Top Semantic Search Engines
Accurate Results based on Language Rules
What is a semantic search engine? Generally speaking, it delivers results based on the science of meaning in language. Originally, many search engines tried to use phrase matching to deliver results, but were supplanted by Google's use of links and other factors to deliver accurate search results to the user. Unfortunately, it is still possible to get vage (or ambiguous) results for certain terms. A semantic engine uses rules of disambiguation to either find out the context of your search, or present you with different options. For instance, a search on "rock" might get you results for music, for stones of the non rolling variety, and for an insurance company logo. A semantic search engine might ask you what you are looking for, or show results around the context of other words in your search query, or present the most common items based on past experience in searching.
Some of the top known semantic engines in 2009 are:
Wolfram Alpha - A newly launched search engine.
Hakia - An engine you can ask quesitons, which lets you limit queries to "credible sites" as well as news and images.
Microsoft "Kumo" - A yet to be released replacement for Live Search
Powerset - Which processes Wikipedia results and can answer questions based on the "facts" therein.
Google Wonder Wheel - Which (at this date) doesn't even give its own result when you search on it!
So far, a general Google, Yahoo, or even MSN search still brings up pretty good results compared to these engines. Wolfram Alpha isn't really a search engine because it does not bring up any other web pages, and it is a bit clunky. These engines may be missing the point for the average surfer who is likely looking for things outside the field of "knowledge." If you ever want to see what PG rated things people are searching for, check out Google Hot Trends and after you weep for the future of humanity, you might note that there aren't too many semantic-friendly queries in the list. Semantic engines are usually popular among academics, and we aren't talking about those people that got a BA degree in distributed studies. If you want to solve a linear equation, or get the mass of the planet so you can calculate orbital velocities, or see your age in minutes, then semantic engines are the way to go. Those of us who got English degrees and had to read Hard Times by Charles Dickens and totally hated it can still remember 20 years later that there was an educator in the book called Mr. Gradgrind, and he only cared about teaching facts. So far, Google is not going anywhere, because it can deliver choices (many of which include the requisite Wikipedia listing) and people still want that.
One last dirty secret for fans of semantic search engines: Searchers aren't that smart. As someone with a background in search, I can tell you that the average query on a search engine is for a website name. Why? Because people know the name of the website, but don't know how to type it into the URL bar. This is probably why Yahoo prefers not to publish an unfiltered list of search queries, because people will go to Yahoo to search for Google. Therefore, when people use a semantic engine and don't get the listing for the site they already typed in, they will not use the engine again. QED. If you are interested in collecting payments online for search engine work, capital processing network may be one option for collecting revenue.
Notes and Special Information
Special note: Search engines update themselves all the time, so the information here should be obsolete about 10 seconds after it is published, and this is a beta site, so if I get any criticism I will just change the content and say you were lying, and that you must be able to manipulate cached data when you show me a screenshot of that. Also, this interface does not have spell check, so if you can't find 5 typos you probably aren't trying hard enough.