Ontology Modeling Of Marine Ecology From Device-function Viewpoint

The21st century is said to be the century of ocean. The combination ofintelligent information processing technology and marine science gives rise to a newresearch field, called “digital marine”. Many researches (e.g., those on energy,ecology, environment, and so on) have been carried out by many countries for a longtime. Throughout these researches, there left many valuable documents such asobservational data, experimental results, papers, and patents as well, which provide anabundant data foundation for marine science research.However, these documents are hardly shared or reused sufficiently amonginstitutes due to the differences both in the format and in teleological or taxonomicalviewpoints. Ontology as “a formal, explicit specification of a sharedconceptualization” captures related domain knowledge, and provides a shared andcommon understanding of a specific domain. The difference between ontology andother concept model is that ontology takes a strictly consistent viewpoint to specifyconcepts, and the hierarchical (is-a) or mereological (part-of) relations betweenconcepts as well. That is, ontology specifies a semantic-consistent agreement to whichall parties should commit, and thus can be used as a middleware to facilitate datasharing and cooperation among these institutes.Marine ecosystems, including oceans, salt marsh, estuaries, coral reefs, the deepsea and the sea floor, are very important for the overall health of both marine andterrestrial environments. Ecological protection and crisis management arecontinuously receiving increased attention in the worldwide marine communityincluding researchers, governments, aqua culturists and other agencies. Certainly,ontology modeling from a fixed viewpoint is crucial for the cooperation among allthese parties.Energy flow and matter cycle are two basic processes in marine ecosystem.Involved in these two processes are various biological factors, non-biological factors, and the interactions among all these factors. Referring to engineering field literaturethat the “function” concept is used to describe these processes, this paper fixes aviewpoint of device-function, which looks processes in matter cycle as functions, andphysical entities (including marine lives and ocean itself) as devices, the agent whichcarries out such functions.By fixing such a viewpoint, this paper presents an upper ontology for marineecology, and then constructs a conceptual model for marine ecosystem. An extensionto OWL is also made for formal representation of the ontology. Based on the formalrepresentation and SWRL rules, a marine ecological knowledge management systemis developed, which provides users a semantic support platform for marine scientificresearch and cooperation. This paper focuses on the following five topics.(1) By studying ontology modeling theory and methodology, a knowledge engineeringapproach is proposed to construct marine ecological concept model and formalontology.(2) According to marine ecological knowledge characteristics, a marine ecologicalknowledge organization model which includes static structure and dynamic functionknowledge is proposed. This paper fixes a viewpoint of device-function, andconstructs a marine ecological upper ontology, and thus provides a unified viewpointfor the systematization of marine ecological knowledge and corresponding researches.(3) By extending the functional knowledge representing framework proposed byKitamura, a functional model of marine ecosystem is built, and then based on the roletheory proposed by Mizoguchi, a marine ecosystem conceptual model (includingmarine biology ontology, marine environment ontology, physical metabolic materialontology, marine hydrological ontology) is constructed.(4) Due to OWL lacks of enough role expressive ability, it is difficult to expresscomplicated dynamic functional knowledge by using the standard OWL. Byextending OWL syntax and semantic, this paper proposes OWL-Process model, andthen constructs marine ecological function ontology and its instances; In Protégéplatform, marine ecological upper ontology and marine ecological knowledgeontology are formally coded in OWL-Process; According to marine ecological abnormal phenomenon principle, this paper constructs the corresponding SWRL rulesets.(5) Based on marine ecological ontology knowledge base, this paper researchessemantic retrieval and ontology rule reasoning, and develops a marine ecologicalknowledge management system, which implements marine ecological knowledgenavigation, semantic query, knowledge retrieval and ecological crisis warningfunction. Through enriching SWRL rule sets, strengthening ontology rule reasoning,this research tries to provide marine ecological crisis warning more accurately, and toplay positive role in marine ecological environmental protection and marineecological crisis warning.