Time is Just Another Attribute - Or At Least, Just Another Dimension

Writers on temporal databases frequently claim that most databases hold only a snapshot of the data to be stored in that database and that commonly-used database design tools such as the relational model are unsatisfactory for the development of database systems in which attribute values change over time. Such writers urge that valid time — the interval over which a data item is valid — be treated differently from time which is an ordinary attribute of a relation. This paper argues that, when considering possible extensions to the relational model for such purposes, a distinction must be made between the data model used to represent temporal data and algebras used to specify processing on that model. It is argued that both valid time and user-defined time (as defined in TSQL2 and by many writers in this area) can and should be modelled as simple attributes of a relation within the unextended relational model. This is contrasted with the need for extra facilities in relational query languages for processing of temporal and other interval data. The extra querying facilities identified in these case studies are those for manipulating overlapping intervals and support for recursive querying of version trees. TSQL2 is criticised in terms of its limitations, complexity and reliance on a specific temporal data model.

[1]  James Clifford,et al.  A Temporal Relational Algebra as a Basis for Temporal Relational Completeness , 1990 .

[2]  Christina Davies,et al.  How helpful is current theory for the representation of time in real databases? A case study and discussion , 1995, ECIS.

[3]  llsoo Ahn,et al.  Temporal Databases , 1986, Computer.

[4]  Umeshwar Dayal,et al.  A Uniform Model for Temporal and Versioned Object-oriented Databases , 1993, Temporal Databases.

[5]  Derek Thompson,et al.  Fundamentals of spatial information systems , 1992, A.P.I.C. series.

[6]  Shamkant B. Navathe,et al.  Temporal Extensions to the Relational Model and SQL , 1993, Temporal Databases.

[7]  Jesse M. Draper,et al.  Guide on Data Models in the Selection and Use of Database Management Systems | NIST , 1984 .

[8]  E. F. Codd,et al.  The Relational Model for Database Management, Version 2 , 1990 .

[9]  Richard R. Muntz,et al.  Stream Processing: Temporal Query Processing and Optimization , 1993, Temporal Databases.

[10]  E. F. Codd,et al.  A relational model of data for large shared data banks , 1970, CACM.

[11]  Klaus R. Dittrich,et al.  Version Support for Engineering Database Systems , 1988, IEEE Trans. Software Eng..

[12]  I. Shearer.  Retrospective Update: Data As It Was Believed To Be (Short Note) , 1992, Comput. J..

[13]  Richard T. Snodgrass,et al.  Evaluation of relational algebras incorporating the time dimension in databases , 1991, CSUR.

[14]  Abdullah Uz Tansel.  Non First Normal Form Temporal Relational Model. , 1988 .

[15]  Thomas C. Waugh,et al.  The GEO VIEW design A relational data base approach to geographical data handling , 1987, Int. J. Geogr. Inf. Sci..

[16]  Michael Stonebraker,et al.  Inclusion of new types in relational data base systems , 1986, 1986 IEEE Second International Conference on Data Engineering.

[17]  Ramez Elmasri,et al.  Fundamentals of Database Systems , 1989 .

[18]  Kemal Koymen,et al.  SQL*: a recursive SQL , 1993, Inf. Syst..

[19]  Richard T. Snodgrass.  A TSQL2 tutorial , 1994, SGMD.

[20]  Constantine D. Spyropoulos,et al.  A temporal framework for managing retroactive and delayed updates: an application to the payroll information system of the Greek public sector , 1993 .

[21]  C. J. Date.  An Introduction to Database Systems , 1975 .

[22]  Arie Segev,et al.  A consensus glossary of temporal database concepts , 1994, SIGMOD 1994.