The Research And Implementation Of Multiple View Consistency Maintenance Technology In The Warehouse Integration Environment

Data warehouse technology is a kind of data integration technology that tries to reserve thelegacy system while maintaining the autonomy and flexibility of information systems inorganizations and makes the new and old information systems work coordinately. Datawarehouse integration is a dynamic integration. In order to reflect the changed data from datasource to warehouse regularly or timely and ensure the data consistency in warehouse,materialized view online maintain technology should be achieved. The data consistency problemincludes the view consistency between a view in warehouse and its base data at the sources, andthe mutual consistency among the multiple views in warehouse which is called multiple viewconsistency.In this paper, a new multiple view consistency maintenance algorithm is proposed on thebasis of the large-scale data integration system which is developed by our project team. Thisalgorithm takes the advantage of a multiple view consistency maintenance algorithm which isproposed by Yue Zhuge from Stanford University and grarantee the data warehouse stateschanges consistently. The algorithm not only enables the flexible combination of the integratedview maintenance mode and incremental submitting mode, but also solves the dependencyproblem of machine time which is attached to the incremental update files. The large-scale dataintegration system after implementing of this algorithm is not necessary to concern the platformenvironment, and is more suitable for the cross-organizational and cross-region informationintegration application to realize self-maintenance.