Vampire

Tuesday, February 15, 2005

 

Amazon Mechanical Turk

The Amazon Mechanical Turk (MTurk) is a beta web service by Amazon.com that enables computer programs to co-ordinate the use of human intelligence to perform tasks which computers are unable to do. Requesters, the human beings that write these programs, are able to pose tasks known as HITs (Human Intelligence Tasks), such as choosing the best among several photographs of a storefront, writing product descriptions, or identifying performers on music CDs; Providers can then browse among existing tasks and complete them for a monetary payment set by the requester. To place HITs, the requesting programs use an open Application Programming Interface.

Requesters can ask that providers fulfill Qualifications before engaging a task, and they can set up a test in order to verify the Qualification. They can also accept or reject the result sent by the provider, which reflects on the provider's reputation. Currently, a requester has to have a U.S. address, but providers can be anywhere in the world. Payments for completing tasks can be redeemed on Amazon.com via gift certificate or be later transferred to a U.S. bank account of the provider.

Contents

  • 1 About the name
  • 2 History of the service
  • 3 Third party programming
  • 4 Related systems
  • 5 External links

About the name

The name Mechanical Turk comes from a certain chess-playing automaton of the 18th century, which was made by Wolfgang von Kempelen. Called "The Turk," it toured Europe beating the likes of Napoleon Bonaparte and Benjamin Franklin, but turned out not to really be an automaton at all: a chess master hid in a special compartment controlling its operations. Likewise, the Mechanical Turk web service allows the machines of today to perform tasks they aren't yet suited for without having human help.

History of the service

The service was launched on November 2, 2005, and is currently in beta. Since its launch, Mechanical Turk experienced an exponential growth in its user base; this was the result of many users who heard of "the Turk" by word of mouth and the many HITs that were in the system. In early to mid November, 2005, there were tens of thousands of HITs, all of them uploaded to the system by Amazon itself for some of its internal tasks that required human intelligence. Web traffic grew to a massive amount near the beginning of December 2005. Since then, the HITs in the system decreased, and by December 20, 2005 there were less than 100 HITs on the average page load. By January 2006, new types of HITs have been set up such as selecting the three best restaurants in a city, and third party HITs are beginning to appear as well.

Third party programming

Programmers have developed various browser extensions and scripts designed to simplify the process of completing HITs. According to the Amazon Web Services Blog, however, Amazon appears to disapprove of the ones that automate the process 100% and take out the human element. Accounts using so called automated bots have been banned.

Related systems

MTurk is comparable in some respects to the Google Answers service offered by Google.com; however, the mechanical Turk is a more general service that can potentially help distribute any kind of work tasks all over the world. The Collaborative Human Interpreter by Philipp Lenssen also suggested using distributed human intelligence to help computer programs perform tasks that computers cannot do well. MTurk could be used as the execution engine for the CHI.

In some Marketplace systems, one can buy/sell particular services, something that could also be done with Amazon's Turk. For instance, both Rentacoder and IPSwap allow people or businesses to request the development of computer programs; software developers can the bid and get the contract to write the program.

Several other computer systems and algorithms use distributed human intelligence, although not in such a general way as MTurk or CHI. Google's PageRank algorithm obtains relevance data for web pages from the web links placed by humans everywhere. The ESP game gets people to collaborate in labeling images. The stardust@home project will recruit humans worldwide to help find interstellar dust particles within millions of images taken from the Stardust spacecraft dust collector.

External links


Comments: Post a Comment



<< Home

Archives

September 2004   February 2005   February 2006   March 2006   July 2006  

This page is powered by Blogger. Isn't yours?

My Photo
Name: