Deep learning based text classification with Web Scraping methods

It is a huge repository of information used to meet the many needs of Internet users. The collection and processing of data in some internet sites in this information repository has become very important nowadays. In this study, categorical news headlines and summaries in a Turkish news agency site were collected by using web scraping methods and test data were classified by using “one hot encoding” method with vector learning methods and depth learning methods. A classification accuracy of 90% has been achieved.