IBM drives text mining via standard search framework is an article about the release by IBM of an open framewrok for knowledge management and text mining.
Another IBM project is TAKMI (Text Analysis and Knowledge Mining) – see Text analysis and knowledge mining system.
Thanks to the Tomorrow’s Professor Listserv at Stanford I cam across the
” href=”http://www.glion.org/?a=6202&p=1512″>Glion Declaration on “The University at the Millenium”. The declaration of the identified IT and alliances as two opportunities:
Two opportunities ó new alliances and the use of information technology ó now offer the possibility of expanding the range and usefulness of scholarship and providing unprecedented benefits to society.
StrangeBanana: Computer-generated webpage design is “Just Another Website” with a twist. You can see a random new design for the site just by pressing “refresh”. They also show you how to use any design you like for your own page. I got this from the Biased History of Interactive Media site blogged before.
The founders of Mackarel, David Groff and Kevin Steele, have created A Biased History of Interactive Media which has a great chapter on When multimedia was black and white which reminds me (an old HyperCard programmer) of the “good old days.” The chapter is a great little intro to early multimedia on the Mac. They mention a number of then new HyperCard stacks like Robert Winter’s CD Companion to Beethoven’s Ninth Symphony leading up to their Mackarel Stack, which I remember. The interface to the chapter is deliberately HyperCard-like.
Chapter 2 is about “When bevelled edges were cool” … remember those days? Shudder?
The Metacombine Multidimensional Visualization Project has been developing Java interactive visualizations for digital libraries. They use prefuse: an interactive visualization toolkit to create the multi-dimensional visualizations. They also have a good section on the dimensionality reduction techniques they use like principal component analysis.
IBM runs a ITtoolbox or “Networking Knowledge Base” for the “Professional IT Community. They have a section on Business Intelligence Text Mining with good links to papers and a link to a section on tools.
StÈfan has a blog entry on Dogpile Search Comparitor tool which produces a venn diagram of the results from different search engines. The interface, however, has a fatal flaw. You can’t link to the resulting sites that are common to more than one search engine (and presumably are the ones you want). The little pills in the middle – which stand for the shared results link to a Dogpile listing, not to the resulting site itself.
The Globe and Mail today had a section on Education with an article about A new world of digital libraries by Kate L. Barrette. The story quotes Michael Ridley, Chief Information Office and Chief Librarian at the University of Guelph to the effect that now they spend 63% of their acquisitions budget on digital resources (31% for print) compared to five years ago when it was 20% digital. That is a big change in the ratio of digital to print.
Continue reading Online Library Budgets
The paper I read every morning, The Globe and Mail has twice in the last week (today, Nov. 2, and Saturday past) had a highly designed front page focused on just one story. They are going “layout crazy” where they don’t present the news, but instead try to get attention. Why?
Continue reading Globe and Mail: Layout Crazy
Wendell Piez pointed me to an article in Ubiquity: An ACM IT Magazine and Forum with the title Why People Don’t Read Online and What to do About It. It is by Michelle Cameron (Vol 6, Issue 40, Nov. 2-8, 2005) and is short and easy to read. It nicely summarizes the reasons people don’t read off the screen and how to write for those who scan. The last guideline is:
Know when to stop writing. Like now.