Welcome to WebmasterWorld Guest from 22.214.171.124
Time for me to eat a little bit of crow. I helped a friend last week by cleaning up a smallish database (about 1,000 products) so that it could be used to generate web pages. The database was from his shipping facility and it was full of strange abbreviations so the field would print out on a small shipping document and a shelf label in the warehouse. So the product descriptions worked for the warehouse staff but they were far from doing the sales job they needed to do.
Google's intention engine did the job. I just pasted in that mishmash directly and I usually got back what I was looking for, or at least I made enough progress so that I could make a second try and nail it.
Maybe Google's idea isn't so far off the mark after all.
Last week I Googled a friend. Her last name is one letter off being the same as a common proper noun. She's a prolific blogger so I couldn't believe that she couldn't be found even after several searches. Then I realized that G was showing search results for the proper noun, not the name that I had entered.
This would be a case of the feature working well for the technically inclined (you, Ted) and not for someone just looking up a friend. Hmmm.
The message I'm getting now is to stop thinking about Google as a search engine in any traditional sense. It's become more of an "information finder" that does use a search engine approach, but combines it with a lot of other technology. As long as I remember that, I'll keep my blood pressure in check.
And yes, I really was impressed by the help Google gave me on this project. It's no small feat.
I'd like to meet the person who thought up that one. Sounds like something George Orwell cooked up.
Nope, I still consider myself to be the decision engine. And even if the top brass at Google would like to be able to return "the right answer" I'm not ready to hand over my own responsibility to any cyber-mind.
But the value in this kind of slightly fuzzy information retrieval has made itself clear to me, and I was a definitely a doubter on the topic.
Just as an example, see the relative search volumes for various misspellings of "restaurants", one of the most often misspelled words on the internet:
I'd like to meet the person who thought up that one.
I think you'd have better luck meeting them in the Microsoft forum.
It's from Bing...
Of course you could try searching Google for the source and maybe to give you the 'one right answer' they'll revise your query to: 'Popular Search Engines and Their Catch Phrases' (: LOL :)
Yeah, I still like getting results for what I searched for...
I guess that makes me 'old school' or something?
it seems that initially only by chance the user interface was designed simple and from then on they only incidentally stumbled upon means that are actually satisfying for the user - which in effect is a basic requirement to make money. sometimes they don't move further at all (look at the development of adsense - nothing worthwhile since years).
instead of having a geat idea beforehand, they often enough disimprove for the worse. they measure, and if the stats indicate it's good, they retain it. but in case of hardly measurable things like user satisfaction with certain new features, they fail. rather make a survey than let stats decide in that case - even then it's hard to decide which way to go, because even relatively few users which are heavily dissatisfied can make a difference. just because a) they have a quasi monopoly on search, b) people are too lazy to change and c) for lack of better alternatives google gets along with it.
simple but annoying things: i hate the suggestion tool in the search box. so often it happens to me that when i type in some letters, it pops out and i accidentally click the first suggestion. and i hate that they show me the "did you mean.." results directly included in the serps when i haven't even clicked the "did you mean..". if i mistype, it's ok, but then i realize i mistyped and consequently click the "did you mean..". otherwise i just wanted to see results exactly for what i typed in. do not force me to use quotes to get exact results for each and every search queue! i'm no dyslexic in my native language. google treats sane users as fools. not to speak of behavioral results, which yet opens another whole different can of worms.
either they have to deploy a seperate "google for dummies" or a "google for pros" search engine, but this mishmash is just bad for overall experience.
monopolies get sluggish and slow down progress. google search experience actually has deteriorated for me. they are clearly tech driven. they badly need usability experts.
At one point I thought "why not try Amazon - they've got a good site search." But even Amazon did not have as good a success rate for me as Google did.
full of strange abbreviations
Are you saying the queries you made could have returned different results if Google hadn't figured out you were looking for book (title/author) type info?
Like, if someone else was searching for a series of widget parts, yet entered the exact same query as you did, and documents contained those same "strange abbreviations", they would instead get the best match for a widget part (and you would get a book) ?
There was real information embedded in them. It was just truncated and not appropriate for a general web page. So i'm assuming that Google's depth of semantic information was able to kick in quite nicely.
The Google blog had a short article [googleblog.blogspot.com] on Tuesday about their handling of semantically similar words that might be of interest. It also included some information about an expansion of bold words in the search results:
Historically, we have bolded synonyms such as stemming variants — like the word "picture" for a search with the word "pictures." Now, we've extended this to words that our algorithms very confidently think mean the same thing, even if they are spelled nothing like the original term. This helps you to understand why that result is shown, especially if it doesn't contain your original search term.
The "very confidently" description of their semantic algorithm is what I feel was clearly in play for the job I was doing.