An Intelligent Named Entity Recognition Method Based on IoT Professional Knowledge

In order to improve the recognition ability and accuracy of professional entities in the field of Internet of things, an intelligent named entity recognition method based on Internet of things professional knowledge is proposed. This method combines the relevant professional knowledge involved in the actual teaching process of the Internet of things, constructs the professional knowledge base of the Internet of things in the form of question-and-answer pairs, proposes a new node triplet model, and labels the corpus. Based on Bidirectional Encoder Representation from Transformers, bi-directional long short-term memory, Conditional Random Field (BERT-BiLSTM-CRF) model, Multi-Head Self-Attention module is added to establish Bidirectional Encoder Representation from Transformers, bi-directional long short-term memory, Multi-Head Self-Attention, Conditional Random Field (BERT-BiLSTM-MHSA-CRF) model. The $F_{1}$ value of the entity in the Internet of things professional knowledge base identified by the model reaches 95.23 % BERT-BiLSTM-MHSA-CRF can more accurately identify the professional knowledge named entity involved in the Internet of things teaching.