Linked Data HOWTO

Jump to bottom Edit New page

nakao edited this page Aug 25, 2011 · 7 revisions

This HOWTO summarises some of the steps recommended to "transform" your DB into a Linked Data resource.

Step 0

Do I really need to transform my data?
What do I and my colleagues gain?

Step 1

Understand what Linked Data really means
Keep in mind that large scale reasoning might be a issue
Accept that I need URI's to identify your pieces of data

Step 2

Extract your data:

from unstructured sources: e.g. NLP, text mining, annotation.
from semi-structured sources: e.g. DBpedia, LinkedGeoData, SCOVO/DataCube.
from structured sources: e.g. RDB2RDF

Step 3

Step 4

Step 5

Step 6

Issues

Questions

You, yes you!, are more than welcome to answer these questions:

What should I ....?
Why is the ...?
In the event of ...?
What's the rules of using perfixes? Is there any conventions?
What's the predicate for thumbnail image urls.
DBpedia uses the dbpedia-owl:thumbnail . example at http://dbpedia.org/page/Berlin

Links