Alkalmazott algoritmusok nagyméretű feladatokra = Applied algorithms for large-scale problems

Alap es alkalmazott kutatast vegeztunk a kovetkező fő teruleteken: - Formalis matematikai modszerek adatbanyaszatban es optimalizalasban; - Nagymeretű adatok elemzese es modellezese, halozatokkal kapcsolatos uzleti intelligencia alkalmazasokban; - Felhasznalo es tartalom osszerendelese, kereses, ajanlas. A projekt resztvevői zart lancban a teljes innovacios lancot lefedik az oktatastol (ELTE es BME algoritmusok, adatbanyaszat, Web informacio-kereses előadasok) az elmeleti kutatasokon at az alkalmazasokig. A kutatashoz kapcsolodo legfontosabb ket ipari partnerunk a Magyar Telekom es az AEGON, amelyek szamara egyedi kereső megoldasokat fejlesztettunk, naploelemzesi es ugyfel-elemzesi feladatokat oldottunk meg. Europai kapcsolataink segitsegevel a jelen kutatasi eredmenyekre epulő Digitalis Konyvtarak es Biztonsag temaju projektben veszunk reszt. A kutatasunk nemzetkozi elismertseget jelzi, hogy felkertek a legjelentősebb europai adatbanyaszati verseny, az ECML/PKDD Discovery Challenge szervezesere, illetve a legrangosabb World Wide Web konferencian Workshop Chair, a WSDM (Web Search and Data Mining) konferencian szenior, tovabbi kapcsolodo temaju konferencian es workshopon (ICALP, AIRWeb, ESA stb) programbizottagi tagot adunk. Legfontosabb eredmenyeink: - Előrelepest a veges testek feletti polinomfelbontas algoritmusaiban; - Dijnyertes megoldas a KDD Cup 2009 feladaton; - Uj Web Spam szűrő modszerek; - Tartalom alapu kepkereső eljarasok. | Our results cover a wide range of areas of theory and application: -Formal mathematical methods in data mining and optimization; -Analysis and modeling very large scale data with applications in the areas of network related business intelligence; -User-content interaction, optimization. The project team covers full innovation chain from Education (Technical University and Eotvos University courses in algorithms, data mining, Web information retrieval), Pure, Applied Research and Innovation. Our industrial exploitation include the Hungarian Telecom Group and AEGON Hungary where we developed custom search engines and conducted log mining and business intelligence projects. Based on the reported results, we participated in several Digital Libraries and Security ICT projects. Our results are acknowledged by being the main organizer of the major European data mining contest, the ECML/PKDD Discovery Challenge 2010 and the invitation to serve as Workshop Chair at the highest prestige World Wide Web conference, senoir program committee member at the Web Search and Data Mining conferences, and PC member of other related conferences and workshops (ICALP, AIRWeb, ESA etc). Our most important research results include -Breakthrough algorithms in factorization of polynomials over finite fields; -Prize winner solution at KDD Cup 2009, in a telco classification task; -New methodologies in Web Spam filtering; -Content-based multimedia indexing methods.