Recommendation engine
First of all the presentation gave us some insight into why they have developed their recommendation engine. The use of statistics and data to provide a better user experience and increase the value on own content through leveraging the use of the service is not something new. This is a trend that Tim O’Reilly described as one of the core principles of the Web 2.0 era. The basic behind the Digg recommendation engine is similar users, and since the release in July (Upcoming) they have had pretty impressive results with for example an increase in digging activity by 40%.

Show your impact
I think that one really innovative use of the graph is to let the user know what kind of effect they actually make up for other users within their activity. More specifically, it means that if I do not like an article and votes it down (bury), I will be told how many people I made sure did not get to look at that article (as in not recommended to by the recommendation engine). This is a great and fundamental part of the networking effect, and takes the effect of direct user interaction to the next level. I hope they will succeed in implementing this innovative thoughts.

Recommendation as a service
Digg is seeing the recommendation engine as part of a more sophisticated network where the engine will act as a service for publishers to help them expose other data within their site. This means that a user visiting BBC may be exposed to articles that is recommended by Diggs recommendation engine. I think Kevin Rose and the rest of his team has some way to go before publishers will team up with this kind of business models, but it is not very difficult to see that from Diggs perspective this would be like a home run.

Targeted advertising – that’s what digg should focus on
I think that a more interesting perspective is looking at advertising. If Digg could use the recommendation engine, all the user activity and the graph (similar users) as a basis for an ad engine that will give more targeted ads to the users, and form the basis for a more effective advertising they will have something that is extremely interesting for others. In its simplest form this could be like this: If I as a user choose to click on an ad, then this ad should be targeted as relevant for all my similar users and therefor be displayed to them. My opinion is that advertising is something they should look more into related to the recommendation engine that they now are implementing.

At the event Dave Morin talked about The Story Behind The Facebook Platform, and the figures he gave was amazing(!). Since the launch of the Facebook Platform on the 24th of may 2007 (kick-off video with Mark Zuckerberg) they have reached:

• 90.000 + developers
• 5000 + apps total
• 100 + new apps a day

Facebook is now the 6th most-trafficked site in the US. They are bigger than eBay, and will (I suppose) soon pass Google(!).

Day one summary (skiped some speakers):

Mike Arrington of TechCrunch talked about start-ups and web companies focusing on how to be successful. I think his bubble 2.0 discussion was interesting focusing on the fact that in 2006 more money was used buying start-ups and web companies than invested with venture capital.

Tara Hunt of Citizen Agency had a presentation about communities. The presentation probably worked well for newbies@communities.

Last.fm with Matthew Ogle & Anil Bawa Cavie, gave a brilliant presentation. Interesting topic from the guys was the use of attention data as a foundation for ranking the songs (popularity). They used tag clouds as example and showed us how much more elegant use of attention data is rather than just focusing on traffic etc. It gave me some nice ideas for a project I am working on.

Werner Vogels from Amazon gave a nice presentation focusing on the hosting-service from Amazon. The company has had great success with this service.

Kevin Rose form Digg announced that they intend to support OpenID. Actually this was the most interesting point with the whole presentation, a some what disapointing talk. Maybe I just had too high expectations to this presentation!

Day two summary (skiped some speakers):

Mark Anders from Adobe gave a great presentation, just in my alley . Use simple tools in an elegant fashion and the geeks are with you. He presented Flex a rich Internet application framwork. I will definitly look more into this interesting technology.

Chris Wilson from Microsoft. I had hopes for a talk around the future of Internet Explorer (MSIE), but Chris’s focus was the past and present and what happend with Microsoft and the web during 2001-2006

Khoi Vinh from New York Times talked about design issues at NY Times. This newspaper is really working hard and develops some pretty nice stuff.

Simon Willison gave an outstanding presentation of OpenID. OpenID may be the solution to web-authentication. It is promising, and several of the big sites out there is now supporting OpenID. Some issues still have to be looked into. Just made a nslookup for larre.myopenid.com and the result was:

P:\>nslookup http://larre.myopenid.com
…
Navn: http://larre.myopenid.com
Address: 67.137.230.67
P:\>

This says me that we might see some issues related to performance. Hope performance will be an issue at some openID articles.

Daniel Applequist from Vodafone gave a presentation about the future of mobile. The presentation may have been interesting for people not so into mobile development.