Amazon Mechanical Turk is a crowdsourcing marketplace for work that requires human intelligence. Mechanical Turk provides an on-demand and flexible work force in the cloud that Requesters can tap into through its web user interface, command line tool, or its web service. Developers can leverage this service to build human intelligence directly into their applications.

While computing technology continues to improve, there are still many things that human beings can do much more effectively than computers, such as identifying objects in a photo or video, performing data de-duplication, transcribing audio recordings or researching data details. Traditionally, tasks like this have been accomplished by hiring a large temporary workforce (which is time consuming, expensive and difficult to scale) or have gone undone.

Mechanical Turk aims to make accessing human intelligence simple, scalable, and cost-effective. Businesses or developers needing tasks done (called Human Intelligence Tasks or “HITs”) can use the robust Mechanical Turk APIs to access thousands of high quality, low cost, global, on-demand workers—and then programmatically integrate the results of that work directly into their business processes and systems. Mechanical Turk enables developers and businesses to achieve their goals more quickly and at a lower cost than was previously possible.

Amazon Mechanical Turk provides access to a marketplace of workers whenever your business needs them. Mechanical Turk can connect you to a variety of skill sets and capabilities via a global workforce that can help you to complete your work whenever and wherever you need it.

One of the difficulties that businesses face is dealing with demand spikes for certain types of work. You may have one million images that need to be tagged today, or 100,000 new catalog entries that need to be validated, but only have that need once per month. Amazon Mechanical Turk allows you to get that work done easily when you need it, without the difficulty associated with dynamically scaling your in-house workforce.

Amazon Mechanical Turk provides several mechanisms to help you ensure the quality of the work you receive. In addition to worker qualifications, plurality is another mechanism to help developers maintain high quality levels. This functionality enables multiple workers to submit answers to the same HIT. When a specific number of workers reply back with the same answer, the HIT is automatically approved. Additionally, you only pay for what you deem to be good results; if the output is not up to par you simply reject the work and do not pay the worker.

The overhead and fixed costs associated with hiring and managing a temporary workforce can often be significant. By leveraging the skills of on-demand workers from around the world, you can significantly lower costs while achieving results that might not have even been possible with just a dedicated team.

Amazon Mechanical Turk enables developers to create applications that were previously not possible. From search results enhanced with human input, to automated image identification, to distributed content creation, to podcast-to-text transcription, companies are relying on human intelligence and Amazon Mechanical Turk to power a whole new class of applications.

Amazon Mechanical Turk is well-suited for processing images. While difficult for computers, it is a task that is extremely easy for people to do. In the past, companies have used Mechanical Turk to:

  • Tag objects found in an image for easier searching / advertising targeting
  • Select from a set of images the best picture to represent a product
  • Audit user-uploaded images for inappropriate content
  • Classify objects found in satellite imagery

Companies with large online directories or catalogs are using Mechanical Turk to identify duplicate entries and verify item details. Examples of this have included:

  • De-duplication of yellow pages directory listings
  • Identification of duplicate products in an online product catalog
  • Verify restaurant details such as phone number or hours of operation

The diversification and the scale of the Mechanical Turk on-demand workforce allow you to gather a breadth of information that would be almost impossible to do otherwise such as:

  • Allowing people to ask questions from a computer or mobile device about any topic and have workers return results to those questions
  • Filling out survey data on a variety of topics
  • Writing reviews, descriptions and blog entries for websites
  • Finding specific fields or data elements in large legal and government documents

The following are examples of companies taking advantage of the power of the Mechanical Turk on-demand workforce to understand and intelligently respond to different types of data:

  • Podcast editing and transcription
  • Human powered translation services
  • Rating the accuracy of results for a search engine