13 November 2006

09. Web 2.0 is So-o Yesterday

Last week was not exactly a slow one for news, so it was surprising to see yesterday’s New York Times devote prime front page space to an article on the drive for Web 3.0. And the article must have struck a chord since it is now listed as one of the ten most popular in yesterday’s paper.

John Markoff gives a competent account of the main elements of Web 2.0, noting many of the characteristics in Tim O’Reilly’s article, and identifying the ‘mashup’ as the exemplar. Markhoff calls Web 3.0 the ‘semantic’ web since its objective is to generate meaningful responses to plain language inquires. Although the effort is in its ‘infancy’, Markoff says that both large firms such as IBM and Google, and a host of small firms, are devoting considerable resources to it because they visualize large pay-backs. Doug Lenat, who has been developing an AI system for a couple of decades, has ‘implied’ that his Cyc AI system can already search the web and answer a natural-language question such as: “what American city would be most vulnerable to an anthrax attack during summer?”

Perhaps in time we will see Web 3.0 systems in operation, and some may eventually trickle down from government and business purposes to meet needs in the academy. But as Dan Cohen has pointed out a couple of times, Web 2.0 applications for scholarly purposes are beggarly by comparison with those for commerce.

That assessment was reinforced by my exploration of a half-dozen applications available on the Progammable Web.

The Ajax Map Comparison links the maps provided by Google Earth, MS and Yahoo, so that when the user alters the zoom level or direction of one, the others change accordingly. Thus, the user simultaneously sees three different maps of the same area. Each service provides slightly different information. For example, Yahoo provides more extensive labelling, while MS provides highway route numbers at much lower zoom levels than Google.
I encountered
two shortcomings. One is that there doesn’t seem to be any way to initially focus on your desired area except by knowing the longitude and latitude, or by using the directional arrows to inch your way towards it. You can’t simply type in ‘
Boston, MA’ for example. Second, all three provide satellite images of an area, and one can toggle back and forth between that and maps (there’s a useful hybrid option also); but only MS provides a ‘bird’s eye view’ which is an isometric perspective from perhaps a few hundred feet off the ground. But when the user invokes the ‘bird’s eye view’, the system crashes, probably because there are no corresponding views in Google or Yahoo.

Utility to the historian? If you are scoping out a particular area or terrain, it could occasionally be useful to have three different mappings of it so as to chose that which best provides the desired information.

Since my primary interests lie in early modern history, much of the focus on current affairs in available mashups is not likely to be useful. What kind of mashups would be of use to me?

- when looking for a book, a mashup that would check a few user-specified libraries, as well as Amazon and Abebooks.

- one that would search a few on-line diaries, identify when there were entries for a common date, and array the results on a timeline, or see when the same proper names occurred and extract the neighbouring information, and if place names, plot them on a map.

I tried Open Searches A9, and searched for L. Lessig in Wikipedia, Amazon and NewsbyLine.com. It brought up generally relevant information in all three, probably the most that can be expected at this point from a simple aggregator. If one could add Abebooks, and a few libraries, it would meet one of the needs I noted above.


Blogger William J. Turkel said...

John, you might also be interested in http://worldcat.org/

9:37 a.m.  
Anonymous Anonymous said...






7:06 a.m.  

Post a Comment

<< Home