GLaRA: Graph-based Labeling Rule Augmentation for Weakly Supervised Named Entity Recognition

Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics, pages 3636–3649 April 19 - 23, 2021. ©2021 Association for Computational Linguistics 3636 GLARA: Graph-based Labeling Rule Augmentation for Weakly Supervised Named Entity Recognition Xinyan Zhao∗ Haibo Ding† Zhe Feng† ∗School of Information, University of Michigan, MI 48109, USA † Bosch Research and Technology Center, Sunnyvale, CA 94085, USA zhaoxy@umich.edu, {haibo.ding,zhe.feng2}@us.bosch.com Abstract Instead of using expensive manual annotations, researchers have proposed to train named en- tity recognition (NER) systems using heuris- tic labeling rules. However, devising labeling rules is challenging because it often requires a considerable amount of manual effort and do- main expertise. To alleviate this problem, we propose GLARA, a graph-based labeling rule augmentation framework, to learn new label- ing rules from unlabeled data. We first create a graph with nodes representing candidate rules extracted from unlabeled data. Then, we de- sign a new graph neural network to augment labeling rules by exploring the semantic re- lations between rules. We finally apply the augmented rules on unlabeled data to generate weak labels and train a NER model using the weakly labeled data. We evaluate our method on three NER datasets and find that we can achieve an average improvement of +20% F1 score over the best baseline when given a small set of seed rules. 1 Introduction Named entity recognition (NER) models often need to be trained with many manual labels to perform well. Due to the high cost of manual annotations, collecting labeled data to train NER models is chal- lenging for real-world applications. Recently, re- searchers have proposed to collect weak labels using heuristic rules, which are called labeling rules (Bach et al., 2017; Fries et al., 2017; Rat- ner et al., 2020; Safranchik et al., 2020). This kind of methods typically first ask domain experts to write labeling rules for a NER task, then use these manual rules to generate labeled data and train a NER model with the weakly labeled data. The ad- vantage of these methods is that they do not require manual annotations. However, during our study, we find that writing labeling rules is also challeng- ing for domain-specific tasks. Devising accurate def rule1( x ): return “Disease” if x.PreBigram==“associated with” else “Other” Rule-1: associated_with_* → Disease def rule2( x ): return “Disease” if x.PreBigram == “cause of” else “Other” Rule-2: cause_of_* → Disease The symptoms were associated with enzyme deficiency. The fragile site is not associated with mental retardation. Migraine is an uncommon cause of cranial neuropathy. The cause of hearing loss after spinal anaesthesia is unknown. matching matching semantic relation Figure 1: Two examples of rules for recognizing Diseases. PreBigram means the 2 tokens on the left of a candidate entity. rules often demands a significant amount of manual effort because it requires developers to have deep domain expertise and a thorough understanding of the target data. To alleviate manual effort on writing label- ing rules, we propose GLARA, a Graph-based Labeling Rule Augmentation framework …

Content truncated to 500 words