A method development based on machine learning techniques for text classification

Text classification is a process of including the text in predefined categories in order to organize it. In recent years, the increase in data in the digital environment has accelerated the work on extracting information from the data and increasing the availability of information. Automatic text classification studies offer very effective solutions in understanding and organizing the text. The problem of text classification is solved by machine learning methods within the scope of this thesis. When machine learning methods are used during text classification, values representing text are determined by natural language processing techniques. The features that represent the text are categorized as word feature vector, name entities feature vector, and part of speech feature vector, meanwhile the effects on classifications of these categorizations have been examined. Additionally, during the classification the success of ensemble methods has been observed by putting different machine learning algorithms together. The use of words, part of speech units and name entities all together as a feature vector was found to have a positive effect on success.