Research and Application on Text Classification Model Based on Keywords

Abstract

Text classification is the process of assigning predefined labels to text by its content. It is a common task of Natural Language Processing (NLP). The traditional way for text classification is machine learning and its effect is greatly depended on the amount and accuracy of the training data set which is difficult to obtain in most cases. The job of building the training data set is inefficient and expensive [1]. This has motivated a research word to break this barrier, with a method using keywords instead to complete text classification. In this work, we explore the usage of label-keywords pairs (each label has a set of keywords) for assigning text documents to one or more categories automatically even without the training data set, obtaining results comparable to those systems that classify the text manually.