Named Entity Recognition of Chinese Intangible Cultural Names Based on Part-of-Speech Tagging

By observing many intangible cultural heritage names, it is found that although the frequency of occurrence of these names is relatively low, they are often composed of multiple high-frequency words. Further analysis of the parts-of-speech distribution of these high-frequency words shows that there is a certain regularity. Consequently, a named entity recognition model based on POS tagging was presented in this study. By including the POS into the word vector, the word vector s information was increased and the name recognition accuracy for intangible cultural assets was enhanced. Compared with BERT-BiLSTM-CRF, the F1 score of the proposed model in the name recognition task of intangible cultural heritage items is increased from 0.854 to 0.933, an increase of 0.079.