Author: Luis von Ahn
Title: Games with a Purpose
Source: IEEE Computer Magazine
Date: June, 2006
Abstract of Major Ideas
Luis von Ahn’s internationally recognized postdoctoral work at Carnegie Mellon University’s Center for Algorithm Adaptation, Dissemination and Integration (ALADDIN) involves developing techniques to harness human computational power to solve problems that computers are currently not very good at solving. In his paper, Games with a Purpose, he describes how he has developed a few very simple online games that are fun and addicting for humans, but also provide valuable information for image search databases that would otherwise be extremely expensive to collect.
Dr. von Ahn describes how computers are very bad at recognizing the content within photographs and images. Image search databases do not return very accurate results because the text that is indexed to describe the images is simply gleaned from image captions or adjacent verbiage. Using his ESP game von Ahn has been able to collect massive amounts of data to help improve image search functionality. The game works by displaying an image to two anonymous players on the Internet. Each player enters words that they feel describe the image. When both players hit on an identical word, a match is made that is then collected and used as a search term for the image displayed. The ESP Game has collected over 10 million player generated labels since it was introduced on October 5, 2003.
Critical Evaluation of Major Ideas
With the power of computers advancing as rapidly as it has since the ’60s, we tend to take for granted tasks that are simple for us to do, but technically monumental for computers to handle. Dr. von Ahn has had the insight to look beyond the complex and obvious solutions to find simple and elegant ways to use human brainpower for a positive outcome. The game industry is one of the largest areas within interactive media. In his paper, Dr. von Ahn outlines several other ways simple games could be used to solve difficult problems utilizing the cognitive abilities of the online community. These include language translation, monitoring security cameras, improving web searches with human aids, and summarizing important documents.
I regularly use Luis von Ahn’s examples as teaching tools in my interactive media design classes. In Advanced Scripting Languages I use Dr. von Ahn’s reCAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart) project as an example of how a web application can detect whether the input received in online forms is coming from a human being or a spambot (a spambot is software designed to collect e-mail addresses from the Internet in order to build mailing lists for sending unsolicited e-mail, also known as spam – en.wikipedia.org). Dr. von Ahn created the CAPTCHA technique and then redeveloped it under the new name reCAPTCHA. reCAPTCHA forces people to type in words that appear semi-scrambled when entering data into an online form. Computer based optical character recognition (OCR) is incapable of deciphering these words. One of the words is the security key, while the other was scanned from a book, but not recognized by OCR. By having a person enter both words they proving they are human and aiding the digitization of books for the Internet Archive (archive.org) at the same time.
Implications for Design Education
We can all learn from Luis von Ahn’s research. His CAPTCHA technique is commonly used and imitated on major sites like Yahoo, Myspace, and Google as well as millions of simple news blogs and personal sites. His ideas are simple, elegant and inspirational. Students in interactive media design are frequently bogged down by the complexity and difficulty of the technology involved in web applications. Examining Dr. von Ahn’s projects helps my students understand that the best answers are not always the most difficult ones.