Chinese question classification method based on text error correction and neural network

Abstract

The invention discloses a Chinese question classification method. The invention aims to solve the problem that the classification accuracy is not high enough due to the fact that faulty wording, wrongly written characters, needless characters and the like exist in input questions and the inherent defect that an existing classification method is single. The Chinese question classification method comprises the steps: 1, obtaining Chinese question text data; 2, preprocessing the Chinese questions; 3, carrying out error correction by utilizing a language model; 4, vectorizing the Chinese questionsby using a word vector tool; 5, obtaining an intermediate semantic matrix vector containing semantic information by utilizing the bidirectional gating circulation unit layer; 6, generating an attention matrix vector by using a self-attention mechanism; 7, extracting a plurality of local features by utilizing a plurality of convolution kernels with different sizes, and obtaining a global feature matrix vector through pooling and splicing; and 8, outputting probability distribution of a corresponding category by utilizing a full connection layer and a normalization exponential function, and taking the category with the maximum probability value as a predicted category, namely, a result of Chinese question classification. The Chinese question classification method is applied to the field ofnatural language processing.

Description

technical field [0001] The invention belongs to the field of intelligent information processing and computer technology, and in particular relates to a method for classifying Chinese question sentences. Background technique [0002] With the rapid development of technology in the Internet age and the influx of a large amount of data, searching for keywords through search engines requires manual screening of the returned results, which is time-consuming and labor-intensive for users. The question answering system can quickly obtain the user's intention, and can return the most accurate answer to the user among hundreds or thousands of candidate answers. [0003] The classification of Chinese questions is the first step of the question answering system, and it is one of the key technologies for the question answering system to achieve accurate answers. By classifying Chinese questions, the question answering system can effectively narrow the range of answers and determine the ...

AI Technical Summary

Problems solved by technology

Current research and applications have proved that based on recurrent neural networks such as long-term short-term memory and bidirectional gated recurrent units, it is better able to learn the upper and lower semantic information of Chinese questions, and based on convolutional neural networks, it is better able to learn local features in sentences and extract sentences in sentences. However, there are many kinds of classifications of Chinese questions, and the classification effect of a single classification method still cannot fully meet the requirements in practice, and there are few current researches that can make full use of the advantages of recurrent neural networks and convolutional neural networks.

[0005] In addition, most of the current methods of classifying Chinese questions using deep learning neural networks hardly consider the presence of language problems, multiple characters, and wrong characters in the input questions, and they are all used as the input of the model for prediction or training, which will lead to training problems. The model has a large bias for similar questions, resulting in wrong predictions

Images