How to use this Page

Daniel Russell, Google Search Quality & User Happiness
2010 Alaska Library Association Conference, opening keynote speaker

Lewis & Clark left without a decent map
it’s a complicated world out there and you don’t want to end up like the Donner Party (hey, go that way; it looks good)
what does the current information map look like?
let’s be adventurers but keep our eyes and minds open

did a demo of Google Earth
cost to put the flyover together = $0 and four minutes of time
Google will crawl it within 48 hours
when Lewis & Clark published about their trip, it took 10 years
we see the world differently, and the library isn’t what it used to be
stacks are no longer a core competence — the information landscape has radically changed

1200 exabytes of new content are generated each year (1.2 yottabytes if that helps or 1.2 billion terrabytes)
3.6 zetabytes per person per year (mostly music and video)
libraries don’t have to curate and manage that — it stream to you
text words per pseron per year = .1% of that total
the good news is that the amount of reading per person per year has gone up by 3X since 1980 (primarily due to internet access); happening online, not print
so need to develop new skills and new literacies

showed Google Books
can click on the places in a book and travel to all of them
can actually recapitulate Huck Finn’s journey down the river

LoC has 10 terabytes of text data or .01 petabytes
he has 2 LoCs at home
an exabyte = 50,000 years of DVD or 10 billion copies of The Economist (there aren’t enough trees in Alaska to print them all)

we’re supporting this renaissance of access to print culture at the same time we’re expanding online content
1.5 million out of copyright books that can be printed for $8 each

do you care about all of this as long as you can get to the stuff that you care about?
what Google is trying to figure out is how can I read your mind from the couple of words you gave me — which pages you want to see of theirs out of all of those exabytes of data?
it’s not just text anymore

mentioned Hans Rosling’s TED talk about visualizing statistics
mentioned Baby Names Voyager
Google bought software to add visual statistics to Google Docs
the cool part is I can type my name and see when my name peaked
is this a book? no. is it a visualization? yes. but it’s also interactive. where/how do I catalog this?
these kinds of interactive documents allow you to understand in ways that were not possible before
showed what happened to names that begin with vowels during the 40s and 50s — “the valley of the vowels“
the answer to what happened is in the hard consonants
no one knew this until they could see it in this visualization
our notion of what constitutes information and librarianship is changing

how do people search now?
suppose you’re Google and you get the query “jaguar” — what do they want?
one of the differences about being Google though is that you’re at a reference desk where a billion people a day ask the question

what about “iraq?” today, it’s the way; 15 years ago, it was probably antiquities
Google sees queries shifting a lot
“latest release Thinkpad drivers touchpad” = I know exactly what they want
“ebay” = in the top 10 most popular queries in English per day
“google” is also in the top 10 queries per day — why?? are they trying to cause the recursive meltdown of Google’s servers?
there are 20,000 ways to mis-spell “Britany Spears” (and they all want pictures of her)

one of the interesting things they do is use machine-generated algorithms
they don’t have to mis-spell a new celebrities name 20,000 times — their users will do that for them
that’s how informaiton works now

he goes to peoples’ ho