"Neuvedeno."@en . . "1"^^ . "Inteligentn\u00ED anal\u00FDza prostoru WWW usnad\u0148uj\u00EDc\u00ED p\u0159\u00EDstup k\u00A0jeho obsahu je dnes v\u00A0ohnisku z\u00E1jmu v\u00FDzkumn\u00EDk\u016F z\u00A0oblasti vyhled\u00E1v\u00E1n\u00ED informac\u00ED (%22information retrieval%22) i znalostn\u00EDho in\u017Een\u00FDrstv\u00ED. V\u00A0prvn\u00EDm p\u0159\u00EDpad\u011B je web ch\u00E1p\u00E1n jako obrovsk\u00E1 kolekce textov\u00FDchdokument\u016F, kter\u00E9 lze reprezentovat formou vektor\u016F p\u0159\u00EDznak\u016F; za efektivn\u00ED p\u0159\u00EDstupovou metodu je pova\u017Eov\u00E1no zejm\u00E9na tzv. indexov\u00E1n\u00ED latentn\u00ED s\u00E9mantiky. Ve druh\u00E9m p\u0159\u00EDpad\u011B je kladen d\u016Fraz na form\u00E1ln\u011B-logick\u00E9 odvozov\u00E1n\u00ED a abstrakci od element\u00E1rn\u00EDch term\u00EDn\u016F astruktur ke konceptu\u00E1ln\u00EDm metadat\u016Fm; s\u00E9mantika pou\u017E\u00EDvan\u00FDch informac\u00ED je definov\u00E1na tzv. form\u00E1ln\u00EDmi ontologiemi. P\u0159edkl\u00E1dan\u00FD projekt usiluje o integraci obou pohled\u016F. Hlavn\u00ED uva\u017Eovan\u00E9 sm\u011Bry pr\u00E1ce zahrnuj\u00ED zejm\u00E9na matematickou formalizaci \u00FAlohy extrakcewebov\u00FDch metadat, konstrukci soustavy konceptu\u00E1ln\u00EDch ontologi\u00ED WWW, v\u00FDvoj agentov\u00E9 architektury pro anal\u00FDzu WWW (vyu\u017E\u00EDvaj\u00EDc\u00ED redundance a komplementarity informac\u00ED vyj\u00E1d\u0159en\u00FDch r\u016Fzn\u00FDmi typy dat), n\u00E1vrh strukturovan\u00E9 reprezentace webov\u00FDch dat na b\u00E1zi XML," . . . . . . . . . . . "Intelligent analysis of WWW content and structure"@en . . . . . "Inteligentn\u00ED anal\u00FDza obsahu a struktury WWW" . . . . "Intelligent analysis of the World-Wide Web content and structure is considered as major application area by both information retrieval and knowledge engineering research; the principal paradigms however differ. IR researchers view the web as an enormouscollection of text documents that can be represented as uniform vectors of features; the hugeness of this collection can be attacked by statistical techniques such as latent semantic indexing. KE researchers focus on logical inference and abstraction ofconceptual metadata from elementary terms and structures; the semantics is formalised by ontologies. The current project attempts to integrate both views. The main tasks addressed are: mathematical formalisation of the metadata extraction task, design ofan integrated collection of web ontologies, development of an agent architecture for web analysis (relying on the redundancy and complementarity of information expressed by different types of data), design of a structured XML-based representation of web"@en . "2009-01-15+01:00"^^ . "0"^^ . . . "C\u00EDlem projektu bylo vyvinout a integrovat soubor r\u016Fznorod\u00FDch metod aplikovateln\u00FDch na anal\u00FDzu obsahu a struktury WWW. Vedle \u010Dinnost\u00ED bezprost\u0159edn\u011B sm\u011B\u0159uj\u00EDc\u00EDch k\u00A0t\u00E9to aplikaci se projektov\u00FD t\u00FDm v\u011Bnoval i v\u00FDvoji metod s\u00A0obecn\u011Bj\u0161\u00EDm vyu\u017Eit\u00EDm, jako je efektiv"@cs . "The goal of the project was to develop and integrate a collection of heterogeneous methods applicable on analysis of web content and structure. Beside the activities directly aiming at this goal, more generic methods were also investigated, such as effec"@en . "GA201/03/1318" . "54"^^ . "1"^^ . "http://www.isvav.cz/projectDetail.do?rowId=GA201/03/1318"^^ . "54"^^ . .