Information Systems Engineering

Simple Solutions to Complex Problem Solving

Information Systems Engineering is a key area for the future. Over the next 12 years we will see as much change as we have seen over the last 12 years. From the computer technology point of view we have domain changes enabled largely by the new technical knowledge that we have, for example the growth in local communications has been greatly enhanced through the provision of SMS messages. Domain demands are leading to increasing demands of software engineers to provide fast development and robust systems. Here is an index of the terms used here.

The sophistication of data management is also leading to increasing efforts to attempt to formally codify the knowledge that we have about the world. Information can be defined as knowledge derived from study, experience, or instruction; increasingly we are able to use machines for remembering the information, all we need to do is interact with the machines in order to extract the data. Search engines are a very basic model. Much effort therefore will be expended in the field of knowledge engineering and management.

Systems are a product of these endeavours, a result of the combination, setting up and establishment intellectual and physical labour. The maintenance and access of these systems will become an important issue. Growth in data systems is high, as this trend continues the needs to develop highly sophisticated algorithms to extract the right data is going to become more and more important.

Systems are having an affect on the space dimension. From a geographical point of view, grid computing is growing and wireless engineering has now begun to roll out. Networked Virtual Environments (NVEs) are being developed.

Some old problems persist: optimising memory requirements, how do we search for information, and how will the net, or the world wide web, develop? Security and encryption continue to grow.

1. Domain Scope and Domain Changes

The domain of our systems development work is changing fast and has a radical effect on the work itself. Mathematics and simple games were early applications, now the applications are many.

Concept Map

2. Systems Development Process

Key factors in the development of systems are the method of abstraction, to what extent we can capitalise on repetition and reuse, how we conceptualise our problems, and how we integrate recursion and reflexivity into our models. Key to all modelling decisions is the order in which knowledge is to be compiled, and the various relevance levels of the data collected. Objects to be modelled need to be assigned identifiers, usually unique, to provide indexing and translation capabilities.

It still remains the case that system development is a tribal activity, with development methodologies created around organisations with specific individuals like the artisans of old. Logica and KPMG may develop system engineering principles, but unification across companies is undone by the cult of personality that is sustained by massive technology shifts. There is also a big research area to emerge on the influence of emotions on human reasoning. Future systems, may focus on even more automation, with the classification of non-system tasks being exception handling. Of course this is renaming old for new, intelligence is the ability to intepret and process exceptions, so nothing has changed there.

Architecture is a word used to describe the structures that are built with computer programming code. Systems Architecture has a number of spin-off terms: "Model Driven Architecture" (MDA) promoted by the OMG Object Management Group. "Security Architecture" to describe the way that the aspect of security is handled. "Information Architecture" describes the architecture of information, especially relevant for the WWW, invloving the principles of library science and architecture.

Java continues to go from strength to strength, although dotnet, or .net offers a strong challenge. Software design remains a stumbling block for many companies due to the twin complications of business complexity and technology complexity. Increasingly software libraries to enhance code reuse will become the norm, and the task will be the sewing together of the stored functions. Finally, the issue of risk remains with us to end, a consequence of grappling with new technology and trying to egg it on quicker than it wants to grow. System development methodologies are also evolving with the Eric S. Raymond's Bazaar model describing the Open Source initiatives (1997 "Cathedral and the Bazaar"), looking at code writing and designing from a non-linear point of view, as in Martin Fowler's "Refactoring".

A critical area of system analysis relates to the method of dividing up a system into components and interfaces. The algorithm required to do this needs to look at optimisation of information flow, information modelling, and relevance modelling.

3. Modelling Systems

Modelling of systems can be done with SDL, or ebXML, or a whole host of other standards including UML Diagrams. What is interesting is to determine whether all the systems of modelling are equivalent, and if they are equivalent, like SDL and UML 2.0 for example, how do we assess the relevance of the what each method is doing. In order to evaluate this we need to model the actual cost of making changes after the event.

Relevance is important for efficiency on the "human" side too - for example, the current UML specifications, maybe because they want to be all things to all men, are far too complex for normal specification for normal people on normal projects.

Related also is the lack of provision for business development within UML, so its current status is too complicated, and does not match the requirements.

There is an attempt at the moment to set a standard an to own it. This is clearly as ridiculous as trying to own a part of mathematics.

Difficulties of operating in the public www environment include revenue generation and supply, viruses, spam (including referrer spam).

4. Managing Systems - Maintenance and Growth

Much of the information development work these days consists of revising legacy systems, data formats, and interfaces and transformations. Processes which support this work are required together with valid models which support the organisational requirements of ( Simple, Inexpensive, Relevant, Zero-risk ) SIRZ systems. On the cost issue, a careful balance needs to be found on expenditure, where the return on investment matches the requirements. In particular, attention needs to be paid to the development of systems which are inexpensive to amend and convert. This requires a careful analysis of what data formats and data interfaces are inexpensive to maintain, both in terms of license fees and the ease of which the data can be manipulated.

5. Developments

Functional Approach to Data Management
ebXML
erlang

References:

Gardenfors, Peter (2002), "The Geometry of Thought", MIT Press
Fowler, Martin (1999), "Refactoring", Addison-Wesley