关键词: 深度学习, 自编码, 神经网络, 招生, 文本分类

Abstract: College enrollment consultation service usually relies on the artificial reply or keyword matching Q&A system, that exists the problems of low efficiency and irrelevant answers. In addition, a consultation text is often a short statement, therefore its vectorization representation may easily lead to the high-dimensional sparse problem. To solve the problems mentioned above, we propose an enrollment consultation algorithm based on Stacked Denoising Sparse Autoencoders(SDSAE). First of all, to improve generalization ability of the algorithm, an autoencoders network is used to extract the feature and reduce the data dimension, dataset enhancement technique and noise-adding technique are introduced to solve the size problem of training sample set and uneven classification. After the low dimensional representation of the short text being obtained, a text classification using BP algorithm is conducted afterwards. From the experimental results, it can be seen that the proposed algorithm has a better classification performance over BP, SVM, ELM algorithm and etc., it significantly improve the classification effect of enrollment consultant texts, it provides a new way for the design of intelligent consultation system, and has a good application prospect in the field of college enrollment consultant.

Key words: deep learning, autoencoder, neural network, enrollment, text classification