Catálogo de publicaciones - libros

In this paper we present a new fixpoint-based approach to bottom-up state generation for stratifiable disjunctive deductive databases. To this end, a new consequence operator based on hyperresolution is introduced which extends Minker’s operator for positive disjunctive Datalog rules. In contrast to already existing model generation methods our approach for efficiently computing perfect models is based on state generation. Additionally, it enhances model state computation based on Minker’s operator for positive disjunctive Datalog rules.

This paper provides an answer-oriented multidimensional analysis environment. The approach is based on a conceptual point of view. We define a conceptual model that represents data through a constellation of facts and dimensions and we present a query algebra handling multidimensional data as well as multidimensional tables. Based on these two propositions, in order to ease the specification of multidimensional analysis queries, we define a formal graphical language implemented within a prototype: GraphicOLAPSQL.

Although the native (tree-like) storage of XML data becomes more and more important there will be an enduring demand to manage XML data in its textual representation, for instance in relational structures or file systems. XML data has to be wellformed by definition and additionally, in many cases, it has to be valid according to a given XML schema. Because the XML column types are often derived from text types (e.g. CLOBs) guaranteeing well-formedness as well as validity is not trivial. And even worse, for frequently modified data it is usually too expensive to re-validate the whole XML data after each update – but waiving re-validation may lead to inconsistencies and malfunctions of applications. In this paper we present a schema-aware pushdown automaton (i.e. a stack machine) that validates an XML string/stream. Using an element/state-index, the pushdown automaton is able to re-validate local modifications of the data while guaranteeing overall validity. Update operations (e.g. SQLXML, XQuery updates) are validated before executing them.

This paper describes a new XML compression scheme that offers both high compression ratios and short query response time. Its core is a fully reversible transform featuring substitution of every word in an XML document using a semi-dynamic dictionary, effective encoding of dictionary indices, as well as numbers, dates and times found in the document, and grouping data within the same structural context in individual containers. The results of conducted tests show that the proposed scheme attains compression ratios rivaling the best available algorithms, and fast compression, decompression, and query processing.

We are interested in evaluating the performance of new matching algorithms for user notification in digital libraries (DL). We consider a subscription system which continuously evaluates queries over a large repository containing document descriptions. The subscriptions and the document descriptions rely on a taxonomy that is a hierarchically organized set of terms. The digital library supports insertion, update and removal of a document. Each of these operations is seen as an that must be notified only to those users whose subscriptions match the document’s description. The paper proposes a notification algorithm dedicated to taxonomy-based DLs, addresses computational issues and report a full set of experiments illustrating the advantages of the approach.

The exponential growth of the web and the extended use of database management systems in widely distributed information systems has brought to the fore the need for seamless interconnection of diverse and large numbers of information sources. Our contribution is a system that provides a flexible approach for integrating and transparently querying multiple data sources, using a reference ontology. Global semantic queries are automatically mapped to queries local to the participating sources. The query system is capable of handling complex join constructs and of choosing the appropriate attributes, relations, and join conditions to preserve user query semantics. Moreover, the query engine exploits information on horizontal, vertical, and hybrid fragmentation of database tables, distributed over the various data sources. This optimization improves system’s recall and boosts its effectiveness and performance.