Retrieve and manipulate data in semantic web technology

As we know the semantic web is exponentially growing to give the meaning to the information and data which machines can understand it.and there is huge amount of information and data stored in internet. My problem is if we have a huge amount of data provided by semantic web standards how we can get the exactly information we want. for instance if we assume that we have a web site like "wikipedia" which is completely understandable by machines.how it's possible to know cities in Austria that start with word "ober". obviously it not possible to go though all of the cities.