Potentials of enriching the Web of Documents
with Linked Data by generating RDFa markup
Benjamin Adrian
1
Knowledge Management Department, DFKI
Kaiserslautern, Germany
2
CS Department, University of Kaiserslautern
Kaiserslautern, Germany
benjamin.adrian@dfki.de
Abstract. Linked Data is out there. It consists of data about various
topics in a range from human health care to Pop Music or product in-
formation. While on a web search users like Sarah, Pete, and Tom would
like to see this data while reading and browsing the web of documents.
The position of this paper figures out the potentials of enriching web
pages with linked data content according to the specific information de-
mand the user has in his current situation.
Tools that automatically enrich web pages with RDFa content from
Linked Data knowledge bases are considered to be the next step to really
use the web of data while browsing the web of documents.
1 Motivation
Please consider Sarah, a PHD student. Sarah wants to buy a laptop and has a
budget of 600$. Sarah has a tight schedule, so her plan is to search online to
quickly compare different offers regarding to product properties and existing re-
views from customers. Fortunately, Sarah knows Linked Data and she was happy
to see the Openlink Virtuoso Sponger that would give her access to Amazon.com
data in RDF format. She knows Amazon to have a large knowledge base about
products and vendors, their offers, and user generated ratings and experience
reports about using these products. Sarah likes Amazon and its marketplace,
but wants to give all online vendors a chance. Finally she wants the cheapest
offer for the best laptop she can get with 600$. Therefore she wants a projection
of Amazon’s product data to products mentioned in web pages of online shops
she found while searching Google products or Yahoo’s Search Monkey.
Imagine Pete, a high school student. Pete is reading an exciting thriller on
his iPad. Pete loves to have a more colorful imagination about concrete sets and
locations where actions take place in. Therefore he installed a fancy App that
uses data from DBpedia and LinkedGeoData to produce a mashup on Google
map and Streetview consisting of scenery pictures, satellite images, and links
to additional background information about heritages, famous buildings, battle-
grounds, etc.
Tom is a young entrepreneur. He really knows about the power of social plat-
forms like Twitter or Facebook. Tom knows many experts, friends and customers
inside these platforms and he likes to know about their opinion about some new
technologies and products Tom is reading about in blogs or other web pages.
Tom installed a browser plugin that uses the RDF data published by Twitter
and Facebook to get the latest tweets, comments and blog entries from twitter
and Facebook from his contacts about topics mentioned in these web pages.
2 Position
The tools Sarah, Pete, and Tom use have one in common: they use RDF data
published as Linked Data to enrich existing instance mentions (textual phrases
that refer to existing instances in the RDF data) with RDFa markup. This
markup explicates the reference between a phrase in text and the instance within
a data set. The tools of Sarah, Pete, and Tom depend on different Linked Data
sets but enrich the document content with additional information that:
– helps Sarah to face more product offers and reviews about a certain Laptop she
found during her web search,
– helps Pete in stimulating his imagination while looking at real pictures and maps
about the primeval forests near the small town Folks in Washington, USA which is
the set of the latest book by Stephanie Meyer about vampires, Native Americans,
and werewolves, Pete is currently reading.
– helps Tom in getting the opinions and comments of well regarded experts from his
Twitter and Facebook account about topics he is reading about.
Behind the scenes the tools start with analyzing text passages in original web
pages . . .
From Port Angeles I carried on towards Forks on highway 101.
. . . then use Linked Data to explicate mentions in text with RDFa markup . . .
From Port Angeles I carried on to-
wards Forks on highway 101.
. . . which is then used by applications to request more information . . .
dbpedia:Folks\%2C\_Washington rdfs:label "Folks" ;
geo:lat "47.950980"^xsd:double ;
geo:long "-124.384749"^xsd:double ;
foaf:depiction
rdfs:seeAlso .
. . . which is finally visualized in useful, inspiring, and interesting mashups.
Conclusion: We recommend to build more browser or proxy based RDFa gen-
erators that automatically enrich web pages with Linked Data that helps the
user in his current situation. Tools such as Epiphany http://projects.dfki.
uni-kl.de/epiphany might be a first step to a usable web of data.