72.3 percent of Americans experience "search engine fatigue" when researching a topic on the Internet ... More than three out of four (75.1 percent) of those who experience search engine fatigue report getting up and physically leaving their computer without the information they were seeking.As Greg Sterling says, this sounds like a common hunger for personalized search.
Kelton asked survey respondents whether they wished that search engines like Google could, in effect, read their minds, delivering the results they were actually looking for. . . That capability is something that 78 percent of all survey-takers "wished" for.
See also a March 2005 post where, in response to Udi Manber statement that "we are not in the mind reading business," I said, "If you need to read minds ... well then you better read minds. [Searchers will] think it's your fault, not theirs, if you don't give them what they need."
4 comments:
As Greg Sterling says, this sounds like a common hunger for personalized search.
Hrm.. it also sounds like a hunger for relevance feedback tools, such as query term suggestions, clusters, etc.
Why? Because if you think about it, a cluster of results is actually a result. If the user sees a list of document clusters, rather than a list of documents, they might be able to see, in those top 10 clusters, that aspect or facet of the word "Leopard" that they really meant (either OSX or animal).
In this way it would appear to the user as if the search engine was reading his mind. Because the engine would show the correct cluster, ranked there, in 2nd place. Wow, how did the search engine know that I meant Leopard the animal? There it is, ranked second! (By then clicking that second cluster, they would be taken to a list of documents, all about their meaning of the word Leopard.)
This is very different from the current search engine approach, in which the user has to expend all this additional effort, typing in 4 or 5 different queries, trying to get the exact set of documents they want. Or figure out how to "turn off" personalization when it makes the incorrect guess, and then figure out how to "turn back on" the correct form of personalization.
But I'm just beating my usual drum.
At least when I search, the prevelance of retail sites is very distracting from the real results, unless that is what I am looking for. Easy enough to grok out, but even if you say put a -sales -purchase in the query they are still there. So even when I clearly state I do not want to shop, I am barraged with shopping sites.
darrell: Yes, that's what I mean, but extrapolated. Just because you like to shop from time to time on the web doesn't mean that you want to see shopping sites every time you search for any kind of item with some inherent commercial market surrounding it.
Rather, it might be better to see two clusters: Non-commercial sites and commercial sites. Or three clusters: Commercial sites, non-commercial product review sites, and non-commercial "product in action" sites.
Then, if your initial search returned a ranked list of these three clusters, you could pick which one was most relevant to you in that moment. To me, this is much better than being overwhelmed with one type of page (letting personalization pick for you). And much better than having the search engine try to mix up the rankings, ranking lower quality pages higher, if they contribute to the overall diversity of the top 10 documents (something I've heard that the engines do).
No, I'd rather like to have the ability to chose that most relevant subset of the 120,000 retrieved documents myself.
To me, the mind-reading part comes in the fact that the search engine should be able to create these categories, these clusters, automatically, without any effort on my part. The search engine should be able to read my mind well enough to know how I would categorize the 120,000 pages, if I had time to do it myself. That would be mind reading!
Thus for a query about "clarinets", the clusters would be all the various aspects of clarinets.. recordings with clarinets in them, compositions written for clarinets, places to buy clarinets, historical origins of clarinets, acoustic properties of clarinets, etc. But for a different query, say about "maps", it would have categorical clusters like maps local to your zip code, world maps, historical maps, transportation maps, census maps, etc.
Basically, each query would have a different set of clusters, based on the documents available on the web. And the clusters would change over time, based on how the web changed.
If a search engine could get that problem right, that would be much more valuable to me than a search engine that recorded the fact that I look at a lot of historical maps, and then gives me another list of historical maps (personalization) when this time I am looking for modern foreign maps.
From what the big players in the industry say, however, I don't see this ever happening. Because the common wisdom is that users are too lazy to click even a single refinement, after their initial query. Users only want to click results links, not refinement links. (Somehow clicking 3 results links, to see which of the pages best meets your information need, takes no effort, but clicking 1 refinement link does take effort? Frankly, I don't understand this reasoning. But that's the general consensus!) So I don't see any development in this direction.
But a search engine that returns one set of clusters to me when I type "clarinet", and another set when I type "maps" is actually, IMHO, reading my mind. It is understanding the fact that my mind thinks about multiple facets simultaneously, that my mind sees many aspects of an information need solution. And then the search engine would show me all the facets that my mind is thinking about. And let me then make up my own mind about what I first want to look at.
I see a difference, then, between reading my mind, and making up my mind. I want the search engine to read my mind. I don't want it to make up my mind.
hmmm...I wish there were a way to see who has paid to have their stuff at the top of the page - I am usually shopping on line and looking for special things - not the shlock that vendors pay to have put up front when the cool stuff by lesser known designers is on page 56.
Post a Comment