Named entity recognition of power communication planning based on transformer

In order to improve the intelligence level of power communication network planning, a power communication planning named entity method based on Transformer is proposed to solve the problems of long text in power communication planning report and low information extraction efficiency. This paper takes the power communication planning text as the research object, uses the transformer model to pre-train the words in the planning text to obtain word vectors, and then uses the Bi-lstm-crf model to train with the pre-trained word vectors as input to screen the appropriate model. parameter to complete named entity recognition. Compared with the experimental simulation, the model named entity recognition accuracy using the transformer model pre-trained word vector is higher.