Creación de un prototipo de chatbot que permita interactuar con la historia del Ecuador registrada en periódico antiguos

Abstract

History shows us an endless series of events that occur daily in the world, leaving their mark on time. In past eras, this knowledge was transmitted orally, preserved through generations. However, with the advancement of technology, the way we access information has changed dramatically, allowing us to explore historical records with unprecedented depth and scope. However, there is the issue of preserving and accessing historical information contained in old newspapers from Ecuador using advanced technologies. Newspapers have been an in- valuable source of historical knowledge, but much of this legacy has been lost due to adverse factors. Digitization and the use of Optical Character Recognition (OCR) have significantly improved the preservation and access to these documents, although there are still limitations in terms of handling and searching large volumes of data. In this context, this thesis work proposes an effective solution for the preservation and access to historical information through digitization and the use of an advanced chatbot. The positive results obtained underscore the importance of integrating modern technologies into historical research and pave the way for future improvements and applications in various fields. The proposed solution involves developing a chatbot capable of understanding natural language queries and providing detailed answers about historical events in Ecuador. This chatbot enhances interaction with the extensive knowledge network derived from digitized newspapers, integrating generative AI models such as ChatGPT and LLAMA 3.