Скачать 103.94 Kb.
Department of Computer Science
University of Georgia, Athens, GA 30602
Department of Computer Science
University of Georgia, Athens, GA 30602
This paper presents the notion of Semantic Associations as complex relationships between resource entities. These relationships capture both a connectivity of entities as well as similarity of entities based on a specific notion of similarity called -isomorphism. It formalizes these notions for the RDF data model, by introducing a notion of a Property Sequence as a type. In the context of a graph model such as that for RDF, Semantic Associations amount to specific certain graph signatures. Specifically, they refer to sequences (i.e. directed paths) here called Property Sequences, between entities, networks of Property Sequences (i.e. undirected paths), or subgraphs of -isomorphic Property Sequences.
The ability to query about the existence of such relationships is fundamental to tasks in analytical domains such as national security and business intelligence, where tasks often focus on finding complex yet meaningful and obscured relationships between entities. However, support for such queries is lacking in contemporary query systems, including those for RDF.
This paper discusses how querying for Semantic Associations might be enabled on the Semantic Web, through the use of an operator . It also discusses two approaches for processing -queries on available persistent RDF stores and memory resident RDF data graphs, thereby building on current RDF query languages.
Categories and Subject Descriptors
H.2.3 [Information Systems]: Database Management–Query Languages
Languages, Theory, Management
Semantic Web Querying, Semantic Associations, RDF, Complex Data Relationships, graph traversals.
The Semantic Web  proposes to explicate the meaning of Web resources by annotating them with metadata that have been described in an ontology. This will enable machines to “understand” the meaning of resources on the Web, thereby unleashing the potential for software agents to perform tasks on behalf of humans. Consequently, significant effort in the Semantic Web research community is devoted to the development of machine processible ontology representation formalisms. Some success has been realized in this area in the form of W3C standards such as the eXtensible Markup Language (XML)  which is a standard for data representation and exchange on the Web, and the Resource Description Framework (RDF) , along with its companion specification, RDF Schema (RDFS) , which together provide a uniform format for the description and exchange of the semantics of web content. Other noteworthy efforts include OWL , Topic Maps , DAML+OIL . There are also related efforts in both the academic and commercial communities, which are making available tools for semi-automatic  and automatic  semantic (ontology-driven and/or domain-specific) metadata extraction and annotation.
With the progress towards realizing the Semantic Web, the development of semantic query capabilities has become a pertinent research problem. Semantic querying techniques will exploit the semantics of web content to provide superior results than present-day techniques which rely mostly on lexical (e.g. search engines) and structural properties (e.g. XQuery ) of a document. There are now a number of proposals for querying RDF data including RQL , SquishQL , TRIPLE , RDQL . These languages offer most of the essential features for semantic querying such as the ability to query using ontological concepts, inferencing as part of query answering, and some allow the ability to specify incomplete queries through the use of path expressions. One key advantage of this last feature is that users do not need to have in-depth knowledge of schema and are not required to specify the exact paths that qualify the desired resource entities. However, even with such expressive capabilities, many of these languages do not adequately support a query paradigm that enables the discovery of complex relationships between resources. The pervasive querying paradigm offered by these languages is one in which queries are of the form: “Get all entities that are related to resourceA via a relationship R” where R is typically specified as possibly a join condition or path expression, etc. In this approach, a query is a specification of which entities (i.e. resources) should be returned in the result. Sometimes the specification describes a relationship that the qualifying entities should have with other entities, e.g. a join expression or a path expression indicating a structural relationship. However, the requirement that such a relationship be specified as part of the query is prohibitive in domains with analytical or investigative tasks such as national/homeland security  and business intelligence, where the focus is on trying to uncover obscured relationships or associations between entities and very limited information about the existence and nature of any such relationship is known to the user. In fact, in this scenario the relationship between entities is the subject of the user’s query and should being returned as the result of the query as opposed to be specified as part of the query. That is, queries would be of the form “How is Resource A related to Resource B?”. For example, a security agent may want to find any relationship between a terrorist act and a terrorist organization or a country known to support such activities.
One major challenge in dealing with queries of this nature is that it is often not clear exactly what notion of a relationship is required in the query. For example, in the context of assessing flight security, the fact that two passengers on the same flight are nationals of a country with known terrorist groups and that they have both recently acquired some flight training, may indicate an association due to a similarity. On the other hand, the fact that a passenger placed a phone call to someone in another country that is known to have links to terrorist organizations and activities may indicate another type of association characterized by connectivity. Therefore, various notions of “relatedness” should be supported.
This paper intends to make two main contributions. First, we formalize a set of complex relationships for the RDF data model, which we call Semantic Associations. Second, we outline two possible approaches for processing queries about Semantic Associations through the use of an operator (-Queries). One of the two approaches is based on processing -queries on persistent RDF data systems such as RDFSuite , while the other is based on processing these queries on a main memory based representation of an RDF model such as JENA .
The rest of the paper is organized as follows: Section 2 discusses some background and motivates our work with the help of an example. Section 3 presents the formal framework for Semantic Associations, section 4 discusses implementation strategies for the operator, section 5 reviews some related work, and section 6 concludes the paper.
|Задача: Предоставляет пользователю унифицированную среду содержащую сервисы и функциональности, варьирующиеся от информационных сервисов и сервисов извлечения |
По аналогии с Semantic Web, Semantic Grid может быть определен как расширение современных grid, где информация и сервисы имеют четкий...
|Semantic Web : a guide to the Future of xml, Web Services, and Knowledge Management|
|Bayesian semantics for the semantic web||Valentina Janev, Sanja Vranes, Semantic Web Tools and Technologies for Competence Management, Lambert Academic Publishing GmbH, isbn: 978-8454-4166-5, Sarbrucken, Germany, 2011|
|Интеллектуальное реферирование: онтологический подход и его реализация в решениях Ontos|
Обсуждению вопросов создания системы реферирования под управлением онтологий, разработанной в рамках решений Ontos для Semantic Web,...
|Smart Semantic Middleware for Ubiquitous Computing”|
|Semantic Execution Environment (see) Background and Related Work||Semantic Execution Environment (see) Background and Related Work|
|Frome a genome database to a semantic knowledge base||Grush The semantic challenge to computational neuroscience (draft V 4) Page|