Information Extraction from Web Pages Based on Their Visual Representation

This research is dedicated to enhancing the efficiency of web information extraction and web accessibility. The motivation behind the research, its aim and objectives are presented, and the performed work on developing web page model for information extraction is described. We also present work on making extracted information accessible to blind users, providing them with the means to navigate and access required information quickly. We also present our ongoing research on creating efficient methods and approaches for information extraction from the proposed model. There are two main approaches considered: 1) development of the library which provides required functionality to the programmer; 2) development of declarative Datalog-like language for information extraction.