Toward Mention Detection Robustness with Recurrent Neural Networks

Nguyen, Thien Huu; Sil, Avirup; Dinu, Georgiana; Florian, Radu

doi:10.48550/arxiv.1602.07749

Cited by 4 publications

(5 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Some studies [87]- [89] employ word-level representation, which is typically pre-trained over large collections of text through unsupervised algorithms such as continuous bagof-words (CBOW) and continuous skip-gram models [90] (see Figure 4 for the architectures of CBOW and skip-gram).…”

Section: Word-level Representationmentioning

confidence: 99%

“…The dictionary contains 205,924 words in 600 dimensional vectors. Nguyen et al [87] used word2vec toolkit to learn word embeddings for English from the Gigaword corpus augmented with newsgroups…”

Section: Word-level Representationmentioning

confidence: 99%

“…The work by Huang et al [16] is among the first to utilize a bidirectional LSTM CRF architecture to sequence tagging tasks (POS, chunking and NER). Following [16], a body of works [17], [18], [87], [88], [93]- [95], [99], [104], [108], [109] applied BiLSTM as the basic architecture to encode sequence context information. Yang et al [105] employed deep GRUs on both character and word levels to encode morphology and context information.…”

Section: Conditional Probability Models For Taggingmentioning

confidence: 99%

See 2 more Smart Citations

A Survey on Deep Learning for Named Entity Recognition

Sun

Han³

et al. 2022

IEEE Trans. Knowl. Data Eng.

966

397

View full text Add to dashboard Cite

Named entity recognition (NER) is the task to identify text spans that mention named entities, and to classify them into predefined categories such as person, location, organization etc. NER serves as the basis for a variety of natural language applications such as question answering, text summarization, and machine translation. Although early NER systems are successful in producing decent recognition accuracy, they often require much human effort in carefully designing rules or features. In recent years, deep learning, empowered by continuous real-valued vector representations and semantic composition through nonlinear processing, has been employed in NER systems, yielding stat-of-the-art performance. In this paper, we provide a comprehensive review on existing deep learning techniques for NER. We first introduce NER resources, including tagged NER corpora and off-the-shelf NER tools. Then, we systematically categorize existing works based on a taxonomy along three axes: distributed representations for input, context encoder, and tag decoder. Next, we survey the most representative methods for recent applied techniques of deep learning in new NER problem settings and applications. Finally, we present readers with the challenges faced by NER systems and outline future directions in this area.

show abstract

Section: Word-level Representationmentioning

confidence: 99%

Section: Word-level Representationmentioning

confidence: 99%

Section: Conditional Probability Models For Taggingmentioning

confidence: 99%

See 1 more Smart Citation

A Survey on Deep Learning for Named Entity Recognition

Sun

Han³

et al. 2022

IEEE Trans. Knowl. Data Eng.

966

397

View full text Add to dashboard Cite

show abstract

“…Compared to Conv-CRF, the BiLSTM-CRF model is more robust and relies less on distributed representations of inputs. Subsequently, a series of works also adopted BiLSTM as the context encoder, such as Lample et al [3], Chiu et al [4], Nguyen et al [5], Zheng et al [6], Ma et al [7], and Zheng et al [8]. Yang et al [9] adopted the GRU-CRF method and simultaneously encoded the word vector and word vector of the context.…”

Section: Related Workmentioning

confidence: 99%

A Multimodal Named Entity Recognition Model for Power Equipment Based on Deep Neural Network

Zhang,

Song,

Zhao

et al. 2023

Advances in Transdisciplinary Engineering

View full text Add to dashboard Cite

Digital empowerment of China’s power energy sector is a key factor in increasing its economic and social benefits, and named entity recognition technology is the most fundamental and core task of information extraction technology in the digital empowerment process. Therefore, we propose a multimodal named entity recognition model PE-MNER for power equipment based on deep neural networks. Compared to text multimodality, text and image multimodality can use image information to supplement missing information in the text, thus enabling more accurate entity extraction. The model first obtains a BERT neural network through incremental training, and then extracts Chinese character features through the network. Then, a hierarchical visual prefix fusion network is used to fuse image information. From the comparative experimental results, it can be seen that the proposed model has the best performance compared to the benchmark model, with an improvement of 4.08%∼7.20% in the F1 score compared to the benchmark model.

show abstract

“…They have been adapted to learn vector representations of words for NLP-based phenotyping [112,136], laying a foundation for computational phenotyping. Deep learning has been applied on various NLP applications, including semantic representation [146], semantic analysis [147,148], information retrieval [149,150], entity recognition [151,152], relation extraction [153][154][155][156], and event detection [157,158]. Beaulieu-Jones et al [136] developed a neural network approach to construct phenotypes to classify patient disease status.…”

Section: Deep Learningmentioning

confidence: 99%

Natural Language Processing for EHR-Based Computational Phenotyping

Zeng

Deng

et al. 2019

IEEE/ACM Trans. Comput. Biol. and Bioinf.

164

View full text Add to dashboard Cite

This article reviews recent advances in applying natural language processing (NLP) to Electronic Health Records (EHRs) for computational phenotyping. NLP-based computational phenotyping has numerous applications including diagnosis categorization, novel phenotype discovery, clinical trial screening, pharmacogenomics, drug-drug interaction (DDI) and adverse drug event (ADE) detection, as well as genome-wide and phenome-wide association studies. Significant progress has been made in algorithm development and resource construction for computational phenotyping. Among the surveyed methods, well-designed keyword search and rule-based systems often achieve good performance. However, the construction of keyword and rule lists requires significant manual effort, which is difficult to scale. Supervised machine learning models have been favored because they are capable of acquiring both classification patterns and structures from data. Recently, deep learning and unsupervised learning have received growing attention, with the former favored for its performance and the latter for its ability to find novel phenotypes. Integrating heterogeneous data sources have become increasingly important and have shown promise in improving model performance. Often better performance is achieved by combining multiple modalities of information. Despite these many advances, challenges and opportunities remain for NLP-based computational phenotyping, including better model interpretability and generalizability, and proper characterization of feature relations in clinical narratives.

show abstract

Toward Mention Detection Robustness with Recurrent Neural Networks

Cited by 4 publications

References 26 publications

A Survey on Deep Learning for Named Entity Recognition

A Survey on Deep Learning for Named Entity Recognition

A Multimodal Named Entity Recognition Model for Power Equipment Based on Deep Neural Network

Natural Language Processing for EHR-Based Computational Phenotyping

Contact Info

Product

Resources

About