The Semantic Web was initiated by Tim Berners-Lee a decade ago with promising ideas regarding the sharing of metadata and knowledge on the Web, complemented with reasoning services for advanced new applications.
Early developments originated important foundational results and a deeper understanding of the issues involved in this endeavour, while identifying important conclusions regarding future developments, namely that:
1. Ontologies are necessary and useful for knowledge representation and the formalisms developed, e.g. OWL, are powerful enough to capture existing modelling languages used in software engineering.
2. Rules are fundamental to overcome the limitations found in OWL and provide constructs that are more natural for software developers (e.g. Closed World Assumption as used in databases).
3. New knowledge-based systems are not isolated and need mechanisms to act, reason, and evolve in the Semantic Web environment.
Indeed, the growing availability of information requires the support of dynamic data and application integration, automation and interoperation of business processes and problem-solving in various domains, to enforce correctness of decisions, and to allow traceability of the knowledge used and of the decisions taken, such as in the following scenario:
The Customs service for any developed country assesses imported cargo for a variety of risk factors including terrorism, narcotics, food and consumer safety and tariff violations. Assessing this risk involves extensive knowledge about commodities, business entities, trade patterns, government policies and trade agreements. Ontological concepts may be used to classify aspects of an import or export, while (nonmonotonic) reasoning rules are required to encode rules, regulations, and classification policies. While some of this knowledge is relatively stable, much of it changes rapidly. Active rules are required to indicate which actions are to be taken, not only when a shipment is classified, but also due to the changes in the available knowledge.
In this scenario, ontologies provide the logical underpinning of intelligent access and information integration, while rules are used to represent regulations and declarative guidelines about information and actionable business policies. Whereas standardisation of ontology and rule languages (e.g. resp. OWL-2 and RIF) fosters large numbers of ontologies and rules with different levels of complexity and scale, current technologies are simply not adequate to support development for this new class of knowledge-rich applications. For example, attempts to combine ontologies with rules often face semantic difficulties in choosing among Open and Closed World Assumptions or computational problems where querying is NP at best [Eiter08,Motik10], often undecidable [Horrocks04]. Update operators for rules [Alferes00] and those for ontologies [Flouris08] follow different principles which, until recently, seemed irreconcilable. Active rules for the web lacked a declarative semantics [Marie10] making their combination with reasoning rules and ontologies harder.
This project aims to address these problems and develop a state-of-the-art platform and tools that provide efficient services for 1) querying and updating (possibly inconsistent) knowledge bases tightly integrating ontologies and reasoning rules originating from a diversity of sources - heterogeneous knowledge bases, and 2) event monitoring to automate the execution of active rules.
Ongoing since March 2012, concludes in February 2015.
Funding entity: Fundação Ciência e Tecnologia (MCTES).
Principal researcher: João Leite.
Researchers: Carlos Damásio, José Alferes, Matthias Knorr, Martin Slota.