Coreference Aware Representation Learning for Neural Named Entity Recognition

Abstract

Recent neural network models have achieved state-of-the-art performance on the task of named entity recognition (NER). However, previous neural network models typically treat the input sentences as a linear sequence of words but ignore rich structural information, such as the coreference relations among non-adjacent words, phrases or entities. In this paper, we propose a novel approach to learn coreference-aware word representations for the NER task at the document level. In particular, we enrich the well-known neural architecture ``CNN-BiLSTM-CRF'' with a coreference layer on top of the BiLSTM layer to incorporate coreferential relations. Furthermore, we introduce the coreference regularization to ensure the coreferential entities to share similar representations and consistent predictions within the same coreference cluster. Our proposed model achieves new state-of-the-art performance on two NER benchmarks: CoNLL-2003 and OntoNotes v5.0. More importantly, we demonstrate that our framework does not rely on gold coreference knowledge, and can still work well even when the coreferential relations are generated by a third-party toolkit.

Context-Aware Feature Attention Model for Coreference Resolution on Biomedical Texts

“…Bio-entity Coreference resolution is a crucial task for artificial intelligence systems to be capable of fully understanding the biomedical texts by improving the performance of several downstream tasks, including information extraction [1], entity linking [2], question answering [3] and so on.…”

“…We further record the frequency ranges and the Pos (part-of-speech) tags of these identical mentions. The results illustrate that mentions with a frequency greater than 100 occupy 1 School of Computer Science and Technology, Xi'an Jiaotong University, Xi'an, Shaanxi, P.R. China Full list of author information is available at the end of the article more than half.…”

“…bedding and head word embedding, we use BioBERT semantic types, we employ MetaMapLite [1] to identify…”

“…Where ⊕ is the concatenation operation and f (x u ) is 38 a linear function to map x u to the same dimension with 39 [1] https://metamap.nlm.nih.gov/MetaMapLite.shtml the feature vector. a j,u is the weight of each feature based on the contexts and ϕ * is the new reweighed feature vectors.…”

End-to-end Deep Reinforcement Learning Based Coreference Resolution

“…Coreference resolution is one of the most fundamental tasks in natural language processing (NLP), which has a significant impact on many downstream applications including information extraction (Dai et al, 2019), question answering (Weston et al, 2015), and entity linking (Hajishirzi et al, 2013). Given an input text, coreference resolution aims to identify and group all the mentions that refer to the same entity.…”