Research On Named Entity Recognition Of Biological Pathogens Based On Neural Networks

Nowadays,public health and safety is the focus of common concern of all mankind.In this paper,named entity recognition of biological pathogens is an important basic task in the field of knowledge extraction in the field of biological pathogens.But at present,in the face of a large number of biological pathogen literature,researchers need to spend a lot of manpower and time on feature design,the selection of feature engineering will affect the performance of the model.At the same time,due to the diversity of entity names in the field of biological pathogens,the problem of inconsistent context annotation in the same chapter is more prominent.In response to the above problems,this paper first introduces the combination of bidirectional long short-term memory(BiLSTM)and conditional random field(CRF)as the basic model of biological pathogen named entity recognition task.By designing the model structure,the BERT model is introduced as a pre-training model.Fine-tune and extract text features and input them into the model.At the same time,the attention mechanism is introduced to calculate the correlation between the current word and the whole word through the weighted processing in the whole chapter,and obtain the assigned attention weight.Finally,the sequential annotation algorithm is completed by the conditional random field layer to identify the biological pathogen named entity name.In this paper,the BERT-Attention-BiLSTM-CRF biological pathogen named entity recognition model eliminates the need to manually select features and embed word vectors of another training model,which improves the consistency of entity recognition in the same chapter abstract on biological pathogen named entity recognition.In the experimental part,In the experimental part,we compared several groups,and proved that the BERT-Attention-BiLSTM-CRF model improved the f-value index on the test set.Through neural network and natural language processing technology,we can quickly and efficiently extract a large number of knowledge from biological pathogen literature,and extract the text corpus of biological pathogens in biomedicine,so as to lay a foundation for the prevention and control of biological pathogens.