This is a sample sentence. it is about someone called Finn Aamodt mentioned on the OilIT site. In this case we didn't link to them or a page about them, but still there is hidden markup identifying them. Here is another mention, but this time of a company: BP.

That's great and all. You can find out which pages mention some company or person. But it is pretty basic and not very Webby.

Let's try adding a bit more hypertext...

This is another paragraph about that same company. BP Global's Web site holds information about its board. If this was in RDFa, we could automatically link it with related information about other people and companies.

For example, we might note that Tony Hayward works at BP. For this we used an ID from the OilIT site. Other pages are also available about Tony; for example he is the primary topic of a Wikipedia page. Consequently there is also a DBPedia identifier for him, as well as for BP.

With potentially multiple identifiers for everything, and the need to distinguish pages from their topics, this can all get a bit confusing. One way to stay sane is to run queries against the parsed output of these pages. So for example, here is the data so far in this sample document.

<http://www.oilit.com/2journal/2index/2peo/21.htm#itself> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://xmlns.com/foaf/0.1/Person> .
<http://www.oilit.com/2journal/2index/2peo/21.htm#itself> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.oilit.com/vocab/basic#OilIndustryPerson> .
<http://www.oilit.com/2journal/2index/2peo/21.htm#itself> <http://xmlns.com/foaf/0.1/name> "Finn Aamodt"@en .
<http://www.oilit.com/2journal/2index/2com/2342.htm#itself> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://xmlns.com/foaf/0.1/Organization> .
<http://www.oilit.com/2journal/2index/2com/2342.htm#itself> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.oilit.com/vocab/basic#OilIndustryOrganization> .
<http://www.oilit.com/2journal/2index/2com/2342.htm#itself> <http://xmlns.com/foaf/0.1/name> "BP"@en .
<http://www.oilit.com/2journal/2index/2com/2342.htm#itself> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://xmlns.com/foaf/0.1/Organization> .
<http://www.oilit.com/2journal/2index/2com/2342.htm#itself> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.oilit.com/vocab/basic#OilIndustryOrganization> .
<http://www.oilit.com/2journal/2index/2com/2342.htm#itself> <http://xmlns.com/foaf/0.1/homepage> <http://www.bp.com/> .
<http://www.oilit.com/2journal/2index/2com/2342.htm#itself> <http://xmlns.com/foaf/0.1/isPrimaryTopicOf> <http://www.bp.com/> .
<http://www.oilit.com/2journal/2index/2peo/21551.htm#itself> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://xmlns.com/foaf/0.1/Person> .
<http://www.oilit.com/2journal/2index/2peo/21551.htm#itself> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://www.oilit.com/vocab/basic#OilIndustryPerson> .
<http://www.bp.com/sectiongenericarticle.do?categoryId=9021976&contentId=7038002> <http://xmlns.com/foaf/0.1/primaryTopic> <http://www.oilit.com/2journal/2index/2peo/21551.htm#itself> .
<http://www.oilit.com/2journal/2index/2peo/21551.htm#itself> <http://xmlns.com/foaf/0.1/name> "Tony Hayward"@en .
<http://www.oilit.com/2journal/2index/2peo/21551.htm#itself> <http://xmlns.com/foaf/0.1/workplaceHomepage> <http://www.bp.com/> .
<http://en.wikipedia.org/wiki/Tony_Hayward> <http://xmlns.com/foaf/0.1/primaryTopic> <http://www.oilit.com/2journal/2index/2peo/21551.htm#itself> .
<http://www.oilit.com/2journal/2index/2peo/21551.htm#itself> <http://www.w3.org/2002/07/owl#sameAs> <http://dbpedia.org/resource/Tony_Hayward> .

And here is how to query to find out what other articles there might be on this person:

PREFIX target: <http://www.oilit.com/2journal/2index/2peo/21551.htm#itself> .
PREFIX dc: <http://purl.org/dc/elements/1.1/> .
PREFIX foaf: <http://xmlns.com/foaf/0.1/> .
SELECT DISTINCT ?what ?title WHERE {
 ?article foaf:topic <http://www.oilit.com/2journal/2index/2peo/21551.htm#itself> . 
 ?article foaf:topic ?topic2 . 
 ?topic2 foaf:name ?what . 
 ?article dc:title ?title . 
} 

Of course we could also ask questions based on other properties, such as having a 'workplace' of BP. But to do this for real would require a larger and more coherent aggregation of information than we have on hand right now.