Thursday, July 26, 2007

Thesis Project - Update

This week, I travelled to the University of New South Wales to review their offline material. Located 3 relevant sources, which will be listed later. Furthermore, I have read the newest paper by Parham (2001), in order to execute the data collection part of this treatise.

Furthermore, I am currently implementing regression in "R" to get an idea of how to perform switching regression in the modelling part of this treatise.

This week, I will be attempting data collection via ABS, as well as coming up with an approach on how to tackle this problem. An idea I have right now would be to concentrate on Bezdek's paper on Switching Regression and Fuzzy Clustering dated 1993, which would allow me to develop an understanding of switching regression in relation to Fuzzy C-Regression. An alternative, which may be easier to perform (but which is considered an alternative, as Fuzzy C-Regression may yield more detailed results), would be that of utlilising Fuzzy C-Means in relation to Switching Regression, as Dr Poon has performed this before, and thus I would be able to draw on his experience if I were to implement fuzzy clustering using the Fuzzy C-Means method.

As an aside, I have created the project plan, which comprises of a table listing the times and requirements, as well as a list of sources (which is not complete, as I may locate sources later on which I have not been able to locate thus far).

Wednesday, July 18, 2007

Thesis Update

This week I spent more time on researching the impact of IT upon industry, as viewed from the production function f(x) = y(ICT, non-ICT), Note: f(x) is a productivity function which is determined by y, with variables both ICT and non-ICT related.

Have read most papers by Bryjolfsson, and have researched into his sources to gain a deeper understanding of the subject area, by viewing his bibliography and extending upon my research from that point. I am now also making arrangements to visit UNSW's library. I have scheduled my trip to Maquarie University for today, and will be looking at the following books:

1. Beyond the IT productivity paradox / edited by Leslie P. Willcocks and Stephanie Lester

2.Information technology and the productivity paradox : assessing the value of the investment in IT / Henry C. Lucas, Jr

3. The trouble with computers : usefulness, usability, and productivity / Thomas K. Landauer

Furthermore, have researched into clustering, in particular, fuzzy clustering, as well as started doing tutorials upon R, available from, to gain knowledge of R in order to gain the required skills to perform both fuzzy clustering and switching regression using R upon industry data sets.

In my research with ABS, I have also located Australian National Accounts: Information and Communication Technology Satellite Account (from, which includes reports on investment in ICT products, and investment in hardware and software, sorted by industry.

Tuesday, July 10, 2007


In the past two weeks, I have met with Dr Simon Poon twice to discuss the progress of my research into the topic area. We have set a schedule for the work, which is as follows:

By August 1st, w should have a clear idea of what we're doing, as well as be up to date on relevant literature

By September 1st, we should have finished all modelling (which in my case involves the application of fuzzy clustering, as well as switching regression, to a data set in order to generate results, and infer findings from these results).

By October 1st, we should have finished analysis of the modelling results.

We would also be using October to finish writing up the final copy of the thesis.

Additionally, I have conducted research via the library's online databases, based on preliminary articles provided by Dr Poon. Furthermore, I have gone to the Fisher, and Badham libraries, to conduct research into this area. I am planning on going to more libraries to conduct further offline research into the areas of clustering, and regression, in the coming week.