Information is power

by Andy Wilson, editor
[email protected]

With the furor surrounding recent developments on Wall Street, it’s no wonder that Oliver Stone has once again teamed with Michael Douglas in a sequel to his original 1987 Academy Award-winning film of the same name.

Due to be released later this month, Wall Street: Money Never Sleeps will reprise Douglas’s role as the unforgettable Gordon Gekko, a Wall Street stock speculator. As many of you will recall, the original film has some very memorable quotes, including some that we can publish here! These include such classic phrases as “lunch is for wimps,” “if you need a friend, get a dog,” and, of course, “…greed, for lack of a better word, is good.”

For those of you who missed the original, the story focuses around Gekko’s influence over Bud Fox, a young and ambitious stock trader who does just about anything to get information from multiple sources and use it to his financial advantage.

In 1987, both men’s behavior was portrayed as somewhat shady and, at the end of the film, Gekko is imprisoned for using “inside information” to his competitive advantage. Wall Street still remains the perfect example of how information—either obtained legally or illegally—can be used to make money. What has changed in the quarter of a century since the film was made is how information is obtained and how it can be leveraged.

Big in Japan

Today, manufacturers mine the Internet using search engines to find information they can use for their own competitive and financial advantage, predicting industry directions, collecting data on competitive products, and looking for lucrative market opportunities. Indeed, in the development of this month’s Product Focus, I used Google to discover a decided lack of frame grabber support for the proposed PoCL-Lite standard, despite the fact that many Asian camera manufacturers now support this standard.

Even to gain this much information still required hours of work. While the information was freely available, it was not easily accessible. Although Google is a good search engine, it is not a very sophisticated data mining tool. Indeed, according to Professor Zoubin Ghahramani and Katherine Heller at the University of Cambridge (Cambridge, England), the concept of querying such search engines with multiple items is still unexploited. If it were implemented, I would have gained more information at a higher level from one or two keystrokes.

Rather than use text-based search tools, Ghahramani and Heller have proposed an item-based searching method based on Bayesian Sets that automatically learn which features are relevant from queries consisting of two or more items. For example, a movie query consisting of “Platoon” and “Natural Born Killers” suggests that the concept of interest is movies directed by Oliver Stone. A Bayesian Set-based search is then likely to return other movies by the same director. Similarly, if I had typed in PoCL-Lite, Camera Link, frame grabber, and Japan, it may have been easier to locate a PoCL-Lite frame grabber.

In the future, improved tools will allow users to extract more information—making facts and figures about specific products, trends, and market opportunities more easily accessible. Because of this, the information itself will become even more valuable to users.

With an eye to making some money using a similar technique, my nephew Paul is planning a project that monitors ticker symbols transmitted by the millions of those now known as Twits Who Twitter. By capturing the data, analyzing them, and projecting the results on stock trends, Paul plans to become financially independent by predicting the direction of the stock market.

I wish he had chosen to develop some software that might help me extract trends from the machine-vision industry instead. It would have made my life a lot easier when writing this month’s Product Focus. But then, as Gekko would say, “What’s worth doing is worth doing for money.”