Biomedical named entity recognition based on recurrent neural networks with different extended methods

Biomedical Named Entity Recognition bio-NER has become essential to the text mining and knowledge discovery tasks in biomedical field. However, the performance of traditional NER systems is limited to the construction of complex hand-designed features which are derived from various linguistic analyses and may only adapted to specified domain. In this paper, we mainly focus on building a simple and efficient system for bio-NER based on Recurrent Neural Network RNN where complex hand-designed features are replaced with word embeddings. Furthermore, the system is extended by the predicted information from the prior node and external context information topical information & clustering information. During the training process, the word embeddings are fine-tuned by the neural network. The experiments conducted on the BioCreative II GM data set demonstrate RNN models outperform CRF model and Deep Neural Networks DNNs and the extended RNN model performs better than the original RNN, achieving 82.47% F-score.