Title: The Ordered Relational Data Model and its Ramifications

Abstract:
The ordered relational model is an extension of the relational model which incorporates partial orderings into data domains. This seminar will demonstrate that the ordered relational model is suitable for modelling many advanced database applications. For example, temporal relations can be modelled as a special case of linearly ordered relations over schemas consisting of the time, time-variant, and time-invariant attributes.

First, a query language, Ordered SQL (OSQL), which is an extension of SQL for the ordered relational model, will be presented. By several application examples the seminar will show that OSQL combines the capabilities of standard SQL with the power of user-defined semantic orderings. The syntax of OSQL is a minimal extension of SQL and thus it should be easy for current SQL users to adapt to OSQL. Although it is a minimal extension, OSQL allows the users to formulate a wide range of queries, such as incomplete or temporal, which are either very awkward or impossible to formulate in standard SQL.

Second, this seminar will discuss how to employ a set of chase rules for Lexicographically Ordered Functional Dependencies (LOFDs) to tackle the implication problem for LOFDs. An LOFD in the ordered relational model captures the semantics of a monotonicity property between two sets of data. The set of chase rules is then justified to be practically feasible.

Finally, this seminar will explain the idea of Temporal Functional Dependencies (TFDs) in order to express the semantics of the temporally ordered data. A TFD can be expressed in term of a set of LOFDs, in which every element has a sequence of time attributes on its left-hand side and a single time-variant attribute on its right-hand side. Finally, a simple, sound and complete set of inference rules for TFDs will be exhibited, indicating that TFDs are a useful semantic constraint for temporal databases.

Biography:
Dr Wilfred S. H. Ng obtained his B.Sc.(Hon) degree and Postgraduate Certificate of Education from the University of Hong Kong. He then received his M.Sc.(Distinction) and Ph.D. degrees from the University of London (UCL) in the U.K. His research interests are Databases and Information Systems, which include semantic data modelling, temporal databases, data warehousing and web data mining. Currently, Dr Ng is a lecturer in the Department of Computing of the Hong Kong Polytechnic University. He is also the principal investigator of the research project "Temporal Data Mining From Ordered Web Log Data", jointly conducted by his department and the Department of Computer Science at UCL.