A weblog by Will Fitzgerald

Even more on index size

Jean Véronis got slashdotted on the index size story.

The NCSA is somewhat disclaiming the index size study published on the NCSA website, and the study has been revised somewhat. Véronis is keeping up: his latest post, Technologies du Langage: Yahoo: Missing pages? (4), includes the original stronger disclaimer, and discusses why the revised study is still flawed.

In conclusion, this new study is just as biaised as the previous one. It still counts numerous junk documents returned by Google, and doesn’t address other important issues.

I’m a bit surprised that no one else has mentioned the English bias of the NCSA study.

In some ways, the thing that most disturbs me about this flap is that the Vernon Burton has had his name taken off the study (including the old study), leaving his two students ‘on the battlefield,’ as Véronis put it. This morning, Google and Yahoo still have a version of the August 8 version of the study the in their respective caches, showing Burton’s name as a co-author.


Comments are closed.

%d bloggers like this: