About Being Semantic

PriceWaterhouseCoopers has just published one of their periodic Technical Reports where they discuss and forecast the impact of the semantic web on business operations. This report is readable - although the concepts are complex, and there are several examples that elucidate the practical implications of adopting some of the theory (BBC).

Fundamental to the development of the semantic web will be the Resource Description Framework (RDF) which builds on the experience of xml and improves on the benefits provided by relational data. Central to RDF are Universal Resource Identifiers (URI) which are 'supersets' of URLs - your garden variety web address. In a semantic context, URIs are more specific than URLs and on page 7 of the report, the authors explain how this interplay works.

There are three key themes noted in this report:

1. Establishment & Use of Ontologies and Taxonomies. Increasingly anyone in data and information management is going to understand and recognise what these terms mean. In the publishing business, far more attention is now being paid to the the way data can/should relate to other data (Ontological) as well as how data can be structured hierarchically (taxonomy). As hierarchical classification schemes, taxonomies can/are limiting - although this factor may not be apparent until an organization considers utilizing one advantage of ontologies which is that they can be linked to other ontologies.
   
   
2. The ability to link your data with data from any third party will ultimately create better business information and support both more accurate and faster decision making. The Linked Data Initiative is a standards based approach to try to make this happen. In a linked data environment data can be shared and the interrelationships between data understood within the confines of a shared understanding of a common ontology. In the report, PWC examine how establishing a new retail location using complex data from a variety of sources can be aggregated and made actionable all using a linked data framework. As the report says, linked data is about both 'supply and demand': Accessing the data you want but also allowing others to access your data as well.
   
   
3. The report also addresses SPARQL and SQL. Theres much more in the report but SPARQL supports interpretation of graphical data representing an improvement over SQL. This discussion challenges the upper reaches of my pay grade so you will have to read this yourself.
   

I would have liked to quote directly from the report but ironically the pdf is disabled for copy and paste. However the report is free and runs about 50pages and I won't spoil the ending for you.

Also of relevance are two presentations I made at the Frankfurt supply chain meeting. Digital Age and Intelligent Publishing Supply Chain.