Tuesday, July 25, 2006

Wiki-ish System For Structured Debate on a Topic - spacedebate.org

I'm very interested in the issue of how to effectively represent evidence and arguments, and this looks promising: spacedebate.org "is an effort to expand the debate on the weaponization of space through a collaborative wiki-like tool for structured debate on a topic".

The site has a quick tour which explains it well.

"Each side of the debate is broken up into positions which are composed of arguments and their supporting arguments, in a structured argument tree".

There's RSS feeds for changes to arguments, and javascript code to include live arguments (and their related discussions, subarguments, and resources) on your own blog or site.

"The project is part of a larger effort, nicknamed the "Open Debate Engine" [the site is here, but it hasn't been released yet], which hopes to create a hybrid wiki platform that will be conducive to handling larger policy debate topics."

There's some rationale for having the structured format and what's wrong with the existing methods for online debate, here.

I think that improving an argument is often quite subtle, and can't be done without major revision of its structure, rather than by making changes within -- or by adding to -- its existing structure... so I think this is an issue that any site like this will have to deal with.

Some more info, from the about page:

Spacedebate.org is an effort to expand the debate on the weaponization of space through a collaborative wiki-like tool for structured debate on a topic. The project is modeled after Wikipedia, but instead of focusing on developing an encyclopedia it invites users to help edit and expand an 'argument tree' that reflects the various positions in the debate over U.S. military space policy. Users can browse the argument tree or an extensive database of resources, including links to relevant news articles, authoritative quotes, and a comprehensive bibliography of sources.

Monday, July 24, 2006

'A Porch and Flowering Meadow, 6 Floors Up'

I reckon this is cool...it's a kinda best-of-both-worlds thing.

Eileen Stukane, David Puchkoff and their daughter, Masha, created a porch and a miniature meadow on top of their West Village [New York City] apartment building.

The 'Super Mario Bros 2' We Got was a Substituion for the Real Sequel

Strange to not have known till now, but it seems the Super Mario Bros. 2 released in most places outside of Japan was actually a slightly reworked version of a game called Doki Doki Panic. The real sequel was deemed to difficult; though a modified version of it was later released here as 'The Lost Levels' on the Super Mario All-Stars game on the SNES.

Monday, July 17, 2006

Favourite Recipes: Chicken in Green Masala Sauce

Chicken in Green Masala Sauce

To me, this recipe is fairly different from the sort of curries you find in restaurants and even in cookbooks of Indian home cooking. The sauce has a fairly light and somewhat fresh taste to it, and is a little sweet. It doesn't have a spicy flavour to it like curries tend to.

It’s quite quick to make, and has pretty broad appeal - so it’s good when you’re cooking for others.

It’s from “Best-Ever Curry: Over 150 great curries from India to Asia” (originally published with the title “Curry”) by Mridula Baljekar. pg 47.

Serves 4

  • sauce
    • natural (plain) yoghurt, 150ml (2/3 cup)
    • fromage frais or ricotta cheese, 45ml (3 tbsp)
    • onions, garlic, ginger
      • spring onions (scallions), 1 bunch, chopped
      • garlic, 5ml (1 tsp), crushed
      • fresh ginger, 5ml (1 tsp), grated
    • other flavourings
      • crisp green eating apple, 1, peeled, cored and cubed
      • fresh corriander, 45ml (3 tbsp - but see note below)
      • fresh mint leaves, 30ml (2 tbsp)
      • fresh green chillis, 2, seeded and chopped
    • salt, 5ml (1 tsp)
    • granulated sugar, 5ml (1 tsp)
  • vegetable oil, 15ml (1 tbsp)
  • chicken breasts, 350-400g skinned and cubed
  • garnishes
    • sultanas, 25g
    • fresh corriander, 15ml (1 tbsp)
      (note: there is also some corriander listed in ‘other flavourings’, above, making it a total of 4 tbsps of it required for this recipe)

  • Blend the sauce ingredients
    • Process the following in a food processor for 1 minute: the apple, 3 tbsp of corriander, mint, yogurt, formage frias or ricotta, chillis, spring onions, salt, sugar, garlic and ginger. Scrape around the outside of the bowl and process for a few seconds more.
  • Cooking the sauce and chicken
    • Heat the oil in a wok, karahi or large pan, pour in the yogurt mixture and cook, gently over a low heat for about 2 minutes.
    • Add the chicken pieces and stir well to blend everything together. Cook over a medium-low heat for 15-17 minutes or until the chicken is fully cooked.
  • Add garnishes.
    • Srinkle the sultanas and the remaining coriander over the chicken (do not mix in, but leave as a garnish).

Serve with Nut Pulao, if you like.

I’ve modified the presentation of this recipe a little, and I've also changed a few of its details. I’ve found an increased cooking time necessary, so I’ve changed that from 12-15 minutes to 15-17 minutes. I also increased the amount of chicken from 225g to 350-400g, as I think the original recipe makes far too much sauce for the amount of chicken in it.

Saturday, July 15, 2006

My PhD Scholarship Application

Here's my PhD scholarship application, from early 2004.

Looking at its content, it does feel quite personal. Largely, I think, because it's so focused on selling myself. Not that I mind it having that focus, as that's what the application required, but it feels a bit funny putting it online.

So why have I? In part because this is the kind of thing I would be interested in seeing if someone else did it. Also because I spent all that time writing it, so it also feels a bit funny keeping it tucked away in a drawer.

I often get asked what my PhD research is about, and it always frustrates me that I can't give a very articulate answer - so this may help a bit for that. Even though two years have passed now, it still provides a pretty accurate description of what I'm doing.

Thanks goes out to the following people. To Bob Colomb for his assistance with the application. To my dad for proofreading the first draft of the application. To Karen Henrickson for helping out on short notice and proofreading the second draft. To Trevor Chorvat for insight into the sorts of things the people evaluating the application would be looking for. And to Zoran Milosevic for giving me the flexibility and support at work that helped make it possible for me to get it done.

Academic Merit and Research Potential

I believe my academic performance demonstrates a strong research potential, in terms of both my overall honours grades and my performance in the most significant research component of the degree, the honours dissertation.

My first semester studying honours was affected by illness, and while I successfully applied for special consideration for the impact of this illness on my study, I believe my performance in the remainder of the course provides a more accurate reflection of my academic merit.

In that first semester, I received grades of six, two fours and a failure, and while the special consideration removed the failure from my academic record it left the other marks standing; this resulted in a grade-point average of 5.33. For the remainder of the course (for five eighths of the honours course load), I achieved the highest possible grades, receiving straight sevens across all subjects. For my academic achievement I was awarded a place on the Dean's List. At graduation I was awarded honours 2A.

My research capabilities are clearly demonstrated by my honours dissertation for which I received a 95% and a grade of seven. My dissertation is now used as (the only) example of dissertation work on the FIT/QUT honours coordinator's web-site, at: http://sky.fit.qut.edu.au/~andersam/honours/.

My academic and research achievements led to employment within the IT faculty at QUT, and I believe this employment further reflects and reinforces my capabilities. As a result of my strong performance in an Artificial Intelligence subject, I was offered a research assistant position where I worked on a data mining project that was undertaken for an external client.

I was also offered employment within the faculty to develop all of the content of an online Java bridging course aimed at programmers already familiar with the C language. At the time, the faculty was switching to Java as their primary programming language for teaching, and this course was required for students that were entering degrees after the first year and as a result had not undertaken the introductory Java subjects.

My academic merit and research potential are also demonstrated by my success in gaining employment after graduation as a research scientist. This work is described in further detail, below.

Match with Research Strengths

A strong alignment between the proposed research and the ITEE School's research strengths in data and knowledge engineering and complex and intelligent systems places the proposed research firmly within the province of the faculty's research strengths.

The topic of the proposed research concerns the nature of information. What, ultimately, is information? In brief, what is its fundamental structure and nature, and what material items and processes does it map to in the real world? The intention is to approach this primarily as a fundamental question, rather than as a means to solve an applied problem. It is difficult to think of questions that are more fundamental, particularly to Information Technology, than the question of ‘what is information?’ and yet this question has so far — it seems widely agreed — eluded an adequate answer.

There is a strong match between the proposed research and the ITEE School's research strength in the area of data and knowledge engineering. The School's Data and Knowledge Engineering research group identifies three main research concentrations and strengths; the proposed research primarily matches their Semantic Issues strength, which covers Ontologies and Knowledge Representation.

The question of 'what is information?' can be considered to be largely an ontological one. This provides a clear and definite link between the Ontologies strength and the proposed research. This link is further strengthened by the way ontologies are conceived in the Information Technology domain. In Information Technology, ontologies are primarily considered from the perspective of 'what is the structure of a particular type of information?'. This question inherently requires, and is reliant upon, some conception of what information is.

For these reasons, the group's research knowledge and experience would provide valuable input into the proposed research, and in turn the proposed research could contribute to the group's strengths, such as through its application, in a collaborative context, to applied Ontological research that was being undertaken.
Primary supervision has been arranged with Associate Professor Bob Colomb, whose research focuses primarily within this Semantic Issues component of the Data and Knowledge Engineering research group. Associate Professor Colomb has undertaken research in the area of the proposed research topic and maintains an active interest in it.

Another of the ITEE School's research strengths there is a strong match with is the area of complex and intelligent systems, as embodied by the School's Complex and Intelligent Systems research group. The group's primary aim is to understand the principles that underlie complex systems. In the view of information I hold at this early stage, information arises primarily from the complexities of the perceptual systems of humans (and other information processing agents) — which are complex systems. Thus, the research knowledge and expertise in this group would provide valuable input into the proposed research.

I also believe that the evolutionary perspective, which is taken by many researchers in this group, aligns well with the proposed research; this perspective figures prominently in my intended approach — at this early stage — to understanding what information is and how it can arise. Additionally, I believe that a better understanding of information could provide leverage in understanding the structure and nature of complex systems — as complex systems are often defined and understood through the way they represent and process information.

For these reasons, the group's knowledge and expertise could contribute substantially to the proposed research, which itself could provide leverage for approaching topics within the research group's areas of interest, particularly in analysing and modelling complex and intelligent systems. The match with the Complex and Intelligent Systems research group is realised through the co-supervision which has been arranged with Dr Tom Mandeville. Dr Mandeville is a member of that research group, as well as being a Senior Lecturer in the Faculty of Economics.

In addition to helping realise the link with the Complex and Intelligent Systems group, Dr Mandeville's involvement would also foster inter-faculty ties between the proposed research in Information Technology, and the Economics faculty, where this research topic is also of interest. Furthermore, Dr Mandeville has done considerable work closely related to the topic of the proposed research.

In summary, there are strong and definite matches between the proposed research and the faculty's research strengths in the areas of data and knowledge engineering and complex and intelligent systems, and these are made operational through arranged supervision with staff members who have research expertise in and relating to the proposed research topic.

Demonstrated research performance

Both my accomplishments while studying and my employment record demonstrate a strong research performance. As mentioned above, I received very high marks for my honours dissertation, and its use by the honours coordinator as an illustration of honours work is a strong recognition of its quality.

While studying, I worked as a research assistant, and after graduating I obtained a position as a research scientist in an IT research center — the Distributed Systems Technology Centre (DSTC) — where I am presently working. DSTC partakes in research under the federal government's Cooperative Research Centre program. As a research scientist, I have three and a half years of research experience, including international collaboration and a number of papers published in well respected forums (as detailed below).

I work on a project at DSTC that undertakes electronic contracts research. The project is focusing primarily on the design and development of an electronic contract monitoring (as in, detecting contract violations) and management system, and consists of five researchers. As an illustration of my level of involvement in this research, I have been entrusted with the primary role in researching and designing our Business Contracts Language, which is used to encode the details of a contract in such a way that it can be monitored by our system.

This language is has several novel features, which include novel constructs for detecting patterns of events (required for detecting occurrences that should/should not occur as per the contract) and a unique model for managing the assignment of contractual obligations to roles, and those roles to people (and other entities).

My research has also involved a number of collaborations, involving researchers located in Australia, the UK, and the US. This has included two months spent earlier this year as a Visiting Research Fellow at the University of Kent in the UK. This time was spent investigating electronic contract issues in collaboration with Dr Steven Neal and Professor Peter Linington. This collaboration was undertaken as part of a DEST grant, "Monitoring for B2B Contracts".

The research has also produced a number of publications, and I have been an author on six papers, including a journal paper (to appear in Data & Knowledge Engineering, Elsevier), two in IEEE conferences and one in a Springer-Verlag conference (EDOC 2002 and 2003, DEXA 2002), and two in IEEE workshops (ITVE 2001, Policy 2001). The paper I presented at the 2003 DSTC Research Symposium, "Iterating Over Time", was awarded best paper.

In summary, my research performance has been demonstrated in my honours dissertation and research assistant work, and more recently in my full time role as a research scientist.

Working at DSTC has been a rewarding experience that has, I believe, equipped me with a number of valuable skills. At the same time, I have long held an interest in the proposed research topic — I have been spending some of my own time over the last few years investigating it — and this interest has gradually increased to the point where pursing this topic has become a primary goal. I wish to take on this topic because I want to research something that is deep and substantial, because I want to grow from the challenges of doing so, and because I have strong belief in this topic’s significance.

Friday, July 14, 2006

Favourite Recipes: Mixed Dhal

Mixed Dhal

I quite like Indian food, and have quite a few Indian cookbooks - and this is the best dhal recipe I've come across. Works well as a meal by itself.

It's from the "Indian Cooking" recipe booklet in The Australian Women's Weekly's "Great Cuisines" cookbook series. It's one of those little ones that you can get in the checkouts in places like Coles and Woolworths for around $5.

I've modified the presentation of the recipe a bit.

Serves 6

  • lentils
    • yellow split peas, ½ cup (100g)
    • red lentils, ½ cup (100g)
    • split mung beans, ½ cup (100g)
      You can vary the proportions of each type no problems, even omitting certain types. (though I've never tried it with more of the split mung beans)
  • ghee or oil, 2 tbsp
  • onion, garlic, ginger
    • brown onions, 2 med (300g), finely chopped
    • cloves garlic, 4, crushed
    • fresh ginger, 1 tbsp, grated
  • spices
    • black mustard seeds, 3 tsp
    • black onion seeds, ½ tsp (not essential)
    • ground cumin, 1 tbsp
    • ground coriander, 3 tsp
    • ground tumeric, 1 tsp
    • chilli powder, 1 tsp
  • tomatoes, 2 x 400g cans
  • veg stock, 2 ½ cups (625ml)
  • for finishing
    • cracked black pepper, ½ tsp (not essential)
    • cream, 1/3 cup (80ml) (not essential)
    • fresh coriander leaves, 2 tbsp, finely chopped

  • prepare lentils
    • Rince peas, lentils and beans, separately, under cold water; drain.
    • Place yellow split peas in a small bowl, cover with water; stand for 30 mins, drain.
  • Heat ghee/oil in large heavy-base saucepan; cook seeds, stirring, until they start to pop.
  • Add onion, garlic and ginger; cook, stirring, until onion is browned lightly.
  • Add ground spices; cook, stirring, for 1 minute.
  • Add split peas, lentils, beans, undrained tomatoes and stock; simmer, covered, about 30 mins or until red lentils are tender.
  • Just before serving, add remaining ingredients; stir over low heat until just heated through.

Tuesday, July 11, 2006

This news story says:

Under an edict issued by a leading Mogadishu cleric, the five-times daily prayer required by the Koran will be enforced under penalty of death, a move that appears to confirm the hardline nature of the city's Sharia courts.

"He who does not perform prayers will be considered as infidel and Sharia law orders that that person be killed," said Sheikh Abdalla Ali, a founder and high-ranking official in the Supreme Islamic Council of Somalia (SICS).

Monday, July 03, 2006

Cringely: It's time to own our own last mile

Interesting article by Robert Cringely (though it seems to drift off into unrelated stuff towards the end):

Frankston points out that we build and finance public infrastructure in a public way using public funds with the goal of benefiting economic, social, and cultural development in our communities. So why not do the same with the Internet, which is an information infrastructure?

A model in which the infrastructure is paid for as infrastructure -- privately, locally, nationally, and internationally can create a true marketplace in which the incentives are aligned. Instead of having the strange phenomenon of carriers spending billions and then arguing that they deserve to be paid, we'd have them bidding on contracts to install and/or maintain connectivity to a marketplace that is buying capacity and making it available so value can be created without having to be captured within the network and thus taken out of the economy.

One Page Comic 'We Can't Teach You This Because....'

Not bad. Here.

Nice Doonsebury on Intelligent Design


Highly Recommended Paper on Hierarchical Temporal Memory

If you're interested in AI, or just more generally, in the ways in which computers can be applied, I'd highly recommend this white paper (PDF, HTML-sans-diagrams) about the Hierachical Temporal Memory (HTM) technology being developed by Jeff Hawkin's research company, Numenta. I think it's a real step forwards.

I think the name 'Hierarchical Temporal Memory' gives the mistaken impression it's just a storage technology. But it's not really. It's really a generic mechanism for learning and predicting causes. Specifically, a hierachy of causes. The reason they refer to it as a memory is that the system encodes a memory that is, in effect, of the sorts of causes it has come across before, which it uses as part of its prediction. The significance of what it does is explained in the paper.

What HTMs actually are is a theory of the general functioning of the neocortex. Neocortex seems to provide such a generic mechanisms for learning and predicting across hierachies of causes.

I've spoken earlier about Steve Grand's excellent book Growing Up With Lucy, which is actually a very similar theory of how the neocortex works. The thing about this paper is that it goes more into concrete details of how the system works.

If you're interested in knowing who the people behind HTMs are:

The founders of Numenta are Jeff Hawkins, Donna Dubinsky and Dileep George. Hawkins and Dubinsky were co-founders of Palm Computing and Handspring and have worked together for a dozen years. Hawkins is known as the architect of mobile computing products such as the PalmPilot and the Treo smartphone. Dubinsky was CEO of Palm and Handspring, and is CEO of Numenta. Dileep George has worked with Hawkins at the Redwood Neuroscience Institute since the summer of 2003 and has extended and formalized Hawkins' theory of the neocortex.[source]

Saturday, July 01, 2006

Ultrasound Tool Developed For Regrowing Teeth and Bones

According to this story, some Canadian researchers have developed a tool for regrowing teeth and bone (a bit more detailed info here).

It is still at the prototype stage, but the trio expects to commercialize it within two years, Chen said.

The bigger version has already received approvals from American and Canadian regulatory bodies, he noted.

Identifiers for News Stories

Just a quick thought... It would be interesting if you could subscribe to a particular news story, so that you could automatically receive any followups to it.

So if you subscribed to a news story about someone being injured, you could receive followups about how they ended up fairing. In the current dynamics, follow ups tend to be buried under 'the news of the day', or are simply not done.

This would require some sort of identification of a particular story. That in itself would be interesting, if different parties reporting on it were using the same identifier.

I'm not saying that this would be a click of the fingers to get right, but I think that it could be worked out. I think it's pretty inevitable that things will evolve towards making the information we deal with more explicit, and this could be a place for a small step -- of many such possible places -- towards that.