Search: Dreams and a Million Miles
May 15, 2014
The write up points to Mr. Page’s “founder’s letter.” I thought there were two founders. Guess not. Anyway, the point of the write up is that Google search is not particularly good.
More interesting is the assertion that Google wants to answer a user’s questions. No problem. Well, a tiny glitch. Users often do not know what question to ask. If one thinks about answering questions, the first step is to figure out what the user needs to know at a point in time in a particular fungible and non fungible context.
This, in my opinion, is a tough problem. Much tougher than figuring out what ad to show to a user. We know how well that works using technology like Applied Semantics’ methods and the cruft of a decade added on to what was Oingo.
Here’s the portion I noted:
But despite these advances Page admitted that “in many ways, we’re a million miles away from creating the search engine of my dreams, one that gets you just the right information at the exact moment you need it with almost no effort.”
Page explained: “Improved context will also help make search more natural, and not a series of keywords you artificially type into a computer. We’re getting closer: ask how tall the Eiffel Tower is, and then when ‘it’ was built. By understanding what ‘it’ means in different contexts, we can make search conversational.”
A couple of thoughts.
Google is one of the few companies positioned to deliver relevant results. The reality is that revenue takes precedence over relevance in my view. I give lectures for a an organization that focuses on law enforcement and intelligence professionals. My most popular lecture provides examples of ways to get through Google’s bullet proof vest of baloney and ersatz information. Why? Useful information IS in the Google index. Finding the information has become a major problem. And getting relevant results is becoming more difficult because as boat anchor access gives way to mobile access, users have neither the screen real estate, time, or expertise to fire queries into a foot of Kevlar.
Because Google is the go to system for metasearch systems, the results in metasearch systems display similar problems with relevance. The Google approach rewards non information, making the metasearch systems output results that are not much more useful than those produced by straight on Google queries.
Google’s belief or fantasy that it is a search and retrieval system reaches back to the dorm and Backrub. The reality is that search is the digital equivalent of a mule. A “query” is generated by some human, system, or algorithmic action. The retrieval system then matches the “query” to one of the indexes Google houses. The results are what you get when you see most Google pages.
The searcher is an award winning technology journalist who now labors in the ArnoldIT vineyards. The situation is looking for a restaurant in Washington, DC, a short walk from Google’s office. The journalist keys in Cuba Libre to get the address. Note that the journalist, one of my law librarians, and I are standing in front of Cuba Libre. The Google system shows a map and the map does not locate Cuba Libre. The journalist looked at me and asked, “What’s up?” My response: The new Google is what’s up.
I am delighted that ZDNet summarized a founder’s letter. I am thrilled that the myth of search is being propagated. I am happy knowing that as long as Google sees search as a million miles away, my three hour seminar will attract a large audience.
Now navigate to Google and try to find out where I am giving my next talk. Let me know how that works out for you using Google search.
And enterprise search? Don’t get me started.
Stephen E Arnold, May 15, 2014