Method for automatically extracting metaphor sentences and anthropomorphic sentences of primary school Chinese composition based on Word2Vec and recurrent neural network

The invention belongs to the field of education informatization, and provides a method for automatically extracting metaphor sentences and anthropomorphic sentences of a primary school Chinese composition based on Word2Vec and a recurrent neural network, which comprises the following steps of: firstly, preprocessing a data set subjected to artificial marking to obtain a data set meeting experimental requirements; and then, training a Word2Vec model in a mode of combining the Skip-Gram model with the HierachicalSoftmax, so that a Word2Vec model can be obtained; and finally, representing the text by using the trained model, and training the corresponding recurrent neural network classifier by taking the text as the input of the recurrent neural network classifier. The invention provides theautomatic extraction method of metaphor sentences and anthropomorphic sentences in a composition, and provides an extraction scheme of the metaphor sentences and the anthropomorphic sentences for constructing a primary school