The Semantic Web is specifically designed to provide contextual information to Search Engines, Adsense engines and the like, so that they can deliver better results. Taking the example of Paris Hilton, this is either a person with an aversion to clothes or a choice of four posh hotels. If we embed "FOAF" data alongside this, we're telling Google we're talking about the person. If we use geotagging, we're telling Google it's a place. With many names being common to people, places, products and/or brands, this information allows us to explain to Google what we ARE talking about and, perhaps more importantly, what we're NOT talking about.
This becomes all the more important if you're placing Adsense adverts on the page, as having either competitive or embarrasing adverts next to your content can be very destructive! It can also mean that because your advertising is far more targeted than your competitors, you can charge more for advertising because the click-through and conversion rates are better than your competitors!
However, to use Semantic Web technologies properly demands a completely different web strategy and it isn't cheap! The site needs to be built differently, we need to develop a rich vocabulary of synonyms and antonyms and build the site in a different way before it is populated with content. The content can also be different because we're less worried about "Google-baiting" within the text, but the text must still be natural as far as Google is concerned!
To be clear, Semantic Web has little directly to do with Social Media which, however you look at it, is a collection of disparate platforms on or through which we interact with people. However, the two working together are incredibly powerful, with Forum content being tagged using Semantic Information that means the content is clearly understood by Search Engines is very good news.
To use Social Media properly as a company, Brand or individual also demands a strategic approach and needs to be done properly and anyone who calls themselves a "Guru" should be instantly discounted - there are about 500 platforms available for people to interact via the web - nobody can possibly claim expertise in more than a handful of these!
To leverage the power available with all these different technologies demands a "practitioner". To use this in a medical analogy, you go to a GP - General Practitioner, they diagnose your complaint/problem and can either address it themselves, or will refer you to a specialist, who is a subject matter expert in this specific area. I don't think Social Media should be thought of any differently!
If you want to engage one or more target segments, you need to know where they "hang out" on the web - LinkedIn, Facebook, MySpace, Twitter, YouTube, etc and you then need someone who can plug you into that segment using appropriate media both for what you want to promote and what "they" want to consume..... and this needs to be seamlessly integrated with your website, etc. Where these two converge is where we're heading right now.
Back in November 2008, I wrote a blog called Maximising-the-value-of-the-web-for-you-and-your-business-–-part-1 which talked about this convergence and it included the following (about half way down)
A very bright lady called Sramana Mitra proposed the theory on her blog that
“Web 3.0 = (4C + P + VS) + Place”
In other words, it’s about Content, Commerce, Community & Context (4C) + Personalisation (P) + Vertical Search (VS)
The combinations of the Semantic Web to give you Content and Context with Social Media to give you Community and Personalisation plus the eCommerce bit and Bob's your Dad's brother!
The key here is about how you search - let's assume you're in a strange town and hungry:
You can either do a Google search for "Restaurants Guildford" or you can do a Google search for "Restaurants" on a device that knows where you are [Place] and gives you tailored answers based on location or you could go to a recommendation engine that allows you to choose from a list of restaurants that have been given ratings by other users.
There's one more option which is emerging now, which is that the recommendations are further biased on a "people like me" basis, so that the answers are the ones that people most similar to you rated most highly, but that demands information about personal preferences and historical choices - information that would come from interaction on relevant Social Media platforms which feed into the recommendation engine's database.
Hopefully this helps clarify the difference between the Semantic Web and Social Media - it's obviously not exhaustive, nor is it intended to be!
If you need more information, just drop me an email.