The World Wide Web has grown too big for our limited human capabilities, all of us web surfers experience this on a daily basis when we type a query into any of the search engines, and try to fish out some information among all the results presented to us by the search engine....
Search Engines play a critical role in our use of the web and our access to information. Through indexing and ranking they determine the content of the internet. By indexing they take note of what content is available on the net, and by ranking they decide what information (meaning web pages) might be more valuable than others.
Material on the web is dynamic, it is very easy to get images and text from somebodies page, and put it again, with or without modification, on another site. Blogs and Wikis have made Web authoring more accessible than ever. One of the consequences of these "easy authoring" systems is that information on the web changes constantly, some sites appear, some disappear, and some lose their relevancy. All these changes are reflected (sometimes immediately, sometimes with some delay) in the results we get from a search engine. If we were to repeat the same query every day, and keep track of the changes in the list of results, we'd see the changes in the result that are served (new ones appear, some disappear), and also in the order of their appearance. This tracking might provide us with some insight as to what the web "thinks" about that query. Regarding the amount of results usually provided by a search engine, this seems a merciless task, which nobody is ready to repeat on a daily basis. Let the machine deal with machines.
The information provided to us by the search engine through the web browsers is encoded in HTML, the browser's language. Since this information is intended to be deciphered by a machine, it is very formatted and patterned. So instead of having some user sit at the search engine, typing the same query again and again, and manually compare the lists of results, let the machine do it - let a script do it.
The tool I propose is a script that will keep track of the changes over time in the search results provided by the Google Image search to the same query. Meaning that once the user of the script submits a search query to the script, then the script will store the results, repeat the query periodically, and will note the difference in the set of results (who is in, who is out), and the order (hierarchy) in which they are presented.
The Image tracer is available here
grew up in Tel Aviv, Israel, where she completed B.Sc's in Mathematics and Computer Science. She spent 2003 at the New Media department of the HGK Zuerich, and in 2004 she joined the Piet Zwart Institute in Rotterdam, where she is pursuing an MA in Media Design. She is interested in the life cycle and evolution of information on the Internet.