National Technical University of Athens (NTUA)
Αλγόριθμοι επανεγγραφής τροποποιημένων ερωτημάτων για βατές περιγραφικές λογικές
Abstract
dc:descriptionThe reasoning problem of answering conjunctive queries over Description Logic knowledge bases via query rewriting has gained a lot of attention in the last few years. Given a conjunctive query and a knowledge base (TBox and ABox) a query rewriting procedure computes a new query that incorporates the constraints of the knowledge base that are described in its TBox, such that for any ABox (dataset) the evaluation of the initial query over the TBox and the ABox can be computed by evaluating only the new query over the ABox. Because the computational complexity in expressive Description Logics is prohibitive tractable Description Logics languages have been developed, such as the DL-Lite family, EL and the Datalog+- family for which there have been presented many rewriting algorithms/systems. However, all these algorithms run from scratch without taking into consideration previous runs, even if successive queries are pretty similar, which is a very common scenario in the Web.In this thesis we study the problem of query rewriting for queries that have been refined in various ways. These are the addition and removal of distinguished variables or atoms. More precisely, we study the problem of computing the rewriting of a query that has been refined taking into account the initial query without having to recompute it from scratch. In the following we present various optimisations that drastically increase the performance of our algorithms. Moreover, we limit the ontology language expressivity to DL-LiteR in order to further optimise the problem of query rewriting for queries that have been refined by the addition of an atom. Interestingly, our approach implies a novel incremental algorithm for computing the rewriting of a fixed query. More precisely, the query can be `decomposed' into its atoms and then incrementally process each atom. We present detailed algorithms as well as several optimisations that as depicted by our experimental evaluation improve the performance of our system and show that it is faster than all the other systems.Finally in this thesis we study the problem of query answering over an ontology network. Namely, given two ontologies and a set of mappings among them we study how Fuzzy Sets and Fuzzy Logic can be used to provide semantics for the mappings so that we can resolve inconsistencies that might arise from them. After resolving these inconsistencies a new (fuzzy) ontology that is consistent is created with an ABox that contains assertions, interpreted using the fuzzy mappings, from both the initial ontologies.
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2014
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Venetis, Anastasios
- Βενέτης, Αναστάσιος
Subjects
dc:subject × 20- Σημασιολογικός ιστός
- Περιγραφικές λογικές
- Πρόσβαση σε δεδομένα μέσω οντολογιών
- Επανεγγραφή ερωτήματος
- Τροποποιημένο ερώτημα
- Συζευκτικά ερωτήματα
- Semantic web
- Description logics
- Ontology based data access
- Query rewriting;
- Refined query
- Conjunctive query
- Φυσικές Επιστήμες
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Natural Sciences
- Computer and Information Sciences
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/39429