In order for mobile applications to efficiently offer all of their services and adjust them according to the end user, they would need to monitor all of his context and surroundings. In the existing researches this is done by a special context server which is responsible for data retrieval and transmission of information application. In the event, when the context server would work on a mobile device, the before mentioned system would gradually not be needed anymore. In the context of the thesis we have successfully implemented the proposed model of context server and tested it with a context application, which operates in the principle of text classification of user entries in the search engines and its current location. The application of this data does not need to be captured manually, since they are automatically transmitted to the server when the new data appears.