There is a growing need for scalable semantic web repositories which support inference and provide efficient queries. There is also a growing interest in representing uncertain knowledge in semantic web datasets and ontologies. In this paper, I present a bit vector schema specifically designed for RDF (Resource Description Framework) datasets. I propose a system for materializing and storing inferred knowledge using this schema. I show experimental results that demonstrate that this solution simplifies inference queries and drastically improves results. I also propose and describe a solution for materializing and persisting uncertain information and probabilities. Thresholds and bit vectors are used to provide efficient query access to this uncertain knowledge. My goal is to provide a semantic web repository that supports knowledge inference, uncertainty reasoning, and Bayesian networks, without sacrificing performance or scalability.
Beyond its usual industrial fields of application, a current body of research explores the use of constraint based configuration to address general AI problems, like for instance automatic composition of semantically enriched web services (SWS). A configuration request is naturally formulated as a fragment of the desired solution, that the configurator will attempt to complete according to constraints. We address here a case where the design of the configuration request may itself be the result of a configuration phase, that helps the user design the request by formulating it on more abstract grounds. Within this framework, the configurator is first used to complete an abstract request formulated in a specific formalism. Then a translation is performed from the goal model to the final model to yield the actual request sent to the second configuration phase. This research builds on previous experience showing the adequacy of using configuration to compose SWS, that raised further issues regarding the nature of queries.
Description Logics (DLs) provide a clear and broadly accepted paradigm for modeling and reasoning about terminological knowledge. However, it has been often noted, that although DLs are well-suited for representing a single, global viewpoint on an application domain, they offer no formal grounding for dealing with knowledge pertaining to multiple heterogeneous viewpoints — a scenario ever more often approached in practical applications, e.g. concerned with reasoning over distributed knowledge sources on the Semantic Web. In this paper, we study a natural extension of DLs, in the style of two-dimensional modal logics, which supports declarative modeling of viewpoints as contexts, in the sense of McCarthy, and their semantic interoperability. The formalism is based on two-dimensional semantics, where one dimension represents a usual object domain and the other a (possibly infinite) domain of viewpoints, addressed by additional modal operators and a metalanguage, on the syntactic level. We systematically introduce a number of expressive fragments of the proposed logic, study their computational complexity and connections to related formalisms.
The machine understandable semantic of information, achieved by using an RDF(S) structure and common-shared vocabularies (ontologies) is the big step in enabling the machine-agent interoperability on the Web. Machine agents can crawl annotated web pages, search for useful information from various sources, use the information to solve tasks at hand by using the internal reasoning mechanism and background knowledge. In order to enhance their inference capabilities, machine- (and also human-) agents need to update their knowledge, using relevant knowledge sources as much as possible. One of the possible scenarios is to search for relevant knowledge on the (Semantic) Web. In this paper we discuss the prerequisites for design, and present an approach for representing rules in the machine understandable form, which is based on the current efforts in achieving the machine understandable semantic of information. Such representation of rules can serve as the backbone for a web-enabled knowledge management process. In the presented usage scenario we focus on the knowledge sharing phase in that process, i.e. on the searching for relevant knowledge (rules) on the Web.
Service-oriented architectures have brought significant progress for more flexible realization of business processes integrating functionality from heterogeneous sources. While more and more businesses adopt the new technology it becomes obvious that many questions are still not addressed to make it keep its promises, especially in the area of human efforts involved in business process composition. We introduce a framework for a possible next generation enterprise software based on, but going beyond that of service-oriented architectures utilizing logic programming taking advantage of formalized explicit policies as substantial constituents of enterprise systems.