SENT: Sentence-level Distant Relation Extraction via Negative Training

Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, pages 6201–6213 August 1–6, 2021. ©2021 Association for Computational Linguistics 6201 SENT: Sentence-level Distant Relation Extraction via Negative Training Ruotian Ma1, Tao Gui2∗, Linyang Li1, Qi Zhang1∗, Xuanjing Huang1 and Yaqian Zhou1 1School of Computer Science, Fudan University, Shanghai, China 2Institute of Modern Languages and Linguistics, Fudan University, Shanghai, China {rtma19,tgui16,linyangli19,qz,xjhuang,yqzhou}@fudan.edu.cn Abstract Distant supervision for relation extraction pro- vides uniform bag labels for each sentence inside the bag, while accurate sentence labels are important for downstream applications that need the exact relation type. Directly using bag labels for sentence-level training will introduce much noise, thus severely degrading performance. In this work, we propose the use of negative training (NT), in which a model is trained using complementary labels regarding that “the instance does not belong to these complementary labels”. Since the probability of selecting a true label as a complementary label is low, NT provides less noisy information. Furthermore, the model trained with NT is able to separate the noisy data from the training data. Based on NT, we propose a sentence-level framework, SENT, for distant relation extraction. SENT not only filters the noisy data to construct a cleaner dataset, but also performs a re- labeling process to transform the noisy data into useful training data, thus further benefit- ing the model’s performance. Experimental re- sults show the significant improvement of the proposed method over previous methods on sentence-level evaluation and de-noise effect. 1 Introduction Relation extraction (RE), which aims to extract the relation between entity pairs from unstructured text, is a fundamental task in natural language processing. The extracted relation facts can benefit various downstream applications, e.g., knowledge graph completion (Bordes et al., 2013; Wang et al., 2014), information extraction (Wu and Weld, 2010) and question answering (Yao and Van Durme, 2014; Fader et al., 2014). A significant challenge for relation extraction is the lack of large-scale labeled data. Thus, distant ∗ Corresponding authors. ? ? Obama is the 44th president of the United States. ① Place_of_birth ② Employee_of ① Place_of_birth ② Employee_of Lived_in (unincluded label) The sentence bag of <Obama, United-States> Which label ? Obama was back to the United States yesterday. Obama United States Obama was born in the United States. Bag labels (We need) Sentence labels Figure 1: Two types of noise exist in bag-level labels: 1) Multi-label noise: the exact label (“place of birth” or “employee of”) for each sentence is unclear; 2) Wrong-label noise: the third sentence inside the bag actually expresses “live in” which is not included in the bag labels. supervision (Mintz et al., 2009) is proposed to gather training data through automatic alignment between a database and plain text. Such annotation paradigm results in an inevitable noise problem, which is alleviated by previous studies using multi- instance learning (MIL). In MIL, the training and testing processes are performed at the bag level, where a bag contains noisy sentences mentioning …

Content truncated to 500 words