Abstract
A problem of the web searching is how to gain only relevant information and how to ensure its completeness. We discusse architectures of search engines, whose mechanisms provide necessary metainformation to make both problems at least partially solvable.