Research On Constructing Ontology Based On Health-care System

With the development of Internet technology and the current environment of tending toimplement informatization worldwide, the hospitals aim at realizing informatization on thepersonnel management and medical management. With the informatization, the resources andthe medical cases in the hospital can be uniformly documented, and some statistical analysiscan be made on the large amount of medical cases, in order to be taken into consideration ofthe hospital’s development and arrangement by the hospital execs.Ontology is defined as a formal explicit specification of a specific field conceptualization.It contains a formal received glossary about the field, as well as the relationship between theconcepts. In this paper, a method on constructing field ontology by extracting informationfrom relational databases is proposed. It is based on the function of description logics and therelationship between relational database and ontology. With analysis on the relationaldatabase, a series of mapping concepts methods is proposed to construct partial ontology. Andthen, By analyzing the modules in the databases, the relationship between the higher conceptsand basic concepts is added to the partial ontologies. Using techniques of description logicsand concept lattices, two or more partial ontologies are combined on the concept level as aglobal ontology. At last, according to the administrators’ flow customization, medical flowsare embedded to the global ontology to adjust the global ontology as the medical fieldontology.The main research work in this paper is listed as follows:1. Constructing mapping rules between relational database and partial ontologyWe first analyze the meaning of the relation、property and constraint in relationaldatabase, and the concept、property and constraint in ontology accordingly, and then construct the mapping rules to generate the partial ontology. These partial ontologies contains the basicconcepts and the related properties and constraints.2. Analysis on the extraction of the relationship among the ontology conceptsAs the resource databases are legacy databases, a large amount of views and triggersexist in the database. These modules also contain the relationship between the concepts. Therelationship could be mapped as the relation of concepts and high-level concepts to enrich theextent of concepts.3. Analysis on mergence and refining on ontologiesWith the access to the related partial ontologies, the description logics is used to build therelated semantic models. The reasoning system and word comparisons WordNet are importedto merge two or more partial ontologies on concept level to generate the global ontology.Then according to the custom-made flow from services administrators, the formalizedmedical flow is described in SWRL language. The flow is translated into the action concepts,and embedded into the ontology net.Finally, according to the method in this paper, a ontology construction system generatingfield ontology based on the relational database is designed and implemented. Then use theprototype system to generate the ontology and merge the ontologies from several resourcemedical databases. With the conclusion, it is illustrated that the system could generate theontologies and merge them into medical field ontology from legacy databases efficiently.