Amazon Mechanical Turk

From Wikipedia, the free encyclopedia

Jump to: navigation, search

The Amazon Mechanical Turk (MTurk) is one of the suite of Amazon Web Services, a crowdsourcing marketplace 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. Workers (called Providers in Mechanical Turk's Terms of Service) 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, or the somewhat limited Mturk Requester site.

Requesters can ask that Workers 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 Worker, which reflects on the Worker's reputation. Currently, a Requester has to have a U.S. address, but Workers 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 Worker's U.S. bank account. Requesters, which are typically corporations, pay 10 percent over the price of successfully completed HITs (or more for extremely cheap HITs) to Amazon.[1]

Contents

[edit] Name

The name Mechanical Turk comes from "The Turk", a chess-playing automaton of the 18th century, which was made by Wolfgang von Kempelen. It toured Europe beating the likes of Napoleon Bonaparte and Benjamin Franklin. It was later revealed that this 'machine' was not an automaton at all but was in fact a chess master hidden in a special compartment controlling its operations. Likewise, the Mechanical Turk web service allows humans to help the machines of today to perform tasks they aren't yet suited for.

[edit] History, HIT types, and user demographics

The service was initially invented for Amazon's in-house use by Peter Cohen, to find duplicates among its web pages describing products.[1]

The service was launched publicly on November 2, 2005 and, as of 2009, is still in beta. Following its launch, the Mechanical Turk user base grew quickly, in part the result of the Slashdot effect. At that time, there were a huge number of "Human Intelligence Tasks" (HITs) 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. Most of these were related to music CD items. Web traffic grew to a massive amount near the beginning of December.

However, the number of HITs in the system soon decreased, and by December 20, there were less than 100 groups of HITs on the average page load. In January, new types of HITs were set up, such as top three lists ranking for the (now defunct) Amazon Unspun site, and third-party HITs began to appear as well. By April 2006, there were only the occasional batch of 25 HIT groups being offered, and the service had slowed to a crawl.

As of January 2007, there were new HITs being offered of podcast transcribing and rating and image tagging. The transcription HITs, mostly offered by Casting Words [1], are still posted regularly as of March 2009. Other common HIT types ask Turkers to write or rewrite sentences, paragraphs, or whole articles. These have rewards ranging from one cent to about $10. HITs which reward people for linking to or commenting on a blog, or friending a person on Facebook are also often encountered, as are surveys.

In March 2007, there were reportedly more than 100,000 workers in over 100 countries.[1]

According to a survey conducted through one MTurk HIT, Turkers are primarily located in the countries of United States and India. 3,548 people answered this question, which was included in a survey about weblogs. 2,420 (68.21%) were from the US, and 434 (12.23%) were from India. The United Kingdom, Canada, and Australia - all under 4% - rounded out the top five. The results were published in April 2007.[2] The survey also indicated that more Turkers were in their 20's than any other age-decade; this was further supported by another HIT survey conducted in August 2008, with 100 participants.[3]

[edit] Searches

In 2007, the service began to be used to search for prominent missing individuals. It was first suggested during the search for James Kim, but his body was found before any technical progress was made. That summer, computer scientist Jim Gray disappeared on his yacht and Amazon's Werner Vogels, a personal friend, made arrangements for DigitalGlobe, which provides satellite data for Google Maps and Google Earth, to put recent photography of the Farallon Islands on the Mechanical Turk. A front-page story on Digg attracted 12,000 searchers who worked with imaging professionals on the same data. The search was unsuccessful.[4] In September 2007, a similar arrangement was repeated in the search for aviator Steve Fossett. Satellite data was divided into 85 squared meter sections, and Mechanical Turk users were asked to flag images with "foreign objects" that might be a crash site or other evidence that should be examined more closely.[2] The search was also unsuccessful - the crash site was eventually found by hikers about a year later.[5]

[edit] 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.

[edit] API

Amazon makes available an API to give users another access point to the MTurk system. The MTurk API lets a programmer submit HITs to MTurk, retrieve completed work, and approve/deny that work. Web sites and Web services can use the API to integrate MTurk work into other web applications, providing users with alternatives to the interface Amazon has built for these functions.

[edit] Related systems

MTurk is comparable in some respects to the now discontinued Google Answers service. However, the mechanical Turk is a more general marketplace 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.

[edit] Criticism

Because HITs are typically simple, repetitive tasks and users are paid often only a few cents to complete them, some have criticized Mechanical Turk as a "virtual sweatshop."[6] Because workers are paid as contractors rather than employees, requesters do not have to file tax forms, and avoid minimum wage, overtime, and workers compensation laws. Workers, though, must report their income as self-employment income. In addition, some requesters have taken advantage of workers by having them do the tasks, then rejecting their submission in order to avoid paying. However, at least some workers on Mechanical Turk are people who are middle class and do the work for fun. [7]

In the effort to use Mechanical Turk in the search for Steve Fossett, a number of factors left contributors conflicted and disappointed with the results. A Wired News report on the effort found little understanding by contributors as to what they were actually looking for, a poor user interface, and a lack of insight into what others had found.[8]

Also, Amazon does not monitor the service and refers all complaints to the poster of the HIT. This hands-off approach can be seen as a prime cause of the actions evident in many of the complaints above.

[edit] References

  1. ^ a b c Artificial Intelligence, With Help From the Humans, The New York Times, 25 March 2007
  2. ^ http://pyxlin.wordpress.com/2007/04/20/relaunching-turk-survey
  3. ^ http://www.floozyspeak.com/blog/archives/2008/08/valley_of_the_t.html
  4. ^ Steve Silberman (July 24, 2007). "Inside the High-Tech Search for a Silicon Valley Legend". Wired magazine. http://www.wired.com/techbiz/people/magazine/15-08/ff_jimgray?currentPage=5. Retrieved on 2007-09-16. 
  5. ^ Jim Christie (October 1, 2008). "Hikers find Steve Fossett's ID, belongings". Reuters. http://www.reuters.com/article/peopleNews/idUSTRE4907G820081001. Retrieved on 2008-11-27. 
  6. ^ Harris, Mark (2008-12-21). "Email from America". Sunday_Times. http://technology.timesonline.co.uk/tol/news/tech_and_web/personal_tech/test_bench/article5370360.ece. 
  7. ^ Mieszkowski, Katharine (2006-07-24). "I make $1.45 a week and I love it". Salon.com. http://www.salon.com/tech/feature/2006/07/24/turks/. Retrieved on 2008-06-15. 
  8. ^ Online Fossett Searchers Ask, Was It Worth It?, Wired.com, Steve Friess, 11/6/2007 5:00 PM

[edit] External links

Personal tools