OWL has several levels of increasingly expressive but general inferences. Much of the domain could be represented in OWL (classes (i.e. sets), instances (i.e. set membership), relationships with domains and ranges, etc.), but there would still be a need for the domain-specific inferences described in the original post.
The other concern about applying OWL when it comes to security and privacy is that OWL is based on Open World reasoning and does not follow the Unique Name Assumption. OWL would say, "Unless these two things are explicitly designated as being different, assume they *might* be the same thing." You have to go a bit further to close off these open questions in OWL. A set of inferences for this domain might be better based on Closed World reasoning and Unique Names (e.g. the proposed "OWL ICV" http://stardog.com/docs/sdp/icv-specification.html ). Using core.logic, you can implement a domain-specific graph logic, which is a well-worn technique of knowledge representation. The idea is to: 1. model your domain as a labeled graph 2. represent a graph in core.logic as triples: ?subject ?property ?object 3. implement a simple graph query "DSL" in core.logic (for example here is one for SWI Prolog: http://www.swi-prolog.org/pldoc/doc_for?object=section(3,'3.1',swi('/doc/packages/semweb.html')) ) 4. implement your domain-specific inferences using the above general graph queries. As Phillip mentioned, depending on the size of your graphs, you may run out of steam using core.logic, but then you can switch over to more capable graph-specific implementations that maintain indexes, etc. on the graph. -pdl On Sunday, May 26, 2013 2:43:47 AM UTC-7, Phillip Lord wrote: > > > Not a direct answer to your question, but you might want to take a look > at: > > https://github.com/phillord/tawny-owl > > This tool wraps OWL, which is a standard language for knowledge > representation, that maps to a formal logic underneath, which is machine > interpretable and reasonable. > > I don't know enough about cryptography to know whether the semantics of > OWL would be rich enough, but you should be able to specific that a > thing that knows a public key of someone is also a subclass of a thing > that knows the encrypted nonce. > > There are some limitations of OWL -- it is not full first order logic, > but in return, it has better computational properties that logic. The > reasoners are also highly developed and will scale to 10's or 100's of > thousands of statements. > > Phil > > > > > Adam Saleh <adamth...@gmail.com <javascript:>> writes: > > I am trying to implement a simple knowledge derivation program in > > clojure.core.logic, > > and I am not sure how to approach it. To be more specific, after I am > done, > > it will be used as a toy cryptographic protocol verifier, > > that tries to infer what knowledge can attacker get based on some > initial > > knowledge and some rules. > > > > For example, I might denote [:pk A] a knowledge of public key of > somebody > > and [:pcrypted [:pk A] :nonce] an encrypted nonce sent to this person. > > > > Then I'd like be able to somehow specify, that if I know [:pk A], I know > as > > well [:pcrypted [:pk A] [:nonce A]] and therefore my program would > derive > > from initial knowledge > > > > [ [:pk :a] [:pk :b] ] the full knowledge [[:pk :a] [:pk :b] [:pcrypted > > [:pk :a] [:nonce :a]] [:pcrypted [:pk :b] [:nonce :b]]] > > > > Or in slightly more difficult case to specify that if I know a secret > key > > [:sk A] and encrypted nonce [:pcrypted [:pk A] [:nonce A]], I also know > the > > [:nonce a]. > > > > Anybody has any hints? > > > > So far I managed to implement some of this, but it is a tangled mess of > > recursive pattern matching, membero, nonmembero and macros :-/ > > > > Thanks for any pointers, > > > > Adam > > > > -- > > -- > Phillip Lord, Phone: +44 (0) 191 222 7827 > Lecturer in Bioinformatics, Email: > philli...@newcastle.ac.uk<javascript:> > School of Computing Science, > http://homepages.cs.ncl.ac.uk/phillip.lord > Room 914 Claremont Tower, skype: russet_apples > Newcastle University, twitter: phillord > NE1 7RU > -- -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
