2022 年 29 巻 4 号 p. 1138-1164
This paper explores how the Distantly Supervised Relation Extraction (DS-RE) can benefit from the use of a Universal Graph (UG), the combination of a Knowledge Graph (KG) and a large-scale text collection. A straightforward extension of a current state-of-the-art neural model for DS-RE with a UG may lead to degradation in performance. We first report that this degradation is associated with the difficulty in learning a UG and then propose three training strategies: (1) Path Type Adaptive Pretraining, which sequentially trains the model with different types of UG paths; (2) Path Type-wise Local Loss, which is an alternative approach of the Path Type Adaptive Pretraining to generate UG path type-wise local error signals so as to prevent the reliance on a single type of UG path; and (3) Complexity Ranking Guided Attention mechanism, which restricts the attention span according to the complexity of UG paths so as to force the model to extract features not only from simple UG paths but also from complex ones. Experimental results on both biomedical and NYT10 datasets prove the robustness of our methods and achieve a new state-of-the-art result on the commonly used NYT10 dataset. The code and datasets used in this paper are available at https://github.com/baodaiqin/UGDSRE. In addition, a DS-RE toolkit developed based on this work is available at https://github.com/baodaiqin/UKG-RE.
