sorted by: recent | see : popular

Content Tagged with marketing + Google

ouch!

In un intervista su Technology Review rilasciata da Peter Norvig, Responsabile Ricerche di Google, alla domanda su come il motore di ricerca riesca a valutare e fornire i risultati con un buon grado di accuratezza ha così risposto:

“A livello macroscopico teniamo traccia dei clic degli utenti. Se selezionano il primo risultato, e non tornano indietro per cercarne altri, probabilmente hanno trovato quello che cercavano. Se invece scorrono più pagine della ricerca o riformulano la domanda cambiando il numero delle keyword e la composizione della frase, allora i risultati ottenuti non corrispondevano a quello che gli utenti cercavano. Un altro modo che utilizziamo è quello di selezionare a caso alcune query e sottoporli ad un certo numero di persone per ricevere delle valutazioni sui risultati ottenuti. Si tratta di collaboratori che assumiamo per avere esclusivamente un giudizio. Li formiamo per identificare siti Web spam o fraudolenti, dopodiché registriamo le loro valutazioni”

….Rimane da capire come i risultati dei “focus group” entrino poi nell’algoritmo e come ed in che modo modifichino le ricerche ed i risultati di Google.

So I RTFM’d: watched VicG’s show, read a gushing GoogleBlog or two as well as the official Google PR, sampled the API docs for containers and apps, read Berlind’s stuff, read the pieces from Marc the first, Marc the second, Mark Cuban and Dion Almaer.

And I still have some questions:

1) Compatability. Is there a test suite to ensure that containers implement this in a consistent (enough) way? While Vic paid homage to Ballmer: “distribution, distribution, and distribution,” Google’s PR utterly bowdlerized McNeeley: “learn once, write anywhere.” Given the heterogeneity of container business models (linked-in, vs. Hi5, vs. Second Life or potentially Adult Friend Finder) it seems likely that this will get implemented quite differently if there isn’t a way to enforce consistency, through a compatibility mark or testing suite or something.

2) “This Standard is Standards-based.” Hmm. I get the “standards-based” part….

3) I like this bit about open sourcing the container side components. Nothing drives de facto standardization like availability of some source. Are container providers going to hold off until they get their hands on this, so they don’t have to re-implement when the new OSS stuff ships?

4) Does Open Social give me a way to gather up friends from different social networks? I.e. punch a hole in the walled gardens and plant some nice veggies? Build a hub app that wires up different Open Social Containers to find a friend that I might like in another network, or what networks my friends belong to? David Berlind describes how difficult it would be to map the identity semantics of different social networks here , but then Joe Kraus spins exactly this vision quoted in this piece in the Times: “The long-term vision, he said, was to enable social networks to be portable: “You want your friends to go with you — you don’t want them to be locked up.” How am I going to connect up friends identities across multiple networks, and break down the walls around these gardens? As Marc1 says here: “Just because MySpace and Friendster say they’re gonna support OpenSocial - is completely different from them actually allowing a user to export their list of friends - with unique emails for each friend. This I gotta see.” Or is this going to be like Unix on minicomputers: the same, but different enough?

5) Where is Amazon? Amazon Associate, probably the second biggest river of dollars on the web (after AdWords) was absent. Wouldn’t it be great if you could query for friends and interests from a service that has every member’s credit card on file (uhh..not to steal from them but to have a very low friction way to effect purchases). There is some cool stuff that uses Amazon on Facebook now (Facebook log-in reqd).

6) I can write an app in Ning’s sandbox now with Marc2’s caveat: “real good-old-fashioned will-probably-break kind of beta!” and for the Orkut sandbox if I’m invited. When can the rest of us write apps? What is the schedule for other containers?

7) If I want to build a container, is there anything I can do beyond putting a plaintive petition in the Google Developers Forum? What if I am not a FOG (friend of Google?) at the moment?

8) Will there be any consistency on business models between containers? I believe the Sherman Act says any such consistency must be a coincidence, which is one of the reasons that open standards bodies exist, to allow competitors to collude in constructive ways that might be illegally anti-competitive behind closed doors. See question 2 above.

9) When Facebook opened up their platform, the stunner was that developers kept all ad revenue. Without this, support would have been a lot slower in coming. In this model, no social network container is incented to give away revenue to get apps onto their networks competitor’s networks, so can app developers expect any sweet spiffs from any of the Open Social containers?

10) I want to go to the campfire next time. Hey Vic, can I be a FOG too?

"What people are really saying is: Why don't Google release more code? And I think that's a valid thing to say. I think it's kind of important they want to see Google release more code, and I agree with them."

Insertion of AuctionAds or Google AdSense into your content is really simple. There are some Bots that are designed to help you to do this, but I feel that they are not required. Just follow these steps and you will get your AuctionAds displayed wherever