I want to talk about a topic that I find very exciting. The roots of this technology are in the area of Artificial Intelligence. This is not just for the tech savvy out there. This will have an impact on how everyone uses the web, over time. The idea is to make the web smarter, or should I say, to make the software and computers that run the web or run on the web smarter.
We all use the web and computers to connect with others, to find information, to share information and so on. If we just focus on the web, we can begin by saying that before the semantic web and related technologies, computers had no idea as to the meaning of what was on a web page. Search engines did not speak our language, despite the fact that it might seem otherwise. For a long time people have been putting into Google questions like we might ask another person. It probably seemed that Google understood our language. However, for the most part, even up until recently, Google has only been looking at what we type into the search box and extracting keywords or key phrases. Then it looks for web sites or web pages that contain those keywords or key phrases. It should be noted that Google is starting to integrate some of these semantic web or semantic search features at the time of this writing.
One problem has been that the web pages did not have anything on them that would help the search engine or any other program running on the web, to understand the meaning of what exactly is on the web page. Previously, I wrote an article on my blog entitled “Semantics and Syntax,” which can be read here. I spoke about the notion that in any language, and also when computers communicate, there is a syntax ( in English this would be the grammar and such) and the semantics, the meaning of what we want to communicate. The Semantic Web is offering a whole range of tools for embedding meaning into web pages and then creating tools to use that information in new ways… to make the web smarter, as it were.
Let’s take an example of a search I did recently. I had introduced a software application called Digsby, which integrates the features of numerous other chat programs, online email addresses and social networking services. She had remarked that she knew of a place named “Digsby” and she was curious to remember where it is located. I went to the search engine Google and to Yahoo and put in terms like “Digsby and geographic location,” among other keyword search terms and various groupings of the terms. I even tried to tell the search engines to not include web pages about a software application. That would be done something like this on Google: digsby -”software application” which you will find doesn’t work too well (the dash – means not). The first result is for the main page of the software application, which I said I specifically did not want. Even this does not work better in Google: digsby geographic location -”software application
The same search at Hakia.com of digsby geographic location will produce a page that has at the top a Spelling suggestion: Digby geographic location which when you click on it does find information about a place named Digby in Nova Scotia, Canada. I had mispelled my location but at least Hakia understood that I meant to find out about a place, a geographic location named Digby. It understood something of the “semantics,” in other words, of what I was seeking, the meaning.
This doesn’t happen through magic. Web developers will have to start putting semantic markup and utilizing semantic web technologies to make this possible. It is an exciting field and an exciting direction for things. We are hearing about things like “natural language processing,” which involves actually understanding the syntax and semantics of our language so that for example a search engine could actually understand a question or request written in natural language.
