Automatický rešeršní systém

Abstract

This thesis focuses on the design and implementation of a system for automated processing of scientific documents in the Portable Document Format (PDF). The primary objective is to develop a solution capable of extracting text, identifying and analyzing visual content, and generating summaries using advanced artificial intelligence techniques. The system leverages technologies for document structure recognition, text element analysis, and image data processing. The outcome is a modular library that seamlessly integrates with a web-based interface, providing an intuitive way to work with scientific texts. The thesis also includes a comparison of the proposed approach with existing methods and an evaluation of its applicability for future research applications.