Prediktivní analytické postupy pro výběry a předzpracování geodat.

Abstract

When developing geoapplications, it is necessary to solve a number of potential problems associated with the users input geospatial data, which can be very diverse not only in terms of content, but on a global scale, in particular by the format, the coordinate system, the character encoding and the amount of information about the data itself (metadata). The absence of such information often leads to data being damaged or inapplicabled. The aim of the thesis is to use the statistical information from already prepared correct geospatial data and its analysis containing records of geodata parameters, geolocation of users and their requirements, to select the most probable variant of the missing input geospatial data parameter, to predict the following user requirements and thus to facilitate the user management of geospatial data in the given geoapplication.