Textmode

Today I received an invitation to the Powerset labs:

Powerset is excited to invite you as one of the first members of our Powerlabs community. In Powerlabs, you’ll be able to interact with demonstrations of our technology, and share your ideas on how we might improve our natural language engine. Powerlabs is all about gathering feedback and we greatly value your input.

Powerset is, in short, a company that wants to revolutionize todays search engines with a natural language based approach of processing the web (and search queries).

Powerlabs is basically a set of demos that show how the whole process is intended to work. You can test some predefined and custom use cases as well as a tool called Powermouse.

Powermouse lets you extract information in little pieces, directly out of the index of facts that the Powerset technology uses (Wikipedia).

It consists of 3 input boxes: something connection something and it should be pretty obvious what they mean. You can start by entering something in the first box, and get a result like this:

As you can see I used Einstein and the result shows something like a tagcloud for the weight of possible connections. The second tagcloud are again Things that the term Einstein is related to.

Below these boxes is the actual list of connections the index has, including a short description from Wikipedia that was used to derive this relationship. If you click a specific connection keyword, the system returns more specific results about the entered connection.

The cool thing about this is that you can enter parts of a relationship, for example if you want to find out how something relates to anotherthing, which will then give you all connections that are indexed.

Some other integrated examples let you build even more natural search queries like:

• What did X write?
• Who painted X?
• What does X make?
• Who works for X?
• Who did X aquire?
• What did someone say about X?
• Who critizized X?

Where the parts are interchangable, I think you get the point.

If you’re interested in this technology, sign up for an invitation, I think this could become very useful, especially in the semantic web corner.

Google Open Social will launch on Thursday. The aim of this project is a set of general APIs for applications that can be included in Social Networks.

What is it good for?

Well, this means a developer writing applications for social networks like Facebook, Myspace etc will no longer have to learn the conventions and markups for each single network. Instead he can write applications using a general interface, which then runs on all platforms that support this API.

This is a huge benefit for developers, since you no longer have to pick out specific platforms with the maximum benefit. I also think that this might help smaller platforms to grow, since most of them are just left out right now.

Initial partners are said to be Orkut, Salesforce, LinkedIn, Ning, Hi5, Plaxo, Friendster, Viadeo, Oracle, Flixster, iLike, RockYou and Slide.