People Search Digest

In this modern age, any luddite would be forgiven for thinking that tracking down information online is quick, easy, and always accurate, but those of us locked into a daily struggle with search engines know that nothing about the Internet is easy and, when it comes to search engine results, hardly ever accurate.

Most people with an Internet connection have had the disheartening experience of inputting a fairly simple search term into their search engine of choice and being rewarded with a list of totally irrelevant, completely useless results. Here is illustrated the great split between the way the human mind thinks and the way the Internet, as of now, has been coded to think. Whereas the human mind makes natural connections between words and phrases using cultural inference, the Internet only makes connections between disparate words and phrases if it has been manually instructed to by the use of tags, keywords, and hyperlinks. Think about it this way: Humans are like a thesaurus - endless catalogues of meaning assigned to one word; the Internet is like the dictionary - a list of strict definitions with no room for embellishment. We connote, it denotes. We traffic in essence; it is the world’s last literalist. We get the gist; it, well, can’t. So the semantic search has come to fill the gap.

But what is the semantic search? It’s one part linguistics, one part network theory, and one part Internet reconfiguration, but in short it’s a way of training the Internet to better reflect the processing power of the human brain. If at present search engines generate results based on the number of times a searched for word or phrase appears in a web document (or, in Google’s case, the number of links leading to a web document), the semantic search model will generate results by combing the web for meaning assigned to word or phrase, rather than just the word itself. Rather than viewing the Internet as a series of pages sewn together with hyperlinks, a semantic search catalogues the Internet as a number of resources connected by the implied meaning of a search phrase. Literally, a semantic search will get to the root of what you mean; it won’t be hung up on what you say. Michel Foucault would be stoked (by which I mean excited).

The implications of the semantic search engine are huge, should it come to fruition. All search technology is presently in its infancy, even though huge strides are being made daily. The beginning stages of a semantic search can be witnessed by running a routine Google search. When a user scrolls through the search results, Google offers links to a few related search terms that may narrow down the results based on what the user is looking for. However, Google as of yet cannot infer meaning from search phrases; it can only suggest a list of related search terms taken from searches that have already been run.

So why haven’t we got the super semantic search engine yet? We just don’t have the technology and Google will not risk its good reputation by fiddling with their already successful formula. Google also assumes, rightly or wrongly, that users are stupid and cannot be relied upon to even know what they’re looking for in the first place. It’s entirely possible that a new company will discover the technology to run a fully functioning semantic search engine and beat the God-like Google at their own game. Already companies such as Hakia are finding ways to move search engines away from the old model, though the process is slow. What is clear is that in ten years, we will most likely be looking back at the current era of search engines as primitive and lacking. That or we’ll all be slaves to the computer overlords who, having become more savvy at understanding humans than humans themselves, will take over the world with the power of language. Dude! Bummer.

By Mariana M.