As far as I can tell, it’s still not exceptionally well-known that the homepage for the Microsoft Zune is actually www.zune.net, and not zune.com. You’ll find two things if you assume the latter: that there’s no page for you at that address, and until recently, if you get forwarded on to your default search engine, chances are pretty good it won’t help you figure out what went wrong. Our result page, in fact, had no results at all. Oops.
In the past, our handling of searches that look like URLs had been very simple: the web results contained only documents from that site. Searching for zune.com, you’d get documents only from zune.com. Sometimes this was great, usually it was okay, and sometimes it was pretty unhelpful. We were unhappy enough with it that we decided to try something a little different.
With our recent release, we’ve significantly changed the way we handle searches that resemble URLs. We show a lot more results instead of a restricted narrow set and that doesn’t just include pages that mention the URL you’ve entered.
Try these searches, I think they’ll illustrate it best: zune.com, cityofredmond.gov, live.com. And if you’re looking for a more literal interpretation of the URL—for instance, to get sites that mention a given URL—try wrapping it in quotes: “zune.net”.
Let us know what you think!
Thanks,The Live Search Team
seems to work pretty well. good luck!
I agree, does seem to work quite well. How will you determine what is the most relevant at that time? For fast changing website (e.g. blogs) you might just want to have the info from the url, instead of info from other url's as well. Cheers.
hello live search team,
this is off topic since i cant find any other way to contact anyone... i am doing a bit of an evaluation of some search engines and have come across some confusions here.
*where on earth are the sliders for changing relevancy rankings i keep reading about?
*if they are gone, why cant i find anything about why this is so?
*why do your help pages still say they are there?
help pages dont have anything about e.g. stemming that's also mentioned in this blog...? are they really out of date?
this is all really for now, dont want to completely hijack your post and apologies for changing the subject. if anyone feels like helping a gal out, you could send me a bit of an email at sovietta@gmail.com
thank you v much
claire.
Below is what happened in search today, as reported on Search Engine Land and from other places across the web....
Nice improvement. But, yet, remains unanswered the matter arised by Interaction Design, that is, how you weigh relevance in such situations.
"We did not find any results for zune.com."
Awesome... Keep up the good work!
I agree, does seem to work quite well. How will you determine what is the most relevant at that time? For fast changing website (e.g. blogs) you might just want to have the info from the url, instead of info from other url's as well. Cheers
Can the relevancy of grouped words be changed?
I've noticed that if I type if a group of words that (in my mind) go together, LiveSearch seems to only find individual words not the pairs or triplets. For example, if I search for Microsoft Word, I want to find articles on Microsoft Word, rather than information on Microsoft and say, the Word (i.e. the Bible). Another example, if I search for cross cultural, I should get information on cross cultural issues like cross cultural medicine, travel, etc. not information on crosses and cultural separately. (I am an anthropologist, hence, the my choice of examples, here.) Some common examples work, such as searching for bread and butter, yields as expected bread and butter, not just references to bread and references to butter. But on the whole, your competitors seem to do this sort of grouped word seasrch much better.
How to improve? Have your ranking engine look for words and rank them based on how far apart they are. For example, if I search for moon rock, I want information on rocks from the moon, not imforation on the moon and Rock music. Otherwise, good.
