Web searchers these days are a sophisticated bunch. We expect more from our search results, and sometimes a list of links just doesn’t cut it. Plus, who wants to muddle around those results trying to find precisely what you’re looking for? Shouldn’t a search engine know what you want? That’s why search engines, including heavy hitters such as Google and Bing, are beginning to look for ways to get you the information you want more quickly.
The latest attempt to make search results more relevant is by peering into the meaningof your search query itself. This is called semantic search.
What is semantic search?
Semantics is the study of the meaning and relation of words together. When applied to search, it allows a search engine to return results to a query based on what it believes the searcher is intending to find. For example take a search for “Philadelphia.” While the standard search may return the city’s official website, its tourist bureau, and other information, a semantic search goes further.
These results are a little more abstract: for example the city’s location, its population, the climate, and other facts about the city. In some cases, search engines are already providing semantic results. A Google search for “population of Philadelphia” displays the result — 1,528,306 — along with a graph of historical population data. Searching for “xanax” brings up a short description of the drug aprazolam and its uses.
Search companies in essence are hoping that they can answer your question without ever leaving the site. This strategy may not be beneficial to those sites who depend on Google and others to drive traffic to their own sites, but it certainly means more page views for the search engines and more chances to sell advertising.
Who’s using semantic search?
Arguably the core concept of semantic search itself and the desire on the part of consumers for such functionality could be traced back to Ask Jeeves. When the search engine debuted in 1996, it brought with it a whole new way of searching the web. Instead trying to figure out the right search terms to use, users were told to phrase their searches in the form of a question. While it was not truly semantic search, this changed the way people interacted with search engines.
True semantic search did not appear on a larger scale until 2005 when Google introduced Q&A. From the search terms entered into the search box, Google attempted to anticipate what the user was looking for across a variety of subjects, such as celebrities, movies, and the elements. WhenMicrosoft debuted Bing in 2009, the company leaned heavily on semantics to differentiate itself from competitors. The company argued that searchers are looking for more than just a list of links.
Wolfram Alpha takes Bing’s concept even further. Instead of combining traditional with semantic search as Google and Bing have done, Wolfram Alpha is completely semantic. Entering queries into the search box returns results from preexisting sets of data. Like Ask, you enter your query in the form of a question but Wolfram Alpha attempts to answer the question directly.
How will it change traditional search?
As long as search has existed, web denizens have played the game known as “search engine optimization.” The goal of SEO is to design websites in such a way that it can be easily parsed by search engine web crawlers, and make it so that the content of that site is optimized to appeal to the engines. Emphasis is usually placed on writing text in such a way that keywords associated with possible search queries appear prominently and repeatedly.
Semantic search changes the way SEO works. Website managers can no longer pack sites with keywords in the hope that Google’s algorithms will look favorably upon it for higher ranking. Additional emphasis is placed in the content surrounding those keywords. Does the text play well with those keywords, anticipating the intent of person searching? In other words, the game will be over — ranking high on a search engine that properly uses semantics will not be easy and will require a good deal more work.
Websites will also be competing with the search engines themselves. Part of the goal of incorporating semantic results into traditional ones is answering the searcher’s question before needing to go elsewhere for it. This means more page views for the search engine, and less for those who may have relied on Google, Bing, and others for a good deal of referral traffic. SEO takes a backseat as the content itself becomes much more important.
For users, semantic search changes the way we interact with search. Instead of being a stop along the way of looking for what we might need, the search engine becomes a destination. Why search around the web when the answer is right there? Search companies like this for the reasons I just mentioned, and users will appreciate it because it drastically simplifies the search process.
Is semantic search the “holy grail?”
Google and Bing (and Wolfram Alpha, too) are embracing semantic search in a big way, but there are limits. While it is nice to enter a question into a search engine such as “What’s the weather going to be in New York City tomorrow?” and get an answer right away, not all searches need to be so specific. Sometimes there’s a need for more vague research when you’re searching, and the standard list of links works well.
Another issue is accuracy. Search engines will need to stay on top of the data they lean on to respond to semantic queries. What if the data is wrong? They search engines won’t just be linking to the data, they’ll also be presenting it. Sometimes it’s good to have that list of links, because one site may be telling you one thing, and the rest something completely different. What if these sites use Wikipedia? Many of those articles aren’t always 100% right. Semantic results will require a great deal more oversight on the part of the search engine itself to ensure accuracy.
Then there’s the issue of playing favorites. With traditional search, getting listed costs you nothing other than time and effort. With semantic search, that result has to come from somewhere. Semantic search could arguably become just another type of sponsored result, a very effective and prominent method of advertising for those who can afford it. Google already sells those top spots to advertisers. Why not make those results semantically aware. That weather could be “powered by The Weather Channel” for example.
It might not lead to a click per se, but that visual reinforcement goes a long way the next time somebody’s looking for weather information. Who might they think of? Hopefully The Weather Channel first.
Semantic search shows a lot of promise to change the way we search. For the webmaster, it changes the game of getting your site high up in search results. For the user, it will hopefully make our searches more relevant as it will attempt to guess our intent rather than a literal interpretation of every search term we type in. Will it also change the search giants’ stance against pay-for-play when it comes to search results? That remains to be seen, but the groundwork has certainly been laid.