Σχολή Επιστημών και Τεχνολογίας της Πληροφορίας
Μόνιμο URI για αυτήν την κοινότηταhttps://pyxida.aueb.gr/handle/123456789/2
Η Σχολή Επιστημών και Τεχνολογίας της Πληροφορίας περιλαμβάνει τα Τμήματα: - Τμήμα Πληροφορικής - Τμήμα Στατιστικής
Περιήγηση
Πλοήγηση Σχολή Επιστημών και Τεχνολογίας της Πληροφορίας ανά Θέμα "++Spicy"
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω
Τώρα δείχνει 1 - 1 από 1
- Αποτελέσματα ανά σελίδα
- Επιλογές ταξινόμησης
Τεκμήριο Schema matching, schema mapping and data exchange: modification and extension of the open source tool, ++SpicyVouziou, Evelyn; Athens University of Economics and Business, Department of Informatics; Vassalos, VasiliosData Integration is the broad and complex scientific field, which copes with semantic heterogeneity, via techniques of adaptive and/or federated query processing, operatingon architectures, which vary from a materialized repository or data warehouse, to a completely virtual infrastructure. Overall, Data Integration aims to increase the completeness, conciseness, and correctness of data that is available to users and applications; fact, which describes the need for quality driven modeling of the data to be integrated and the procedures, which would best achieve integration. There are many both academic and commercial tools, addressing the challenging enterprises’ environment or particular information technology domains. At the core of Data Integration lie the processes of schema matching, (as the generation of semantic correspondences between heterogenous data), schema mapping (as the consolidation and translation of these correspondences for the unification over the source and target schemata) and data exchange (as the detection of the best schema target and the processing of schema mappings to executable queries on the source). Spicy, a research project at University of Basilicata, Italy, performed simultaneous schema matching and mapping, letting the user validate the mappings via the interface (helping her choose among the different candidate mappings), using a first generation schema mapping algorithm. Schema mappings are expressed as sets of tuple generated dependencies, and schemata were relational or nested relational. Spicy also required source and target instances, enriching the user’s reasoning on the mapping task. Spicy later evolved into the +Spicy tool, which also performs rewriting of the tuple generating dependencies, eliminating redundancy, generating runtime scripts; for an instances input, the output is core solutions. Later, ++Spicy was developed, and provided as an open source tool, which uses equality generating dependencies, as target functional dependencies –constraints for the generation of executable scripts, and provides global solutions to the data exchange problem for other nested data models as well, apart from the relational and nested relational. In this thesis, ++Spicy was modified and extended, by the implementation of various functional requirements, such as: the optional (previously mandatory) import of source instances, the full operability support over extra types of data sources (in CSV and SQL files format), the import of correspondences in a conjunctive way, graphically and via an editor. Aiming at the tasks’ implementation, refactoring and redesign were necessary, also allowing its further integration with other tools. The in depth study of, and hands on experience with ++Spicy also led to the tool’s evaluation in terms of software engineering, useful for further work regarding redesign and implementation.