If it was really semantic content, then your client (browser) could walk the graph of related (advertised) documents from those links and provide all sorts of information. For the advertising to be semantic, it would need to be wrapped in some sort of standard API or descriptive (semantic) access method that flagged it as advertising. You could then, in a good client, turn off all the advertising links, and even substitute dictionary entries with the same keyword.
Semantic access is exactly that; providing the data for the client to make decisions based on, so that you can choose between different things that have the same keywords, depending on their meaning. If it isn't associated with a new browser feature, it probably isn't a semantic document at all, unless it is just a REST-based catalog that is easily client-walkable. Then it might be primitive "semantic web."