The Amoeba Project

I realized in the meeting yesterday, Node London is much about decentralization. Just like the internet, the nodes should be distributed and mainly working on its own without a central organ.

The discussion yesterday reflected this attitude. It was very open and de-centralized. This, as you can imagine, makes decision processes more difficult. The group relys on peers to voluntairly make decisions and carry the consequences.

My personally, I believe that there are advantages as well as disadvantages in decentralization. The advantage is its fairness. Every peer has the same influence. On the other hand, one of the big disadvantages is, the system is not efficient. I can speak from experience, because the Swiss state is struggling with its inert nature. However, the Swiss citizens have much more to say about the faith of the country.

The reason, why it is so inefficient is, because the opinions are "floating" around and everybody tries to be as fair as possible. However, it is not possible to take a decision that pleases everyone. And none of the members of the group want to be in the position to decide something that has negative influence on the others. 

There is no discussion about numbers. Numerologists do not lose time disputing if 7+4=11 or not. This is because they have agreed on the operations and the mechanisms beforehand. This idea can be mapped to a group of people discussing in a group. If everyone agrees to a certain mechanism of decision making, there will not be any discussion about any decision because it is at the end the system itself and not an individual person that takes the decision.

It is a big challenge to lay out these mechanisms ant this is basically what I will be diealing with in the next couple of months.

Today was a meeting for the Node London events that took place in March 2006 in London.

The goals of the meeting was part review, part how to move forward. I was invited by Ruth Carlow to join the meeting as Armin Medosch advised me to apply the Amoeba project within the Node London context.

During the discussions of the organization committee i took notes in order to get a feeling of what the tool would need to be capable of.

I will post the mind map later, however there are a couple of points on which the group was taking decisions on:
- dates (e.g. for meetings, deadlines, things that need to get done)
- subgroups (who is member of which subgroups that are attached to subtopics - like task forces)
- what needs to be done next moving forward?(setting priorities)
- where to spend money? (e.g. more in marketing, more in publishing, more in artists?)
- what happens to left over money? (how is it invested etc.)
- who represents Node London abroad or in meetings? Who goes to speeches?
- who decides on which software to use and which hardwarde?
- how are funds to be raised?

If anyone knows or is interested in providing a trial field for the project, let me know. I am currently planning a trial with the mazine. However, I will also contact other sites for cooperation. Who these sites are is will not be disclosed until further notice.

The Amoeba Project will be a centralized system. For this particular purpose, a centralized system seems more suitable. Not only does it ease the integration but also the maintenance and updates. The most popular services on the web including google, myspace, flickr, del.icio.us, digg and others are centralized system. I do not know of any networked de-centralized project (except p2p that does it for legal reasons)

Yesterday was my presentation in front of the assessment team. The feedback was constructive, and the following steps will be undertaken in order to bring the project to the next stage:

- search for actual community to make a prototype. (there is a Node London event in May I will join)
- speak to various people that have worked on similar projects
- analyse existing projects that have failed. Try to analyze what exactly may have caused the failure. Make field tests. Analyze the interface between social systems and technical solutions
- make the predicate more accurate. something is missing in the predicate bit of the data triple.
- how to deal with statements about statements?
- what does the fact that a peer may be related to other statements make their statments more or less valuable?
- don't kill complex systems with easy workarounds.
- i am making explicit what is usually implicit. outline.
- read the essay from jamie king about.. (?)
- read literature from Ted Nelson from Oxford University
- get into Tim Burners Lee's critical discussion about the semantic web.
- is there any way of integrating multimedia into the project?

There will be four options on how existing third party communities will integrate with the folksonomy:

1.) No tags at all. Just two options for the rating person: Agree/Disagree

2.) Fixed tags (as categories) where the community administrator(s) can define which tags should be available

3.) Dynamic tags where there are no restrictions for the tags and they can be made up by the rating person "on the fly".

4.) A combination of suggested tags plus an option in the form of an input field where the rating person can make up a tag "on the fly" in case none of the before tags should apply to the content or requirements.

The only restriction regarding tagging should be, that the language of the tags should be preferrably English, could however vary but should certainly not vary within a community.

Since the concept of the Amoeba project underlies the idea of the Subject, Predicate, Object triple, like for examle in semantic Web architectures or RDF, RFID would perfeclty fit into the concept as the identifier of real world objects.

Imagine yourself walking down the streets, going into a restaurant and scanning the RFID tag of the location you find at the door and either rate or retreive ratings of this particular tag. This would be an aspect that is perfectly thinkable in this context.

The technical framework that will be used are web based technologies such as PHP, MySQL, XHTML1.0, CSS2.0, JavaScript and various flavours of XML (RSS, AJAX). Wherever possible, the technical infrastructure should be free and open source.

In order to get maximum recognition and spread, the API (Application Programming Interface) will be designed as open as possible and it will be using standards such as XML, RDF and WSDL. This should theoretically allow any third party community to plug in their user base into the rating system and therefore into a potential voting/decision making system. The latter however is just a suggestion. Obviously the metadata gathered within the community can be used in any kind of context beyond our current imagination.

Amoeba is, sometimes more, sometimes less, related to a number of other projects in the web that are dealing with the attempthttp://www.blogger.com/img/gl.link.gif of creating information about information in a semantical way.

* RDF (the idea of metadata)
* Tim Burner Lee's "Semantic Web" (metadata)
* Digg (rating/digging of text articles)
* Slashdot (moderation)
* Stumble Upon (rating/tagging of websites)
* Flickr.com (tagging)
* Del.icio.us (tagging)
* Technorati (tagging)

"Amoeba" is the code name for the my project at the Ravensbourne College of Design and Communication. The project Amoeba is, as the name sais, an amorph and still to be formed attempt of creating a piece of software, that allows to create dynamically context related profiles of peers that are being used to weigh opinions at a later stage.