This project aims to apply existing Artificial Intelligence techniques for Natural-language Generation in the development of a voice assistant app, whose main goal is to provide media coverage in local elections. The application would be able to inform of any of the more than 8000 towns in Spain, as well as answer user’s questions about these matters. The language models GPT-2, T5 and BERT have been used for text generation tasks, text summarization and question answering throughout the development of the project, besides many online platforms such as DialogFlow or HuggingFace. All the code has been written in Python.
This was a final project of the Telecommunications Engineering Degree cursed at University of Seville, by Sergio Olivares (https://idus.us.es/handle/11441/126873). Outstanding mark.
