the "Data Sciences Group"

The "Data Sciences Group" has been at the forefront of research design, data analysis and data analytics since the late 1970's.

We are a privately owned company incorporated in Australia as Data Sciences Pty Ltd (ACN 005 602 595) in 1979, and have traded profitably and pleasurably ever since, primarily in Sydney and in Melbourne, under the same ownership.

Our directors are John (physicist, statistician, programmer, systems designer) & Tricia (economist, market researcher) Aitchison, and you may contact either on 0427-791495 or via email

As to what we do, and what products and services we provide, there are links on this page to our subsidiary websites for more specific information.

We have diverse interests : broadly speaking they lie in the areas of pattern recognition and automation. We like to understand things more deeply, make things work better, make tools that make the "working better" faster and cheaper and more human scale.

We like ..

• Patterns. Patterns as a way of making sense of the world (the world as it presents itself in the shape of a "problem" or an "opportunity") . Patterns of behaviour, patterns in data and processes, patterns in images and text.. recognizing patterns : the fundamental "stuff", the leap in understanding, the "ah ah... that is an XYZ" or " that looks like that or maybe that".. .
  
  Finding the patterns, describing them, categorizing and classifying new data into them - the fundamental way of human thinking, the stuff of taxonomy and clustering and discriminating and classifying. And patterns are ubiquitous, everywhere .. in the regularity of business cycles, in accounting and cash flows, in purchasing behaviour, in images, in market research and survey data ..
• Breaking new ground excites us ... using creative and analytic processes in fields where they have not been applied before. To be reductionist about this, it is a matter of mapping the new and unknown field to the tried and true research and analysis techniques.
• Writing .. websites, and reports, and position papers. Writing in English and in program code, literate programming, accessible and pattern driven object oriented programming. Writing to clarify, to expound, to propose next steps..
• Democratization. Making processes and 'advanced analysis' techniques accessible to even small organizations and businesses .. even to private individuals for use in 'everyday life'.

Big things? .. perhaps there are and perhaps there should be no "big things", no "ambitious projects" : just an accretion of little things. Baby steps, "small fish are sweet".. easy to get your head around. Incremental learning and applied case based reasoning.

Some things are clearly "too big".. suppose you asked us to look at image retrieval by similarity, or modelling consumer choice for as yet un-designed products, or forecasting currency movements, or predicting the success of movies, or optimal budget allocations, or patterns in the stock market, or global warming or .. these are all interesting problem areas, but intrinsically too ambitious as stated.

There is work that is do-able in each of these areas, but first let's more clearly define our terms and objectives and define the scope of each stage. Baby steps.

We use languages and write computer languages, we build experimental designs specific to a problem, we automate market research design and data analysis, we use advanced statistical analysis to sieve through masses of data to build models, we analyze and build text, we design ontologies, we research, we write, we clarify, we abstract and synthesize ..

In short, we look at lots of real world problems - client initiated, or in areas where we speculate there may be a market opportunity - and apply existing statistical and programming and communication tools, or design and build new tools, new systems.

Thank you for your interest.