In case you hadn’t heard, there’s a new search site called Blekko that wants to tempt you away from Google, Bing, and the other big boys.
I will always have a soft spot for new search engines, because search is where I got my start. Way back in the mid-80s, when I was but a poor grad student, I became quite enamored of heuristic search algorithms. So enamored, in fact, that I wrote my dissertation about them. Regrettably, the powers that be of the time (the powers that were?) had determined that there was nothing interesting to say about search, and that the topic was no longer suitable for consideration by AI researchers.
Still, search did play a central role in my thinking and research for five years or so, and it showed up in my teaching assignments (my favorite term project assignment was a dating service--commonplace today, but still interesting back in '87). Which is precisely why I still perk up when I hear about--yet another--new search engine.
The entrance of search into the commercial realm raised some interesting questions. Back in the day, we used to assume that all users were at least reasonably sophisticated, educable to the point that they would devise reasonable sets of key words. In some cases, we actually made them learn query languages. In others, we just assumed that they knew what they were doing.
Those assumptions are passe. Perhaps the greatest challenge facing commercial search today is the range of keyword selection skills that the public demonstrates. Some users, of course, are quite good. They (we?) select keywords closely related to the information they seek, and know how to zero in on the terms most likely to differentiate the useful from the useless. Others, shall we say, do not.
How can a commercial search engine cater to such a broad range of users? And how can a tech journalist devise a reasonable test sequence capable of determining search engine performance across a broad range of users?
These are the sorts of questions that take me back--with a smile. They were the sorts of questions that engaged large portions of my thought in the 80s and 90s.
I'm always pleased to see when something that engaged me in my own little corner of the computer lab broadens into a matter of general interest.
Recent Comments