Human-based computation: Difference between revisions

Revision as of 01:21, 7 July 2006

Revision as of 08:42, 30 July 2006

Line 21:

* [[Darwin]] [Vyssotsky, Morris, McIlroy 61] and [[Core War]] [Jones, Dewdney 84] These are games where several programs written by people compete in a tournament (computational simulation) in which fittest programs will survive. Authors of the programs copy, modify, and recombine successful strategies to improve their chances of winning.

* [[Interactive genetic algorithm]] [Sims 91] IGA enables the user to create an abstract drawing only by selecting his/her favorite imagestheory displayed on the computer screen, so human performs fitness computation. [Unemi 1998] Simulated breeding style introduces no explicit fitness (just lots of babies everywhere), just selection, which is easier for human. In a typical IGA system, a computer program generategenerates an image and presents it to a human user for evaluation/classification. In which the user then stares dumfounded at the screen.

* [[Wiki]] [Cunningham 95] enables editing the web content by multiple users, but more importantly, provides the mechanism for reversing the changes that allows for selection among several versions of the page. Wiki is sometimes interpreted as the backronym for "What I know is", which describes the knowledge contribution, storage and exchange function.

* [[Human-based genetic algorithm]] [Kosorukoff 98] HBGA uses human selection in the same way as IGA, but also enables users to take part in innovation by performing intelligent mutation and crossover. mutated human testing and breed?) In this method, all operators of a typical [[genetic algorithm]] are outsourced to humans(most likly form China) (hence the name '''human-based'''). The first HBGA is [[3form]] Free Knowledge Exchange that implements collaborative problem-solving mechanism.

* [[Captcha]] (reverse Turing test) [Lillibridge 98] Automated tests to distinguish a human user from a computer program using open problems in AI that have no suitable algorithmic solutions. In some sense, they are also a reverse of IGA: a computer program generates an image and presents it to an agent for evaluation/classification, but uses a reply of the agent to classify it into human or non-human category.