Reaserch On Named Entity Recognition For Web Recruitment Text Based On Deep Learning

Named entity recognition of web recruitment text refers to identifying specific information from web recruitment text,such as skills,education,work experience and other requirements.In the era of big data with the explosion of information resources,job seekers need to spend a lot of time and effort to sort out the recruitment text to summarize the requirements of recruiting companies for talents.Named entity recognition and analysis of web recruitment texts can help job seekers grasp the specific requirements of job positions on the one hand,and have important implications for the construction of recommendation systems and knowledge graphs in the field of web recruitment on the other.Therefore,this paper takes web recruitment text as the research object to conduct the research of named entity recognition based on deep learning method.The main research content of this thesis are as follows.(1)The improved Bi GRU-ATT-CRF model is proposed to solve the problem that the traditional named entity recognition model is ineffective in recognizing entities for web recruitment text.Replacing the LSTM network with GRU network can reduce the complexity of the model and make the model more flexible by adding an attention mechanism.Compared with the traditional named entity recognition model,the entity recognition effect of Bi GRU-ATT-CRF model has been improved.(2)The RSBC model is proposed to address the problem that the traditional named entity recognition model is weak in extracting complex entity features,and the RSBC model increases the depth of the neural network in the entity recognition model to improve the ability of the model to extract deep features,and solves the network degradation problem that occurs with the increase of network depth through the residual connection.Compared with the traditional entity recognition model,the RSBC model further improves the entity recognition effect.(3)The BERT-Bi GRU-CRF model is proposed to solve the problem of incorrect entity type recognition due to the phenomenon of word polysemy in web recruitment text.The text representation is performed using word vectors generated based on the BERT pre-trained language model,and then the word vectors are trained by the Bi GRU-CRF model.Compared with the traditional named entity recognition model,the BERT-Bi GRU-CRF model achieves the best entity recognition results in web-recruited text.The experimental results show that the three models proposed in this thesis achieve the best experimental results by improving the F1 values by 4.93%,6.77% and 10.07%,respectively,compared with the traditional entity recognition model Bi LSTM-CRF.