompt tuning models are directly implemented on a raw input. It is weak to encode semantic dependencies of a relation instance. In this paper, we designed a cueing strategy which implants task specific cues into the input. It enables PLMs to learn task specific contextual features and semantic dependencies in a relation instance. Experiments on ReTACRED corpus and ACE 2005 corpus show state-of-the-art performance in terms of F1-score.
Paper Type: short
0 Replies
Loading