Artificial Intelligence-based Text Generation

OData support
Supervisor:
Dr. Szűcs Gábor
Department of Telecommunications and Media Informatics

Artificial Intelligence is present in more and more usage areas nowadays. In informatics there are various applications dealing with languages. In the category of artificial languages, programming languages are widespread and have been used for a long time now. However, in the last years, processing natural languages with machines have been evolving fast. One area that focuses on natural language, is Natural Language Generation that is performed with Artificial Intelligence methods.

This thesis focuses on Natural Language Generation methods. First, the required literature is studied through the most important concepts and tendencies. This is followed by providing an insight into application areas by two tasks and their combination. These tasks are: a translator between two foreign languages and a poem generator. Their combination is also presented in this thesis which generates a poem based on an input provided in a different language. In the practical examples I focused on deep learning networks, because the tasks have been completed with these methods; at the end of the thesis an evaluation part is included using both automatic metrics and human evaluation techniques.

Downloads

Please sign in to download the files of this thesis.