Recently I have been reading and rereading the book "Semantic Web for the Working Ontologist: Effective Modeling in RDFS and OWL".
The need for this book is very high since studying the RDF, RDFS and OWL specs is not enough.
It doesn't stop at knowing their syntax and semantics.
The main aim is to develop good models, but how is this done. Where are the (anti-) patterns?
Seen the huge fame of both authors, my expectations were very high,
Unfortunately the more I read it, the more disappointed I get.
However, during the first read, the book was a real eye opener, in this sense that the authors make very clear that modeling for the semantic web is different from the more traditional ER or UML modeling.
The main reason for modeling for the SW is to allow inferencing of new data. Lesson learned.
But there it stopped for me.
The book is full of typos and errors (where was the copy editor?).
Most of all suffer the illustrations, which is really a pity since, aimed at better understanding, they are now confusing the reader.
One example: figure 9.2 contains 7 errors.
The authors jump also too quickly into examples without explaining upfront the heuristics to tackle the challenge.
When in the middle of reading the examples I felt too frequently lost: why are we doing this, where does this lead to?
Be also aware that one of the examples of an RDFS model presented in the book is the 2005 version of SKOS (Simple Knowledge Organization System) and that some of the issues discussed are not longer applicable.
E.g. the properties 'skos:broader' and 'skos:narrower' are not transitive anymore; 'skos:Collection' is not the be used anymore as range of a 'skos:semanticRelation'; 'skos:subjectIndicator(s)' are left out. Since these decisions were known early 2008, they could have made the book.
Conclusion: Hasty work - shame for the craftsman.
Who takes the opportunity and writes the reference book on modeling for the SW?
This book surely isn't yet in this first print.
to correct all errors are available online now.