Research on English Corpus Tagging Based on BiLSTM model

Abstract

The recognition and tagging of special words in English corpus can effectively improve students' learning efficiency. Based on BiLSTM model and CRF model, a BiLSTM-CRF model model is constructed to recognize and automatically label special words in English corpus. The results show that the average accuracy of BiLSTM-CRF model is 95.35% and the average recall rate is 94.83%, which are much higher than other models. We can know from the above that BiLSTM-CRF model can label English professional corpora well and is a practical method.