Azara Blog: Information Technology for Sustainable Development

Blog home page | Blog archive

Google   Bookmark and Share

Date published: 2005/04/20

The sixth and final lecture of the university's Third Annual Lecture Series in Sustainable Development (2005) was given today by Nazli Choucri, a political scientist who is behind the Global System for Sustainable Development (GSSD) project at MIT. (Their home page is one of those confusing front pages where it is not obvious how to proceed. Going via google to find the real home page was easier.)

Now Choucri flew over from Cambridge, Massachusetts, and then will go back again after a couple of days in the UK. Meanwhile her talk was beamed back to MIT via a video link. Nobody seems to notice the irony in all this. Apparently the practitioners of "sustainable development" see nothing wrong with flying 10000 miles round trip to give a lecture when instead you could give the lecture on your home ground of MIT and make the video link go the other way. At least she did not talk about "sustainable transport". In last year's series, Fred Salvucci of MIT did exactly the same round trip, again unnecessarily, and he did talk about "sustainable transport" (needless to say, the car being the source of all evil in the world). Some people have no shame. (Well, he was the originator of the "Big Dig" in Boston, one of the worst public sector projects of all time.)

There is a tendency among many in the "sustainable development" community to use cringe-making jargon, and this lecture was certainly full of that: "e-knowledge access", "new technologies for knowledge networking", "de-massification" (huh?), moving from "'supply chain' to 'knowledge chain'", "strategic cyberpartnering", "enabling new actors and leverages", etc., and with a meaningless chart with "knowledge content" on the x axis and "knowledge value" on the y axis. Social scientists seem to love this stuff, and it's a good reason to move money from that into science and engineering, where people solve problems instead of just talking fashionable jargon.

However there is at least some point to the GSSD project. If you get away from the dreadful jargon, what they seem to be trying to do is to catalogue work in "sustainable development" on the internet so that other people can more easily find it. Unbelievably they seem to have been given a patent for this work (the first for the School of Social sciences at MIT, according to Choucri), but then again, in the US you can patent anything.

They have introduced fourteen top-level categories or "domains" (labelled "energy", "trade", "industry", "mobility", "agriculture", "land use", "water", "conflicts", "urbanization", "consumption" "unmet needs", "population", "migration" and "governance"). Each of those is split into five sub-categories (labelled "activities", "problems", "technical solutions", "social solutions" and "international responses"), and then each of these is sub-divided further (dependent on what the top-level domain is).

This is a classic exercise in data modelling, as has been done by librarians, and more recently scientists and others, for centuries. It is a non-trivial job and largely thankless. No domain is black and white, and almost anybody can come up with examples which seem to show the categorisation does not quite make sense. (Someone in the audience asked about "air" since "water" was listed explicitly. It's hiding somewhere but only an expert would know.)

And with google around, you have to wonder if all this work is worth it. Basically GSSD is a "sustainable development" portal. Is going to such a portal going to save you time and effort when you are trying to find something in the "sustainable development" field? First of all, google only searches on syntax (word matches independent of context). As it happens that is a very powerful methodology if you combine it with a good ranking system for webpages, which of course google has.

Many people believe (but with little success so far) that you should be able to do better searches if you take account of semantics (what something means, so in particular in what context it appears). GSSD would be one approach to doing that. You get experts to review documents, those that are good enough get listed and categorised, and those that are not (according to the reviewer) do not. This can be an extremely expensive procedure, not only to review the documents in the first place, but then also to maintain the information (because documents on the internet move, etc.).

The proof will be in the pudding, and as with all portals there is a chicken and egg problem (if you are not known you might not be found, unless google ranks you high, and if you are not found you will never become known). Choucri mentioned that if you put "sustainable development" into google you would get millions of hits, suggesting GSSD would be much better for that kind of search (since meant specifically for it). But why would anybody put such a wide query into google. Much more likely you would ask about a specific topic, or paper, or author.

For many searches google wins hands down, but it depends on the search. If you have a paper copy of the Financial Times and want to find an article on the FT website you could go to the FT website and try to find it using the categories the FT has set up, or you could go to google and type a half dozen words from the first sentence. The latter would get you there faster almost always. Google does a better job of drilling down into websites than websites do themselves. But if you wanted to know how much a Polish zloty (stroke through the "l") is worth in British currency today then you would probably just go to the FT website, because it is dead simple to find that information on their website, and google might well flounder. So the FT website has been set up to make it easier to find currencies than specific articles.

One suspects that GSSD will be the same. If what you want falls easily into place with their categories then GSSD will be useful, otherwise probably not.

And it is a bit ironic that with "decentralization" featuring in Choucri's talk (as it does in many "sustainable development" talks) that GSSD is a large central service with a central authority deciding what can appear in the system. As someone from the audience pointed out in the question and answer session, a more decentralised approach would be for people in the "sustainable development" community just to put appropriate keywords into their HTML and let google do the rest.

Choucri said in the question and answer session, in response to a specific query, that one of her primary objectives was to get the UN to use this system. The idea is that there are lots of UN agencies and committees and lots of international agreements, and how does anyone know, for example, if a clause in one agreement contradicts that in another. Well, it's hard to see how GSSD or any computer system could ever do that since you need to understand the legal implications of the obtuse and arcane jargon found in such agreements, and that is down to human experts.

All material not included from other sources is copyright For further information or questions email: info [at] cambridge2000 [dot] com (replace "[at]" with "@" and "[dot]" with ".").