Educational and Non-educational Text Classification Based on Deep Gaussian Processes

With the development of the society, more and more people are concerned about education, such as preschool education, primary and secondary education and adult education. These people want to retrieve educational contents from large amount of information through the Internet. From the technical view, this requires identifying educational and non-educational data. This paper focuses on solving the educational and non-educational text classification problem based on deep Gaussian processes (DGPs). Before training the DGP, word2vec is adopted to construct the vector representation of text data. Then we use the DGP regression model to model the processed data. Experiments on real-world text data are conducted to demonstrate the feasibility of the DGP for the text classification problem. The promising results show the validity and superiority of the proposed method over other related methods, such as GP and Sparse GP.