Judging Tomorrow’s Technology with Today’s Behavior
I have stumbled across a nice blog article, titled “The (Big) Problem For NY’s Hakia and Other Google-Killing Search Engines” . In the article, Henry Blodget mentions the difficulties for new startups like hakia and Powerset to be the users’ behavior and satisfaction levels of the current search engines. Points are well taken, yet there is a missing component which I could not resist to write about.
In retrospect, I was one of the first users of analog cell phones. You know… the Motorola era. At that time I was quite happy with it. Reception was OK, it did the job. If someone came to me those days and said, “hey look there is a digital cell phone coming up so you can take pictures with your cell phone.” I would, and probably did, say “who cares.” Like me, many people may have been responding to future ideas in this manner. Namely, judging tomorrow’s technology with today’s behavior.
Semantic search is tomorrow’s technology. It is really not a choice, niche, or fad. It is the unavoidable evolution of language-related technologies. Even in Star Trek episodes we were introduced to it: a computer that talks back with some intelligence embedded.
To be more concrete, I can immediately say that today’s expectation and satisfaction levels are directly related to “popular” queries, the fat-tail, or the tip of the iceberg. What is greatly missing is the “unpopular” queries, the long tail, or the bottom of the iceberg. These “unpopular” queries are what makes search experience a personal experience challenged with the uniqueness of each individual brain, and its wishes. That’s where statistics end, and the voting algorithms fade away.
When approached in this fashion, Henry Blodget’s question becomes more accurate if we put a time-frame to it. For example, I was not expecting videos shot from a cell-phone to be widely used in a 5 year time-frame since the day I first heard about the idea. Accordingly, one must ask “when?” There will be a time when semantic search will make a significant entry, and it will eventually take over. What is the time frame? Otherwise, one will be forced to imagine Mr. Spock talking to the computer using keywords.. That would be intolerable. Don’t you agree?
September 29th, 2007 at 4:46 am
In my IE7 I have made Hakia as my default search provider. I used the procedure provided by Microsoft. I dumped Google and Live for Hakia.
What I found is that when I think in English and use English (rather than keywords) I find Hakia right on to the result I want.
With Hakia alone, I will see the on the horizon the English use to be above 10% after 2 years.
With Hakia acquired today by Google or Microsoft and English use actively promoted, I think that English use may reach over 50% after 2 years.
September 29th, 2007 at 5:05 am
To add to my previous comment, there are some queries which are totallly impractical or not even thought of when your mind thinks keywords. With your mind free of keyword limitations you generate much more English queries than what you would use if you used Google or Live. Hence 50% with English promoted is practical.
As an example, with Google or Live my Bollywood query may be
shah rukh khan movies
but with Hakia you may do
list all Shah Rukh Khan movies
lish all Shah Rukh Khan movies with heroine Kajol
list all Shar Rukh Khan movies in which he is a bad character
you may fill in your favorite Hollywood artist instead of Bollywood’s Shah Rukh Khan
I have not tried these on Hakia. I am going to.
September 29th, 2007 at 5:16 am
Unfortunately, Bollywood information is not complete enough or Hakia does not go deep enough
September 29th, 2007 at 5:22 am
do a Google on any of the above and they really suck!
December 19th, 2007 at 2:50 pm
Oh, and did not know about it. Thanks for the information …