Research On Knowledge-Engineering-based Domain-Specific Modeling Method For Emergency Management

Recently,emergency management has aroused high attention along with increasing various emergencies.Luckily,computational experiment based on artificial society gradually becomes the efficient way to study this kind of society problems.Modeling and simulation(M&S),which rely on existed domain knowledge and generates new knowledge by computational execution,is the activity of knowledge generation.Simulation resource is the basis of domain simulation,and models,viewed as the cores of simulation resource,are the integration of domain knowledge.In this paper,the presentation and transition of domain knowledge about artificial society are studied from the perspective of knowledge engineering,and the transforming process from domain knowledge to simulation models are especially discussed.Combined with ontology technology and meta-modeling technology,this paper tries to construct the domain-specific modeling environment to support the model development and maintaining.The main works and light spots are summarized as follows:This paper makes a thorough research of the presentation,transformation and innovation of domain knowledge in M&S,and brings up the evolution process of domain knowledge in the process from the perspective of knowledge engineering.Firstly,the relation between domain knowledge and simulation resource is analyzed,along with the problems of constructing the simulation resource architecture.Secondly,the modeling process of domain knowledge is discussed,and the features of domain knowledge are also discussed,such as knowledge storage,knowledge flow,knowledge fluctuation and knowledge granularity.Combined with emergency-management-oriented computational experiment,the integration and involution mechanism of domain knowledge are discussed,and the innovation of domain knowledge is brought up,which contains the process of domain knowledge,simulation resource,simulation application and new knowledge.Combined with the computational experiments,domain knowledge about emergency management is analyzed,based on which comceptual models are built.Firstly,the knowledge architecture of emergency management is analyzed,and the methods of knowledge modeling are summarized.Secondly,the theory and technology about ontology modeling are especially discussed,and the ontology library is built to depict the corresponding objects along with the relationships among them.Then the description of scenario based on ontology is studied,and the description framework of scenario is proposed which combines ontology,metadata and scenario together.Both simulation case about prediction and controlling infectious diseases and the reusing mechanism of simulation cases are discussed.Meta-models are distilled and abstracted from domain models,which are the integration of domain knowledge and are easy to reuse in engineering.The corresponding theory and technology of meta-modeling are summarized in this paper.Combined with the emergency management in public health,the population meta-models and environment meta-models of artificial society are designed along with emergency meta-models and corresponding intervention meta-models.Finally,emergency-management-oriented domain-specific modeling environment is also constructed.Model transformation and code generation are studied,and the realization methods of simulation models are given based on these technologies.In a sense,the process of model realizing is also the process of model transformation,which contains model-to-model transformation and model-to-texture transformation.In this paper,the building and realizing process of domain models about emergency management are studied,and the code generation framework is also designed,which is able to transform PIM to PSM and support simulation execution.Additionally,the validation,integration and sharing of models are also discussed in this paper.Simulation case about prediction and controlling infectious diseases is also studied,and the reusing mechanism of simulation cases is also discussed.Simulation case is the summarization of simulation process,corresponding simulation resources and the experience we gain in the process.It is significant to the knowledge integration,reusing and innovation.Combined with the prediction and intervention of Ebola emergency,the description and management of simulation case are studied,and CBR-based reusing mechanism of simulation cases along with the corresponding similarity matching algorithm are also discussed in this paper.In a word,this paper provides a new perspective to recognize the M&S process,and teases the knowledge presentation,transformation and innovation from the perspective of knowledge engineering.Combined with emergency-management-oriented computational experiment in public health,the transformation process from knowledge to models are especially discussed,which contains the transformation from domain knowledge to conceptual models and the transformation from conceptual models to simulation models.In the paper,a new method is proposed to construct the domain-specific modeling environment combined with technologies about ontology modeling,meta-modeling and model transformation.Moreover,the reusing of domain knowledge in the form of simulation cases is also studied by introducing CBR-based technology.In short,this paper discusses the process of M&S,which is not only significant to recognize the essence of simulation but also to reconstruct the modeling and simulation environment.