A Bidirectional LSTM and Conditional Random Fields Approach to Medical Named Entity Recognition

Medical named entity recognition is a fundamental and essential research for medical natural language possessing, aiming to identifying medical concepts or terminology such as diseases, drugs, treatments, procedures, etc. from unstructured medical text. A model based on a bidirectional LSTM and conditional random fields (Bi-LSTM-CRF) is proposed for medical named entity recognition. Our model contains three layers and relies on character-based word representations learned from the supervised corpus. BiLSTM-CRF model can learn the information characteristics of a given dataset. Experiments on a publically available NCBI Disease Corpus as an evaluation standard dataset shows our approach achieves a 0.8022 F1 measure, which outperforms a number of widely used baseline methods.

[1]  Hua Xu,et al.  Research and applications: A comprehensive study of named entity recognition in Chinese clinical text , 2014, J. Am. Medical Informatics Assoc..

[2]  Zhiyong Lu,et al.  NCBI disease corpus: A resource for disease name recognition and concept normalization , 2014, J. Biomed. Informatics.

[3]  Jürgen Schmidhuber,et al.  Learning to Forget: Continual Prediction with LSTM , 2000, Neural Computation.

[4]  José Luís Oliveira,et al.  A modular framework for biomedical concept recognition , 2013, BMC Bioinformatics.

[5]  Hua Xu,et al.  A study of active learning methods for named entity recognition in clinical text , 2015, J. Biomed. Informatics.

[6]  Zhiyong Lu,et al.  TaggerOne: joint named entity recognition and normalization with semi-Markov Models , 2016, Bioinform..

[7]  Jeffrey Dean,et al.  Efficient Estimation of Word Representations in Vector Space , 2013, ICLR.

[8]  Haibin Liu,et al.  Natural language processing pipelines to annotate BioC collections with an application to the NCBI disease corpus , 2014, Database J. Biol. Databases Curation.

[9]  Ramón Fernández Astudillo,et al.  Not All Contexts Are Created Equal: Better Word Representations with Variable Attention , 2015, EMNLP.

[10]  Zhiyong Lu,et al.  DNorm: disease name normalization with pairwise learning to rank , 2013, Bioinform..

[11]  W. John Wilbur,et al.  Meshable: searching PubMed abstracts by utilizing MeSH and MeSH-derived topical terms , 2016, Bioinform..

[12]  Guillaume Lample,et al.  Neural Architectures for Named Entity Recognition , 2016, NAACL.

[13]  Pascal Vincent,et al.  Representation Learning: A Review and New Perspectives , 2012, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[14]  Tao Chen,et al.  Disease named entity recognition by combining conditional random fields and bidirectional recurrent neural networks , 2016, Database J. Biol. Databases Curation.

[15]  Zhiyong Lu,et al.  An improved corpus of disease mentions in PubMed citations , 2012, BioNLP@HLT-NAACL.

[16]  Alexander A. Morgan,et al.  Evaluation of text data mining for database curation: lessons learned from the KDD Challenge Cup , 2003, ISMB.

[17]  Jürgen Schmidhuber,et al.  Framewise phoneme classification with bidirectional LSTM and other neural network architectures , 2005, Neural Networks.

[18]  Lei Liu,et al.  Extracting important information from Chinese Operation Notes with natural language processing methods , 2014, J. Biomed. Informatics.

[19]  Saad Alanazi,et al.  A Named Entity Recognition System Applied to Arabic Text in the Medical Domain , 2017 .

[20]  Devanshu Jain,et al.  Supervised Named Entity Recognition for Clinical Data , 2015, CLEF.

[21]  Wang Ling,et al.  Finding Function in Form: Compositional Character Models for Open Vocabulary Word Representation , 2015, EMNLP.

[22]  Graciela Gonzalez,et al.  BANNER: An Executable Survey of Advances in Biomedical Named Entity Recognition , 2007, Pacific Symposium on Biocomputing.

[23]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[24]  Benjamin M. Good,et al.  Microtask Crowdsourcing for Disease Mention Annotation in PubMed Abstracts , 2014, Pacific Symposium on Biocomputing.

[25]  Wei Xu,et al.  Bidirectional LSTM-CRF Models for Sequence Tagging , 2015, ArXiv.

[26]  Min Song,et al.  PKDE4J: Entity and relation extraction for public knowledge discovery , 2015, J. Biomed. Informatics.