Keynote: Tim Berners Lee: Reflecting on 20 years of the world wide web.
Note by Dom: as usual this is not a real transcript but a mashup of a summary and my own interpretation of the talk.
Tim starts by talking about the program of www 2009 and in particular he sees that: Search and search optimization has a big place.Advertising science is also appearing as an important topic, which shows the tendency to also study how to efficiently monetarize the Web.The move to the mobile web is also definitely reflected in this year’s program.
He then moves on talking about the changing environment, mentioning the great hardware developments, the ever-growing data set. He also talks about things which remain (and should remain?) the same: the use and active creating of standards, the still very important web of document and the still growing web of communities.
He goes on with more technical concerns, talking about mashups he mentions the fact that we now really need a serious and clean trust system. Decentralized modular installation is also a feature he’d like to see up on the web soon. Just as we get Debian packets (e.g. using aptitude) we should be able to easily download and deploy applications and parts of the Web infrastructure directly from the Web independently of our platform.
Modularity in general is highly important for Tim. He would like to see a more modular Web where features can really be re-used (Dom: Fulllllllllly agree with that ;-))
The next topic is open social networking. He starts talking about the WAP which did not take off because it was not the web, it was an isolated island not quite the Web. Mobile phone really became first class citizens when they actually implemented Web! access. He sees similarities in the was social networks currently behave, Facebook is an island so is Twitter. However, this has started to change. OpenId and more generally transportable identities should enable us to transport our identities and contacts from one island to the other. A tremendously important point for Tim. We should really get our data back!
He then talks about Linked Data which is also important in terms of social identities. Linked Data can help to infer on our social networks as well as on our several identities (Dom: e.g. TimBL is Time Berners Lee is creator of the web, etc.). He further explains that Linked Data is an attempt to extract our knowledge and capture it on the Web. For instance the answer to a question such as “how do you cure cancer” should be captured by organized, linked data on the Web. He would like to see a system appear using which you could extract the semantics of stored data. He takes the example of a relational database system containing company data, having a system that could understand this data and capture the company internal knowledge of that database would be very valuable.
He then talks about social concerns, in the field of privacy we should have systems going beyond simple access control we should also be able to capture acceptable use. For instance a doctor should be able to get access to all my data to save my life but if this doctor is also a friend of mind he should not be able to access this data out of the “saving my life” context.
To conclude Tim explains our role as Web scientists should be to build a platform for others that follow. We should also avoid assume what they are going to use it form. We should focus on our bit, other will do theirs!