Text Classification And Recommendation Research Based On Primary And Secondary School Books

With the rapid development of Internet, network information is becoming more and more, and much richer than before. Faced with such a huge amount of data, how to accurately and effectively obtain the information that we need in a limited time becomes one of the current research focuses. The discovery and development of relevant technologies such as data mining comes with this situation, as an important part of which, the text categorization is rapidly developed. Text categorization means a process through which large numbers of text will be automatically classified by using a variety of text processing technology. With this technology, the large amount of data text will be classified accurately, the artificial classification work will be reduced, and the overall classification efficiency will be improved.The text classification process is roughly divided into the text preprocessing, feature selection, weight calculation, text vectorization and training corpus formation classifier.The text classification and recommended study of primary and secondary school books are specially designed for the primary and middle school students who do not have much time under such circumstances. This article mainly hopes that through the use of text classification and relevant technologies combined with the digital library, the primary and secondary school students can be able to determine whether the book is needed in a very short time by classifying the books.At the same time the the relevant or similar books in the digital library will be as alternatives to be recommended back to primary and middle school students.This paper mainly has a brief introduction on the theory of text classification. And the designation and implementation of the system are also be put forward in this paper. Besides, on the choice of classification algorithm, a simple experiment analysis has been carried in this paper. The results of the naive Bayes algorithm and KNN algorithm experiments have also been compared. From these experiments, the KNN classification algorithm is proved to be a better choice as the research classification algorithm.System module design, similar to the process of text categorization, is also divided into preprocessing module, feature selection and weight calculation module, as well as the classification and recommendation module. Of each module, the design idea, function demand and implementation of the design will be made a detailed description in the article.The purpose of this study is in the service for primary and middle school students, saving their limited spare time and helping students gain valuable meaningful books classification and recommendation, so as to improve students’ reading in their after-school life.Based on the analysis of design achievements, to some extent, the classifier can be used to classify and recommend materials accurately.