[Sigia-l] IA and the MS Office Paperclip

[Joe Dombroski]

Hello,

The seemingly simple Office Assistant (the paperclip "Clippy" character 
in Microsoft Office who pops up to offer help), has interesting 
connections to information architecture.

The amazing aspect of the feature is its ability to infer goals and 
model events based on categorizing user events with Bayesian 
classification systems (see http://research.microsoft.com/adapt/ for 
more information), but I found the implementation annoying as other 
users noted in a study of Clippy: 
http://xenon.stanford.edu/~lswartz/paperclip/ which also discusses how 
labels (another IA connection) influence user reactions.

As people who are interested in the way information is structured and 
presented, though, what do you find helpful and/or unhelpful about the 
Office Assistant?

Thank you,
Joe Dombroski

PS. Since I noticed many people are regulars on the list and I've been 
lurking for a few weeks now, I'll introduce myself briefly as an 
undergraduate student at the College of William and Mary in Virginia.  
I'm double majoring in linguistics and psychology.  Information 
architecture is something of a hobby that might be put to use if I 
decide to go into usability, product design, or a related field.  If 
you've got any recommendations for someone interested in the field, I'd 
  be grateful to hear them.