# Recurrent neural networks with specialized word embeddings for health-domain named-entity recognition

arXiv: https://arxiv.org/abs/1706.09569 · HTML: https://mlsanity.com/paper/1706.09569/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 29 Jun 2017.

## Built on by (most cited first, 7 of 7)

- [Model Tuning or Prompt Tuning? A Study of Large Language Models for Clinical Concept and Relation Extraction](https://mlsanity.com/paper/2310.06239/) (2023)
- [Zero-shot information extraction from radiological reports using ChatGPT](https://mlsanity.com/paper/2309.01398/) (2023)
- [A Transformer-based Diffusion Probabilistic Model for Heart Rate and Blood Pressure Forecasting in Intensive Care Unit](https://mlsanity.com/paper/2301.06625/) (2023)
- [Clinical Concept and Relation Extraction Using Prompt-based Machine Reading Comprehension](https://mlsanity.com/paper/2303.08262/) (2023)
- [A Cross-institutional Evaluation on Breast Cancer Phenotyping NLP Algorithms on Electronic Health Records](https://mlsanity.com/paper/2303.08448/) (2023)
- [A Biomedical Entity Extraction Pipeline for Oncology Health Records in Portuguese](https://mlsanity.com/paper/2304.08999/) (2023)
- [InspectionV3: Enhancing Tobacco Quality Assessment with Deep Convolutional Neural Networks for Automated Workshop Management](https://mlsanity.com/paper/2505.16485/) (2025)

## Abstract (arXiv)

Background. Previous state-of-the-art systems on Drug Name Recognition (DNR) and Clinical Concept Extraction (CCE) have focused on a combination of text "feature engineering" and conventional machine learning algorithms such as conditional random fields and support vector machines. However, developing good features is inherently heavily time-consuming. Conversely, more modern machine learning approaches such as recurrent neural networks (RNNs) have proved capable of automatically learning effective features from either random assignments or automated word "embeddings". Objectives. (i) To create a highly accurate DNR and CCE system that avoids conventional, time-consuming feature engineering. (ii) To create richer, more specialized word embeddings by using health domain datasets such as MIMIC-III. (iii) To evaluate our systems over three contemporary datasets. Methods. Two deep learning methods, namely the Bidirectional LSTM and the Bidirectional LSTM-CRF, are evaluated. A CRF model is set as the baseline to compare the deep learning systems to a traditional machine learning approach. The same features are used for all the models. Results. We have obtained the best results with the Bidirectional LSTM-CRF model, which has outperformed all previously proposed systems. The specialized embeddings have helped to cover unusual words in DDI-DrugBank and DDI-MedLine, but not in the 2010 i2b2/VA IRB Revision dataset. Conclusion. We present a state-of-the-art system for DNR and CCE. Automated word embeddings has allowed us to avoid costly feature engineering and achieve higher accuracy. Nevertheless, the embeddings need to be retrained over datasets that are adequate for the domain, in order to adequately cover the domain-specific vocabulary.