Noise Mitigation for Neural Entity Typing and Relation Extraction

Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers, pages 1183–1194, Valencia, Spain, April 3-7, 2017. c⃝2017 Association for Computational Linguistics Noise Mitigation for Neural Entity Typing and Relation Extraction Yadollah Yaghoobzadeh* and Heike Adel* and Hinrich Sch¨utze * These authors contributed equally to this work Center for Information and Language Processing LMU Munich, Germany yadollah|heike@cis.lmu.de Abstract In this paper, we address two different types of noise in information extraction models: noise from distant supervision and noise from pipeline input features. Our target tasks are entity typing and rela- tion extraction. For the first noise type, we introduce multi-instance multi-label learn- ing algorithms using neural network mod- els, and apply them to fine-grained entity typing for the first time. Our model outper- forms the state-of-the-art supervised ap- proach which uses global embeddings of entities. For the second noise type, we propose ways to improve the integration of noisy entity type predictions into re- lation extraction. Our experiments show that probabilistic predictions are more ro- bust than discrete predictions and that joint training of the two tasks performs best. 1 Introduction Knowledge bases (KBs) are important resources for natural language processing tasks like ques- tion answering and entity linking. However, KBs are far from complete (e.g., Socher et al. (2013)). Therefore, methods for automatic knowledge base completion (KBC) are beneficial. Two subtasks of KBC are entity typing (ET) and relation extraction (RE). We address both tasks in this paper. As in other information extraction tasks, obtain- ing labeled training data for ET and RE is chal- lenging. The challenge grows as labels become more fine-grained. Therefore, distant supervision (Mintz et al., 2009) is widely used. It reduces the need for manually created resources. Distant su- pervision assumes that if an entity has a type (resp. two entities have a relationship) in a KB, then all sentences mentioning that entity (resp. those two entities) express that type (resp. that relation- ship). However, that assumption is too strong and gives rise to many noisy labels. Different tech- niques to deal with that problem have been in- vestigated. The main technique is multi-instance (MI) learning (Riedel et al., 2010). It relaxes the distant supervision assumption to the assumption that at least one instance of a bag (collection of all sentences containing the given entity/entity pair) expresses the type/relationship given in the KB. Multi-instance multi-label (MIML) learning is a generalization of MI in which one bag can have several labels (Surdeanu et al., 2012). Most MI and MIML methods are based on hand crafted features. Recently, Zeng et al. (2015) in- troduced an end-to-end approach to MI learning based on neural networks. Their MI method takes the most confident instance as the prediction of the bag. Lin et al. (2016) further improved that method by taking other instances into account as well; they proposed MI learning based on selective attention as an alternative way of relaxing the im- pact of noisy labels on RE. In …

Content truncated to 500 words