dc.contributor.advisor | Σκιαδόπουλος, Σπύρος | |
dc.contributor.author | Παρράς, Γεώργιος | |
dc.date.accessioned | 2024-08-27T09:24:34Z | |
dc.date.available | 2024-08-27T09:24:34Z | |
dc.date.issued | 2020-11-06 | |
dc.identifier.uri | https://amitos.library.uop.gr/xmlui/handle/123456789/8164 | |
dc.identifier.uri | http://dx.doi.org/10.26263/amitos-1666 | |
dc.description | Μ.Δ.Ε. 74 | el |
dc.format.extent | σελ. 104 | el |
dc.language.iso | en | el |
dc.publisher | Πανεπιστήμιο Πελοποννήσου | el |
dc.rights | Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα | * |
dc.rights.uri | http://creativecommons.org/licenses/by-nc-nd/3.0/gr/ | * |
dc.title | Παραλληλοποίηση αλγόριθμων για αποτελεσματικό φιλτράρισμα πληροφορίας | el |
dc.type | Μεταπτυχιακή διπλωματική εργασία | el |
dc.contributor.committee | Βασιλάκης, Κώστας | |
dc.contributor.committee | Τρυφωνόπουλος, Χρήστος | |
dc.contributor.department | Τμήμα Πληροφορικής και Τηλεπικοινωνιών | el |
dc.contributor.faculty | Σχολή Οικονομίας, Διοίκησης και Πληροφορικής | el |
dc.contributor.master | Πρόγραμμα Μεταπτυχιακών Σπουδών στην Επιστήμη και Τεχνολογία Υπολογιστών | el |
dc.description.abstracttranslated | In the information ltering paradigm, clients subscribe to a server with continuous queries that express their information needs. Such queries aim to retrieve relative documents that are published on the server. More speci cally, whenever a new document
is published on the server, the continuous queries satisfying this document are found and noti cations are sent to the respective clients. More formally, given a database of continuous queries db and an incoming document d, an information ltering process nds all queries q 2 db that match d. We concentrate on queries that are expressed in the AWP data model. This model is based on named attributes with values of type text, and its query language includes Boolean and word proximity operators. In this thesis, we consider the e cient parallelization of the information ltering procedures. To this end, we employ appropriate data structures, indexing methods and parallel techniques. Using the aforementioned machinery, our parallel methods achieve
an improvement of more than 98% in ltering performance for large databases (up to 3 million queries), expressed in the AWP model. | el |