A global query processing system handles a large number of databases scattered in a large network. These databases may be independently designed even if they store related information. Ordinal multidatabase approach may not be effective in open environment because user may want to get information from many sites whose database schemas are not known. We propose an approach based on an incomplete database concept in order to retrieve necessary information by simultaneously querying various databases without knowing the exact schema of each database. The incomplete database concept for global query processing is different but related with incomplete database with null values. This paper discusses semantics of these incomplete databases.
展开▼
