I have had a long involvement in temporal data management. The major results from this were the development, with my then PhD student, Dr Nikos Lorentzos, , of the first use of a temporal interval data type to support temporal data, . We subsequently generalised this to a generalised interval, . This work was later influential with the ISO Standards Committee on Database Languages in considering potential temporal extensions as noted in the new book by Date, Darwen and Lorentzos, “Temporal Data and the Relational Model”.
I continued with this work and explored the use of nested relations with my research students Yung Jang and Georgia Garani. With Yung Jang, we addressed issues relating to managing temporal data with a nested data model,  and subsequently with an object model, . I continued this work with Georgia Garani and we published the first generalised nested join algorithm, . A fuller discussion is in her recently examined PhD thesis .
Arising from a comparison of two potential temporal extensions, ATSQL and IXSQL, my research student, James Green, and I perceived a need to evaluate which extension was the more “usable” and why. Basing our work on the HCI literature on the subject, a series of experiments is being conducted at the moment which involves subjects in studying a temporal extension and then attempting a series of tests involving writing queries in the relevant SQL extension.
During the design of the tests it became clear that results could be influenced by whether subjects did a “pencil and paper” test or were tested in an online simulation. A generalised package was developed to build simulations of SQL extensions.
1. Extending Relational Algebra to Manipulate Temporal Data (with N A Lorentzos)
Information Systems Vol 13 (3), 1988 pp 289-296.
2. An Extension of the Relational Model to Support Generic Intervals (with N A Lorentzos)
Proc of the International Conference on Extending Database Technology, Venice, Italy, 1988 in Advances in Database Technology ‑ EDBT'88, Ed. J.W.Schmidt, S.Ceri & M.Missikoff, Springer‑Verlag, 528‑542 (1988).
3. A Formal Extension of the Relational Model for the Representation and Manipulation of Generic Intervals
N A Lorentzos, PhD thesis, London University, 1988.
4. A Heterogeneous Temporal Nested Relational Data Model (with Y P Jang*)
7th IEEE-CS International Symposium on Computer and Information Sciences,
5. Evolutions of Object States in Temporal Object-Oriented Databases (with Y P Jang*)
of 22nd ACM Computer Science Conference CSC'94,
(with G Garani*)
Information Systems Vol 25, 4 (2000) pp 287-307.
7. ProSQL - A Prototyping Tool for SQL Language Extensions (with J Green*)
Proceedings of 20th British National Conference on Databases
8. A Temporal Database Model using Nested Relations
The School taught possibly the first Masters degree course in computing when it began the M.Sc in Numerical Automation in 1957. At the same date the Computing Laboratory was re-designated the Department of Numerical Automation and was amongst the world’s first formally established academic computing departments in contrast to “computing laboratories”.
I have created a (very large) Powerpoint version of the history.
I was Programme Chair of the BCS@50 conference marking the BCS 50th Anniversary, held from July 12-14th 2007, celebrating the British contribution to computing. My talk is available here.
Another strand of research is being undertaken by my
research student, Bruce
Ready Reckoners (with B O
IEEE Annals of the History of Computing Vol 27, No 4 (2005), pp64-80.
(with B O B
Under revision for IEEE Annals of the History of Computing.
3. HEC 1200 Computer - In preparation.
For submission to IEEE Annals of the History of Computing.
Software re-use has been a major c
1. Trends in Java code changes: the key to identification of refactorings? (with Y Hassoun*, S Counsell)
Proceedings of 2nd ACM International Conference on the Principles
and Practice of Programming in Java,
Open Implementation and Java’s Dynamic Proxies (with Y Hassoun*, S
Counsell, K Mannock and
Proceedings of 2nd ACM
International Conference on the Principles and Practice of Programming in Java,
3. Reusability and Open Implementations in Distributed Environments (with Y Hassoun*, S Counsell)
Software - Practice and Experience, Wiley, Vol 35, No 1, pp75-99.
4. A Dynamic Runtime Coupling Metric for Meta-Level Architecture (with Y Hassoun*, S Counsell)
5. Dynamic Coupling Metric: Proof of Concept (with Y Hassoun*, S Counsell)
IEE Proceedings Vol 152, No 6 (2005) pp273-279.
Processing operations involving interval data types gave rise to considerable computational loads in executing the additional operators needed. At the time the School had a Meiko Computing Surface with 32 transputers. With my colleague Nigel Martin and research assistant Xiaoding Zhao we designed a parallel implementation of an SQL database engine. This was successfully implemented and demonstrated proof of concept. The architecture employed mapped the execution tree derived from the SQL query on to the transputers. Each node in the tree could be allocated to a number of transputers reflecting the expected volume of tuples to be processed and the complexity of the instruction. The results then “flowed” on to one of the transputers allocated to the next node in the tree.
engine was, to the best of the implementors’ knowledge, the first working
parallel database engine in the
1. PPS ‑ a Parallel Partition Sort Algorithm for Multiprocessor Database Systems (with X.Zhao*, N Martin)
Proceedings of 11th International Conference on Database and Expert System Applications (DEXA).
IEEE Computer Society, 2000, pp 635‑644.
2. DBJ - A Dynamic Balancing Hash Join Algorithm in Multiprocessor Database Systems (with X Zhao*, N Martin)
Information Systems, Vol 18 (2), pp 89-100, April 1994.
3. Dynamic Distribution Estimation for Hash Partitioning in Multiprocessor Database Systems (with X Zhao*, N Martin)
Proceeding of High Performance Computing Conference
94, National Supercomputing Research
4. An Improved Dynamic Balancing Hash Join Algorithm (with X Zhao*, N Martin)
My involvement with intelligent
data analysis started with the SCAMS (Structure Characterisation and Automation
for Mass Spectrometry) project which was a collaboration between our
Similar techniques were also applied to medical and bioinformatics data and reported in [4, 5, 6]. The work reported in  led on to further work on protein structures although most of the work was related to the integration of data from multiple sources into a re-implemented CATH database, see ref .
1. Knowledge-Based Data Generation (with H Dettmar*, X Liu et al)
Knowledge‑Based Systems, Vol 11, pp167‑177, 1998.
2. Advanced Data Pre‑processing in SCAMS (with X Liu, M Phalp*)
European Congress on Intelligent Techniques and Soft Computing,
3. The Acquisition and Application of Meta-Knowledge in the SCAMS System (with Dettmar H J* et al)
Proceedings of the 7th Annual Conference
4. Molecular simulations on proteins: A software protocol for analysis and comparison of simulation data in relation to experimental crystallographic data (with M Phalp*, M Knaggs and J Goodfellow)
Technical Report, Computer Science Department,
5. Domain Knowledge in Intelligent Data Analysis (with X. Liu)
Proc of the 16th IFIP World Congress, August 2000,
6. Soft Computing for Intelligent Data Analysis (with X Liu et al)
Proc. of the 18th International Conference of the North American Fuzzy Information Processing Society, pp 527‑531,
7. PFDB: A Generic Protein Family Database integrating the CATH Domain Structure Database with Sequence Based Protein Family Resources (with A Shepherd*, P Kellam, N J Martin, C Orengo)
Bioinformatics Vol 18, No 12 (2002), pp1666-1672.