In the spirit of full disclosure, I had to write a journal entry for my Engineering of Enterprise Software Systems class, and I figured, hey, I wrote the damn thing, why not post it as a blog entry, since it's about search(-ing), and thus relevant to libraries? And yes, leave it to a librarian to have 8 reference for a page-and-a-half long paper.
--------------------------------------------------
This week, Google (NASDAQ:GOOG) announced “Google Instant”, featuring predictive, real-time search results, meaning you can see your search results update according to each character you type in the search box, as you type it. The company claims this speeds up searches by an average of 2 to 5 seconds per query.
Based on what I understand about databases and search technology, this would require a database call for each character typed, which, given the millions of people using Google’s search at any given moment (they recently reached 1 billion users a week, according to USA Today), would place a huge load on their servers/system.
Google’s new real-time search functionality uses AJAX (Asynchronous JavaScript and XML). This technology (or really, groups of technologies, including JavaScript, XML or JSON, XHTML, among others) allows web applications to run independently of a web page, so data can be retrieved from the server while the rest of the page loads (asynchronously). In other words, it allows different portions of a page to load/update separately, or asynchronously.
Apparently, the underlying technology is so simple, users have begun creating imitation real-time search applications for popular sites such as YouTube and Twitter (15-year-old web developer Stephen Ou built “iTunes Instant” in just three hours.)
Although it’s still pretty new, users have begun weighing in on Google Instant. Reactions range from love to hate, as is usual with any new technology. In an opinion piece in The Guardian, Charlie Brooker laments:
I'm starting to feel like an unwitting test subject in a global experiment conducted by Google, in which it attempts to discover how much raw information it can inject directly into my hippocampus before I crumple to the floor and start fitting uncontrollably.
In general though, people seem to like the speed, but are unsure if a faster search was really necessary, especially at the cost of some well-liked features that Google has dropped with this launch, including the search box that used to be located at the bottom of any page of search results, and the retention and auto-suggest feature for recently-completed searches.
Although some people may be underwhelmed by this new search feature, they are voicing concerns about some aspects of it.
The company is apparently filtering search results, and not returning items from queries that are deemed “offensive”. This not only brings up censoring issues, but can have unintended consequences, like in the case of Irina Slutsky, a reporter for Advertising Age, who complained that results related to her were being blocked because of the first four letters of her last name.
There is also concern over how Google is choosing results, and the affect that this, along with quickly appearing (and disappearing) will have on sponsored search results and ads. IBM’s Todd Watson was quoted as saying:
Let me explain: go to www.google.com and just type in the letter "A." What comes up first in the listing? "Amazon." Followed by "AOL." Followed by "ATT." I skipped on over to "I," thinking that IBM might come up first. But no, it was "IKEA." ... what in the world is IKEA doing coming up when I search for the letter "I"??
Academic Librarian. Author, website builder, lover of science, and superwholock fangirl. Cat mom, misanthropic humanist and inconsistent oxford comma user. Full of snark and boxed wine. More @
Posts
