Not many investment information sites have a "chief scientist" in the management lineup, especially somebody who has a Ph.D in computational linguistics. In fact, I never encountered such a title until I checked out MindfulEye.com, a Vancouver-based outfit that recently launched a fascinating new tool for public companies and investor relations outfits.

Why would an Internet site hire somebody with such a rarefied academic background? There's a clue in Maite Taboada's curriculum vitae. Her master's degree in science is from Carnegie Mellon University, where she worked in the Center for Machine Translation. That's right, she patrols the frontier of communication between man and machine.

At one time this field offered academic satisfaction but little prospect of commercial utility. But the staggering quantity of information flooding through the Internet -- the good, bad and malicious -- has created an opportunity for an automated service that can not only filter it but also assess the mood of the market.

Enter Lexant, which MindfulEye calls "a new class of artificial intelligence technology that can read and understand language as it occurs on the Internet." The site says Lexant uses natural language processing neural networks and hundreds of rules of grammar "to determine the true meaning of words in context."

Is this gobbledegook? Possibly. But if it works, it will help investors deal with the most baffling aspect of the stock market -- so-called market sentiment.

Beginning investors have trouble with this concept. They learn about such concrete indicators as price-earnings ratios, which are used by investors to determine whether a stock is cheap (low ratio) or expensive (high ratio). But then they see investors price one stock earning $1 a share at 10 times, and another stock also earning $1 a share at 30 times.

Even more confusing, investors may suddenly decide one day that the stock with the P/E of 30 should really be priced at only 15. Why? Because of market sentiment.

Largely because of the Internet, those changes in market sentiment are communicated with unbelievable speed, often through myriad online chat groups. Top managers of public companies who fail to stay on top of them can look out of touch.

MindfulEye began offering Lexant last month to public companies and to those who advise them on investor relations in a service called Surveillance. The software monitors online news and discussion for developments which influence market sentiment, gathering 250,000 to 500,000 pieces of financial content a day.

But it goes beyond just the words, says Scott Kerr, who heads business development for MindfulEye. "Lexant does a deep linguistic analysis, looking at phrases and idioms and putting comments in context, to produce a comprehensive snapshot of what the online world thinks about a company right now."

A good analogy is the kind of interested but skeptical attention an experienced investor pays to the latest pitch from a broker. "I would never suggest that anybody should trade stocks on this sort of stuff alone, but I believe that it will be useful to individual investors as well as to the public companies themselves," Kerr adds.

That's why he sees online brokers as a market for MindfulEye, too. He also believes they could find it a good tool to offer to their clients and plans to offer something designed for that group next year.

Kerr says the site has several corporate customers for the Surveillance service. And there are also benefits for securities regulators as they try to monitor what information and commentary is passed around the Internet.

And computational linguistics? The Association for Computational Linguistics -- yes, there is one -- describes it as the scientific study of language using electronic digital computers. So now you know.