Improving Long-Tail Relation Extraction with Collaborating Relation-Augmented Attention

Proceedings of the 28th International Conference on Computational Linguistics, pages 1653–1664 Barcelona, Spain (Online), December 8-13, 2020 1653 Improving Long-Tail Relation Extraction with Collaborating Relation-Augmented Attention Yang Li1, Tao Shen1∗, Guodong Long1, Jing Jiang1, Tianyi Zhou2 and Chengqi Zhang1 1Australian AI Institute, School of Computer Science, FEIT, University of Technology Sydney 2University of Washington, Seattle {yang.li-17,tao.shen}@student.uts.edu.au, tianyizh@uw.edu {guodong.long,jing.jiang,chengqi.zhang}@uts.edu.au Abstract Wrong labeling problem and long-tail relations are two main challenges caused by distant su- pervision in relation extraction. Recent works alleviate the wrong labeling by selective attention via multi-instance learning, but cannot well handle long-tail relations even if hierarchies of the relations are introduced to share knowledge. In this work, we propose a novel neural network, Collaborating Relation-augmented Attention (CoRA), to handle both the wrong labeling and long-tail relations. Particularly, we first propose relation-augmented attention network as base model. It operates on sentence bag with a sentence-to-relation attention to minimize the effect of wrong labeling. Then, facilitated by the proposed base model, we introduce collaborating relation features shared among relations in the hierarchies to promote the relation-augmenting process and balance the training data for long-tail relations. Besides the main training objective to predict the relation of a sentence bag, an auxiliary objective is utilized to guide the relation- augmenting process for a more accurate bag-level representation. In the experiments on the popular benchmark dataset NYT, the proposed CoRA improves the prior state-of-the-art perfor- mance by a large margin in terms of Precision@N, AUC and Hits@K. Further analyses verify its superior capability in handling long-tail relations in contrast to the competitors. 1 Introduction Relation extraction, as a fundamental task in natural language processing (NLP), aims to discriminate the relation between two given entities in a plain text. As recent data-driven algorithms, e.g., deep neural networks (Han et al., 2018; Du et al., 2018; Ye and Ling, 2019), have shown their capabilities in tackling NLP tasks, labor-intensive annotation and scarce training data become the major obstacles of achieving promising performance on relation extraction. Instead of costly manual labeling, distant supervision method (Mintz et al., 2009) is proposed to auto- label the training data for relation extraction. It labels a sentence containing a pair of entities with the relation between them in a knowledge graph, e.g. Freebase (Bollacker et al., 2008). A strong assumption behind this is that a sentence containing two entities only expresses the relation existing in the knowledge graph, but this assumption will not always hold (Lin et al., 2016). Hence, it leads to two main problems regrading the training data. First, when the assumption is invalid, wrong labeling problem appears and degrades algorithms by introducing noisy supervision signals. This problem has been well-studied by recent works (Lin et al., 2016; Han et al., 2018; Ji et al., 2017; Li et al., 2020) operating at bag level for multi-instance learning (Hoffmann et al., 2011), where the “bag” denotes a set of sentences with the same entity pair. Most of them use selective attention to avoid wrongly-labeled sentences. Second, the long-tail problem …

Content truncated to 500 words