Scenario-based Data Interoperability In Industrial Interconnection Environment

In the whole-link industrial interconnection,data is generated by multiple subjects and used by different organizations in different parts of the data value chain.The data interoperability problem caused by the heterogeneity of multi-source data,as well as the heterogeneity of dynamic user demand is the key to hinder data application.According to the service-oriented principle of "paying equal attention to data resource integration and usage",data sharing needs to ensure the findability,accessibility,interoperability and reusability of data,as well as to realize the maximum integration,high connection and matching of data,methods and user requirements.Intend to realize above purposes,on the one hand,data sharing needs to manage,map and integrate multi-source heterogeneous data,so as to provide stakeholders with a visual panoramic map of data resources.On the other hand,data sharing needs to be service-centered and reduce the interoperability barriers between users and complex data sources.By comparing typical data sharing cases at home and abroad,this study found that most data sharing address data heterogeneity from the technical level,ignoring the complex business logic and business semantic heterogeneity.Specifically,data resources and information system that generate them are strongly coupled,and business knowledge behind the data has already reached a potential tacit understanding between the user and the supplier in a small scale.However,such knowledge is not well migrated when data resources disconnected from information systems,shared and applied on a larger scale.As a result,data sources lack sufficient semantic information in the process of identifying reasonable semantic inconsistency between them.When it comes to data usage,in order to reduce the interoperability barriers between users and data,most of the existing solutions are to define a collection of domain knowledge and present them to users in a multi-level catalog for querying data resources.The directory-based data resource query approach emphasizes on management rather than service,and favors resource-and space-based rather than user-centered operation mode.It is found that with the expansion of user groups in industrial interconnection and the difference of users’ respective business domains,the heterogeneity of expertise(or taxonomy)among users and the dynamic changeability of user scenarios are increasingly prominent.The seemingly similar expertise in different user scenarios essentially has significant business semantic differences.This brings new challenges to the study of interoperability between user requirements and data resources.The heterogeneity of expertise among users and the dynamic nature of user scenarios make it difficult for the existing directory-based data query methods to provide proactive data services to users.In summary,data sharing faces the problem of identifying data heterogeneity between multi-source data caused by business complexity and uncertainty,as well as the problem of interoperability between multi-source data and scenario-based user requirements.To this end,the study proposes to build a set of scenario-based data interoperability methods to address the data complexity of both the supply side and the demand side,as well as to explicit the complexity of the business,organization,and system that behind data.That establish an operable sustainable data sharing service method based on knowledge fusion,i.e.the identification of multi-sense terms by semantic context fusion,concept-extended data model alignment,and an implementation method under collaborative governance.Through research on the identification and dissolution of semantic heterogeneity of complex multi-source data and scenario-based semantic identification of dynamic user needs,our research realize the closed-loop design of interoperable services from the data supply side to the data usage side,so as to promote the discovery of data value-added opportunities in industrial interconnection.In order to address he heterogeneity between scenario-based user requirements and complex multi-source data.This paper proposes to identify and reorganize the differentiated expertise embodied by users in data usage as an intermediate knowledge base connecting users and data resources to support further research on scenario-based data interoperability.To this end,this research proposes a set of polysemy recognition and modeling methods based on semantic context fusion,that builds a terminology base with the help of static concept relationships and dynamic usage contexts,thus establishing semantic association paths between polysemy and their applicable contexts.The constructed taxonomy can be used to solve the problem of implicit knowledge reasoning in intelligent response to user needs,so as to achieve the purpose of user-centered and reducing interoperability barriers between users and data.In short,the study proposes to build a multisense term recognition method based on semantic context fusion to reorganize the differentiated expertise of users in data usage to solve the difficulty of providing proactive data services to users by the existing directory-based data resource query method.In response to the heterogeneous problems caused by business complexity and uncertainty that cannot be identified and solved at the technical level,this study proposes to systematically sort out the problems faced by data resources detached from information systems from the level of conceptual data models.Since the context of the study is crosssystem,cross-professional and cross-organizational data interoperability,it is difficult to reach consensus on a certain model among multiple organizations due to the requirement of autonomy,therefore,the study mainly adopts meta model-based protocol technology,which researches on how should the conceptual data model meta model elements be defined to accommodate the heterogeneity of multiple source data as much as possible.In addition,in order to assist data experts in identifying heterogeneity problems,the study proposes a semi-automatic approach to achieve conceptual data model alignment,that conducts a similarity comparison study of each instance of the conceptual data model based on a combination of structural similarity,semantic similarity,and name similarity.Conflict resolution are systematically given for different heterogeneous problems.In short,the study proposes to construct a conceptually extended data model meta model and use a semi-automatic collaborative approach to identify and disambiguate data heterogeneity problems caused by business complexity and uncertainty between conceptual data models from different sources.Existing studies mostly adopt the technical perspective to evaluate data interoperability methods.In the industrial interconnection environment,the characteristics of data such as multi-subject sources,multi-business links,multiorganizational sharing,and complex data value chain make it more necessary to evaluate data interoperability in terms of the operability of organizational collaborative governance and shared services.Therefore,this study proposes scenario-based data interoperability implementation and application methods infected by collaborative governance idea in the context of complex industries,complex organizations,complex businesses,and complex systems.In short,most of the existing studies evaluate data interoperability approaches from a technical perspective,without verifying whether the proposed approaches can truly meet the data usage needs of users in specific scenarios.To this end,the study proposes a scenario-based data interoperability implementation and application method under collaborative governance.