Automatic text categorization of news articles

Abstract

To categorize the data reduces the access time. Nowadays, the Internet is one of the biggest data resources. However, most of the data on the Internet is written in natural language. To use the Internet more efficiently, it needs to be categorized. The amount of data and increment rate is so high that this process can not be done by hand. Hence, the necessity of automatic text categorization systems is increasing. In contrast to other languages, there is not much study on Turkish texts. In this study, a system is developed for automatic text categorization of news articles. The articles are classified into 5 different classes and 76% success ratio is achieved.