Research On Chinese Question Classification Method Based On Convolutional Neural Network

With the rapid development of computer technology,people can obtain the information that they want from the mass information.This system can better meet the people's retrieval requirements compared with the traditional keyword based search.In question answering system,problem analysis is the foundation,and it is the foundation to classify modules in problem analysis.The question and classification can reduce the candidate space of the answer and reduce the time of finding the answer,and the quality of the question and answer system determines the overall quality and performance of the whole question answering system.So in the question and answer system,the natural language questions that people put forward are processed to make the machine understand the human language,and the question is excavated and analyzed to provide the most desired answer to the human.Automatic question answering system is an important research direction in the field of Natural Language Processing.The main task of the system is matching the user questions with knowledge library,and then providing some accurate feedbacks to users.At present,the after-sales service has gradually become one of the core departments of many enterprises and the scale of customer service is expanding with the enlargement of service users.The efficient and practical automatic question answering system will help to liberate a large number of human resources.This paper proposes two sentence classification methods,which are a word vectorbased Chinese question classification method and a Chinese multi-pool convolutional neural network-based Chinese question classification method.(1)The word-vector based Chinese question classification method uses word2 ve.Tools to automatically learn and extract eigenvectors and use low-dimensional vector space representations can solve dimensionality catastrophes caused by sparse dimensions and can also tap the associated attributes between words and words.Improve the semantic accuracy of the vector,while also reducing the complexity of the model.(2)Research on Chinese Question Classification Algorithm for Dynamic Multicell Convolutional Neural Networks.The training-derived distributed word vectors are used to initialize the input layer of the DMCNN,further extract features,train a classification model based on the network,and then use the classifier to classify the input questions.Experiments compare the traditional machine learning algorithm with the cnn model of deep learning random generation vector and the method proposed in this paper.Finally,the experimental results show that the method has improved the accuracy of Chinese question classification.