The Elements of Search, Part Two
Last week we barely scratched some fundamentals of searching, specifically How to Search and Where to Search. These need more elaboration.
1. How to Search
In searching, you can reasonably assume the answer is not in the same form as the question you’re asking. So if you ask a question, why can’t search engines take it and devise a search target to match an answer? It’s the opposite of, but no different from, what IBM’s Watson did on Jeopardy. Instead of helping us frame the question, search engines tend to search for the words just as we entered them. But there’s a better way to work towards finding your answer.
2. Where to Search
Wouldn’t it be nice if we could search the Internet the way we found that jar opener in the last post? That is, search by narrowing the field until the right answer appears. We could enter just one or two words, look at the results, then keep adding words, narrowing the results until we hit the target. It’s the Hotter-and-Colder game all kids play. In computer terms, it’s called an iterative search. These are fast because each successive step only looks at the last set of results instead of the entire Internet.
Why isn’t searching nice and easy? Why do search engines search the way they do? Lots of reasons. One reason is that they see searching as a computer problem. To them it’s not a human problem. Another is that their goals are not necessary the same as yours or mine.
Search engines may be free to use, as are browsers, but that doesn’t mean they don’t care about making money. In fact, I can make the case they care more about making money than meeting our search needs.
Next week: What search engines actually do.