20 Nopember 2008 11:26
Hi there, just wondering what the volume limitation is for using the "Live Search API" in my web application?
21 Nopember 2008 1:35Pemilik
If you are using the legacy API 1.1 (the endpoint you mention), you can send up to 25,000 requests per day.
If you use the new API 2.0 (see http://msdn.microsoft.com/en-us/library/dd251056.aspx ) you can send unlimited requests provided that:
you are building a customer facing application (no bots, scrapers, keyword miners, etc)
you show all the results that you are retrieving
you attribute the results to Live Search
If you intermix web results from live search with results from other sources, you make the other sources clearly distinguishable of coming from another source
you have no other advertisement on the results page but advertisement that you got from an agreement with Live Search or Microsoft.
21 Nopember 2008 14:52
if I take a look at the TOU I read
"Whether we include trademarks or logos or not, you will attribute us as the source of the Live Search results (in the manner that we may specify from time to time in applicable services documentation) and display the "attribution" data field (if a non-blank field is returned as part of the Live Search results)."
What is the "attribution" data field ?
How do we have to attribute the search results to Live search?
Thank you very much.
01 Desember 2008 17:18Pemilik
Sorry for the delay in replying - I forgot.
Currently we have not specified and "attribution" field for any public source type, so there is no action required on your part.
Should there be one, per the ToU it must be displayed.
We have this to make sure that the right attribution for third party instant answers would be correctly displayed.
Concerning the attribution to Live Search, we are not mandating any specific way to do this for the beta timeframe and we are still debating among ourselves what the most flexible way is. The tricky part is finding something that works for web apps and clients, showcasing our brand without being overly intrusive
I'd love to hear any suggestions or direct feedback on how you would like this done.
26 Januari 2009 7:52In your blog announcement for the new API you wrote:
My answer then and now is: No.
If Microsoft really wants to do something better than Google with search, it can fill a huge void in the market that Google has chosen to either ignore or brush off: commercially oriented searching. That means paying an affordable amount (based on query/result volume ranges) to be able to do search queries that are *not* customer facing - can retrieve any or all of the results whether 1 or 1 million - and for any purpose whatsoever. As far as I can tell, no company offers this capability. Rather than trying to emulate Google's advertising model, think business-to-business commerce. If you want some idea of the demand for this capability, I suggest that you ask developers, in a prominent place, if they want something like this. The size and intensity of the response may surprise you.
Amazon's Alexa offered something like this, but apparently it's been discontinued. A major reason is no doubt that their web crawl was incredibly incomplete. So it needs to actually have good, relatively complete, timely data as well.
27 Januari 2009 8:55PemilikThe ToU as it is only covers who comes in as a user of the vanilla,self-serve search offering.
Exceptions are allowed and there is a simple way to submit the request: just contact api_tou at microsoft dot com and provide more details about your intended usage.
Each request is evaluated and, if accepted, we quote access to the system based on the traffic generated.
This requires an explicit contract with Microsoft and comes with restrictions and limitations different from the standard self-serve product (for example with a daily query quota or a limitation on the query rate).
You would not be the first one.