All the studies show that one of the most common reasons for going online is to search for health information.

People who are skilled at health searches have mastered their own little tricks and shortcuts, sparing themselves hundreds of cites which don’t apply to them. That kind of searching skill takes time to acquire and, not to be picky, still is harder than it ought to be.

Perhaps not for much longer.

There is a good possibility that the searcher will be able to enter a question, rather than a term, and get results which intelligently link keywords even when they cross scientific boundaries.

We can’t wait. Not only will this be faster for the skilled searcher, it will mean anyone can do a skilled search.

Don’t race to your toolbar yet but it does not appear we should not have to wait too terribly much longer.

Even now, a searcher can try that method. It can be surprisingly successful….or not. This question-method is worth one quick try first, just to see what pops up, until this latest-greatest tool is available.

Of course, we have to say, if...it becomes available. Not guaranteed yet.

The Economist

…”A few weeks ago, your correspondent witnessed a demonstration of a medical diagnosis and treatment model built by NetBase in Mountain View, California. It was far more impressive than Wolfram Alpha. The machine parsed a statement such as “Magnesium is known to help with high blood pressure often due to stress” and extracted not only the keywords (“magnesium”, “known”, “help”, “high”, “blood”, “pressure”, “stress”) — as Google and any other keyword search engine would do. But it also recognised that “magnesium” was a chemical entity and “high blood pressure” and “stress” were medical conditions — in much the way a semantic search engine might. Then it went on to define “magnesium” as a possible treatment, and the phrase “is known to help with” as a problem-solution relationship and “often due to” as a causal relationship.

“The important thing about such relationships is that they are the “connective tissue” between problems and answers—and the key to a whole new approach to asking questions and getting meaningful answers. NetBase calls such relationships “semantic lenses”.

“Apart from actually understanding statements like the one above, the NetBase engine retrieved all the benefits and problems associated with magnesium as well as products containing the element, and organisations selling it. It even identified the various causes, drugs, complications, treatments and useful foods for dealing with hypertension. In one instance, the NetBase model found, in minutes, the best drug for treating a rare disease that had taken a skilled researcher months to identify. (Emphasis added)

“The health model was just one example. NetBase has built a library of semantic lenses that can be applied to practically any topic. It has already delivered a research-and-development engine (called Illumin8) to Elsevier, an Anglo-Dutch technical publisher. It is now building a market-research engine for a large household-goods company capable of surveying up to a billion people at a time.

“It cannot be long before NetBase (or one of its fledgling rivals) creates an engine capable of inventing things people never realised they needed. Just imagine the productivity gains such an innovation would unleash.”

Source: The Economist, May 29, 2009