Research On Some Problems Of Formal Ontology Engineering

Ontology is a term borrowed form philosophy, where an Ontology is a systematic account of Existence. After imported into Information Technology domain, ontology acts as"a formal specification of a shared conceptualization"and is applied to everywhere of Information Technology domain. With the development of ontology theory and application, a new research field called"Ontology Engineering"appears. Ontology Engineering is formally defined as"the set of activities that concern the ontology development process, the ontology life cycle, and the methodologies, tools and languages for building ontologies".Although ontology engineering receives extensive attention, it is still in its elementary phrase of research and application. There are many important problems waiting to be solved, which have blocked the fast development of ontology engineering. Based on the analysis and summary of the stat-of-the-art of ontology engineering research, this paper provides an idea of formal ontology engineering. Formal Ontology Engineering emphasizes to make the best use of the characters of formal technology, such as rigorous, verifiable, inferable and automatic, to improve formalization of ontology engineering activities, to realize the automation of ontology engineering activities, to ensure the controllable procedures and the verifiable results. We do some research on some problems of formal ontology engineering, such as formal representation of ontology, ontology security, ontology modeling, ontology evolution, ontology slicing and ontology application in this paper and get some achievements which improve application of formal technology to ontology engineering. To be more specific, works are done in the following aspects:1. To provide a process calculi based formal representation for ontology As a knowledge representation, there are many formal representations for ontology, such as set theory, logics, and ontology languages. Each representation has its character and application domain. An ontology representation based on a process calculus– Bigraph theory is provided in this paper. A descript logic (DL) for ontology representation is described in bigraph, and the Tableau algorithm for reasoning the descript logic is described in reactive rules of Bigraph Reactive System (BRS) in this paper. The new representation bridges ontology and process calculi theoretically. It provides theoretical base for research ontology engineering on process calculi.2. To do some research on process calculi based secure ontology engineering Bigraphical semantics of a secure process calculus– Seal calculus is provided in this paper. Static representations of Seals in bigraphs and translation rules from Seals to bigraphs are provided, and then structure correspondence and operational correspondence between Seal and its bigraph representation are provide, thus the bridge between Seal and Bigraph is built, and more, the relation between security calculus theory and ontology theory is built. They provide theoretical base for research ontology security theory based on process calculi.3. To provide an Ontology Definition Metamodel based ontology modeling methodResearchers found that the two standalone domains, ontology engineering and software engineering, have similarities, and research achievements of the two domains can be referenced each other, so they suggest to bridge these two domains. Object Management Group (OMG) adopts Ontology Definition Metamodel (ODM) as a standard for this goal. This paper provides an Ontology Definition Metamodel based ontology modeling method in context of Model Driven Architecture (MDA) of software engineering. We also implement a visualized ontology modeling tool prototype using software engineering tools. The modeling tool solves some problems of current ontology development tools, such as professional requirement, complex operation, non-friendly UI, and inextensible architecture.4. To do some research on ontology evolution1) To provide a model transformation based ontology evolution framework. OMG's ODM bridges ontology engineering and software engineering, thus a new research field called Model Driven Semantic Web (MDSW) appears. In MDSW, many ontology problems can be solved in software engineering technologies. Using model transformation technologies of software engineering, a QVT-based ontology evolution framework is provided in this paper. Under the framework, ontology evolution can be looked as a process of model transformation. Structure of the ontology evolution framework, ontology procedure model and applications of QVT in each phrase of evolution procedure are also provides in this paper.2) To provide a ripple-effect analysis method for ontology evolution. Ontology evolution can affect instances and applications depend on original ontologies, and more complex, ontologies that depend on original ontologies, their associated instances and applications can also be affected. This kind of effect is called ripple-effect. In this paper, an ontology graph model, which creates ontology adjacency matrix and ontology reachability matrix, is described. Depending on matrix shift and calculation, ripple-effect of ontology evolution can be analyzed and its quantity can be ascertained. Every ripple-effect caused by ontology change operation is described. At the same time, approaches for calculating ontology element contribution, ontology cohesions and effect degree applied to every ontology element during ontology evolution, are provided. Ripple-effect analysis for dependency-unknown ontology evolution is discussed, and a server model for ontology evolution is also provided. All are credible foundation for management, control, usage and evaluation of ontology evolution, and are foundation for ontology evolution automation calculation in computer.5. To provide a dependency graph based ontology slicing methodSlicing is a method that can extract required data segments from operated data with internal relations according some special criteria. Program slicing and model slicing are two familiar slicing techniques. By introducing the idea of these slicing techniques into ontology engineering, we provide an ontology slicing method in this paper. In the method, an ontology dependency graph is derived from Ontology Definition Metamodel (ODM), and then ontology slices are generated automatically according slicing criteria given by users. This method can be applied to many domains, such as ontology reasoning, ontology evolution, ontology merging, ontology modularizing and so on.6. To do some research on ontology engineering application1) To provide an Ontology Definition Metamodel based consistency checking of UML models. We first derive a mapping for MOF to World Wide Web Consortium (W3C)'s Web Ontology Language (OWL) from ODM, and then a representation of UML metamodel in OWL based on this mapping is provided. After transforming UML models to OWL instances, we can check consistency of these models by using logical reasoning and query techniques. Thus, we implement a semantic level consistency checking method for UML models.2) To provide a Description Logic ontology based checking method for software process models. An approach of representing OMG's Software Process Engineering Metamodel (SPEM) in Description Logic (DL) is provided in this paper, and an example of using it to represent and reason a concrete software process model– a portion of DMR Macroscope is presented, too. This approach is built on two mappings. One is a mapping from Meta-Object Facility (MOF) to DL and the other is a mapping from Object Constraint Language (OCL) to DL. Using those two mappings, OMG's SPEM consisted of a metamodel that is an MOF extension and constraints represented in OCL can be mapped to a representation in DL. Thus, software process models based on SPEM can be represented in DL, and process analysis and manipulation techniques, such as reasoning, consistency checks can benefit from DL techniques.The research mentioned in this paper solves some problems of formal ontology engineering in some degree. The achievements will provide valuable reference for ontology engineering research and application, and will provide theoretical base for formalization and automation of ontology engineering activities.