Text Classification in Architecture Field Based on Naive Bayes Algorithm

Abstract

With the rapid development of the Internet, it has become a problem to quickly and accurately find the information we need in the massive text information. And the method to solve the problem is text classification. This paper proposes a text classifier for the architectural field. The feature selection adopts the TF-IDF text classification method, which effectively reduces the computing time and creates conditions for improving the computing power of the text classifier. In the classification stage, Naive Bayes classifier, K nearest neighbour classifier and decision tree classifier are introduced for comparative analysis. Experiment shows that the prediction results of Naive Bayes are significantly improved compared with KNN and decision tree. Among the three groups of classifiers, the Naive Bayes classifier has the best performance. The classifier can effectively improve the recall rate and accuracy rate of the text classification in architectural field, and has a good use effect.