94 slides / 12 secs per slide / As mixing data / As dealing with reality
Seb Chan, "some people think a web site should be neat... a lot of neat, and a lot of order. But if you look closely at order, there is a chaos. Our sites at the macro level look modular, and at the micro level resemble favelas." How to agree on standards, when we haven't made anything yet? We can start with museum calendars... because really, how hard can calendars really be in 2008?
WORD SALAD SOLUTIONS:
solution: scape > aggregated third party feed, > nice backend > slapped over a "trust-o-meter" > Frontend interface with microformats RSS+ GeoRSS+ Django... But how can we really be certain that any of this is headed in the right direction?
EXTRACTING MEANING FROM THE MESS:
If you Query Wikipedia > "show me all the sitcoms set in NYC," then all the sitcomes set in NYC are pulled up. So why can't we do this in museums?
Google 2007... we still do this by hand... Each of us are creating multiple collection records...
biographies, by (other) hands, crowdsource using our audiences ... resulting in an aha-moment: so maybe tags are our answer! Natch.
There is valuable data created by our users - like tags, that help with over 11.2 million successful searches per day, misspelled is auto-corrected. But we are often faced with "Nothing Found. Did You Mean the Following?" Reuters bought out ClearForest, which essentially is a text analysis tool to wipe over your entire collection.
TAGGING WITH CALAIS TO CREATE META-DATA TAGS
If you Query Wikipedia > "show me all the sitcoms set in NYC," then all the sitcomes set in NYC are pulled up. So why can't we do this in museums?
Google 2007... we still do this by hand... Each of us are creating multiple collection records...
biographies, by (other) hands, crowdsource using our audiences ... resulting in an aha-moment: so maybe tags are our answer! Natch.
There is valuable data created by our users - like tags, that help with over 11.2 million successful searches per day, misspelled is auto-corrected. But we are often faced with "Nothing Found. Did You Mean the Following?" Reuters bought out ClearForest, which essentially is a text analysis tool to wipe over your entire collection.
TAGGING WITH CALAIS TO CREATE META-DATA TAGS
Calais creates tags that are now meta-data... and now begs the question, who creates meta-data? Librarians can create meta-data, but they're aren't that many who are doing this now. Do we still need to clean all of our data by hand? Essentially, we are left to parse and prune, parse and prune.
About NSW beta: a locality explorer in Sydney that maps the news over ariel images found on the web, cross-referenced with heritage museums and wikipedia pages. When one clicks on "read more," you get the object from the Heritage Trust site. One can pull up art, the bios of the artists, and then 5 years from now, it can be sent to you again.