A Relational Model of Data for Large Shared Data Banks

[abstract] "Future users of large data banks must be protected fromhaving to know how the data is organized in the machine (theinternal representation). A prompting service which suppliessuch information is not a satisfactory solution. Activities of usersat terminals and most application programs should remainunaffected when the internal representation of data is changedand even when some aspects of the external representationare changed. Changes in data representation will often beneeded as a result of changes in query, update, and reporttraffic and natural growth in the types of stored information."Existing noninferential, formatted data systems provide userswith tree-structured files or slightly more general networkmodels of the data. In Section 1, inadequacies of these modelsare discussed. A model based on n-ary relations, a normalform for data base relations, and the concept of a universaldata sublanguage are introduced. In Section 2, certain operations on relations (other than logical inference) are discussedand applied to the problems of redundancy and consistencyin the user's model."