Sometimes, the story is as good as the moral. Sometimes, the journey is as good as the destination. So in Information Retrieval, why are we too often satisfied with producing results, but not with explaining how the results were arrived at?
Daniel Tunkelang talks a lot about the need for transparency in search, in being able to give the user enough understanding and control over the results produced, so that the user can engage the search engine in a dialogue and instruct it to the point that results improve. Only with transparency can the user truly engage in Exploratory Search.
I would like to suggest a name for this goal of transparency: Explanatory Search. If the search engine can explain to the user why certain results were retrieved, then the user has the best chance of explaining back to the engine a refined description of his or her information need. This can only make the user experience better. Exploratory Search requires Explanatory Search.
Pingback: Information Retrieval Gupf » Creating Passionate Search Users
I’m for it, but what about the IP and competitive advantage for commercial engines? For example, Wikia Search could afford to expose everything – it didn’t invest its time and money in building the search engine/crawler/etc. they used. GYM spent lots of millions of dollars on their sw to gain competitive advantage.
It’s a changing world. Just like people are saying that the newspapers have to evolve or die. Or that the music industry has to evolve or die. Find new business models, find different ways of competitively distinguishing themselves. There is no point in propping up a failing system, just because they spent lots of money to gain competitive advantage. Same with search engines. They too have to adapt, and find a different way to gain competitive advantage. Will the entrenched search engines change? Probably not. No one likes to take their own medicine. But then we, the users, need to continually call attention to what we’re missing out on when the search engines essentially lock up all the good/interesting/useful information behind their “digital rights management” (DRM) opaque and non-transparent algorithms. Because only then will conditions be right for something better to come along.
Hmm.. did I just compare lack of search engine transparency to music DRM? I guess I did. I think the analogy fits.
Pingback: Information Retrieval Gupf » Is the Ad-Sponsored Web Search Market a Conversation?
Pingback: FXPAL Blog » Blog Archive » Search is not Magic
Pingback: Explanatory search « IREvalEtAl
Pingback: Who has sighted (not just cited) Belkin 1980, “ASK”? « IREvalEtAl
I don’t think I’ve ever “sighted” Belkin 1980. But I have read (“sighted”) Belkin 1982, “ASK for Information Retrieval: Part I Background and Theory” from The Journal of Documentation, Volume 38, Number 2, June 1982. In that, Belkin, Oddy, and Brooks lay out the entire ASK hypothesis.
So I apologize for citing 1980 in my blogpost, when I’ve only read 1982. But I have indeed read the theoretical basis of the ASK notion, written by Belkin, from that time period. Were I writing a formal paper, I would make sure and track down 1980.