On Tue, Dec 12, 2017 at 12:56 AM, 'Nil Geisweiller' via opencog < [email protected]> wrote:
> Hi, > > On 12/11/2017 09:44 PM, Mindey I. wrote: > >> Hello everyone, >> >> is there a simple way to download OpenCog KR model in a machine-readable >> format, so I could try representing knowledge, learning relations and >> reason with OpenCog links and models without installing the software per se? >> > > I'm afraid not, certainly not easily. What format would you have in mind? > Prolog? KIF? > Well, technically, opencog does not have any KR model. It is a language for describing KR models. In some vague philosophical sense, it maybe could be able to describe itself, but there's just a huge amount of stuff buried in C++ code that this would never be workable in practice. I don't think we could ever quite capture what the pattern matcher, pattern miner, rule engine, or PLN actually do, simply by writing some atomese (or kif or prolog or ...) that describe it. You would need to have a C++-to-kif, or C++-to-prolog translator to capture the intricacies. Opencog has something like half a dozen different KR models in it: -- that used by PLN and R2L mostly consisting of ConceptNodes, PredicateNodes, InheritanceLinks and MemberLinks -- that used the language subsystem, consisting of WordNodes, and the like -- that used by the language learning subsystem, using SectionLinks, etc. -- that used by ghost, using.. whatever ghost uses. -- misc space and time KR subsystems -- an old obsolete KR system used by the old/obsolete question-answering subsystem -- and old obsolete system used for the minecraft virtual world > Beside the logic you'd also have to represent truth values (second order > probabilistic distributions representing uncertainty, though in practice, > for now we only use SimpleTruthValue that can be represented by 2 numbers > https://wiki.opencog.org/w/TruthValue#SimpleTruthValue). > > Sure, everything is possible but it looks like you'd have to write the > translation tool yourself. > > Nil > > >> For example, I'd like to get a list of all "Links", with machine-readable >> mathematical, linguistic, and ML model-based definitions, and reproducible >> code to re-learn them, or re-implement them. >> >> My interest is in integration of all learning and reasoning systems, and >> I'd like to simply collect them into a machine-readable collection of >> models and processes, with more accessible human interface for reproducible >> engineering and science. (Wikipedia and arXiv are not good enough...) >> >> Best regards, >> Mindey >> >> >> >> >> -- >> You received this message because you are subscribed to the Google Groups >> "opencog" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to [email protected] <mailto: >> [email protected]>. >> To post to this group, send email to [email protected] <mailto: >> [email protected]>. >> Visit this group at https://groups.google.com/group/opencog. >> To view this discussion on the web visit https://groups.google.com/d/ms >> gid/opencog/62287087-a54c-47fa-98a4-4cc68be85bf4%40googlegroups.com < >> https://groups.google.com/d/msgid/opencog/62287087-a54c-47f >> a-98a4-4cc68be85bf4%40googlegroups.com?utm_medium=email&utm_source=footer >> >. >> For more options, visit https://groups.google.com/d/optout. >> > > -- > You received this message because you are subscribed to the Google Groups > "opencog" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To post to this group, send email to [email protected]. > Visit this group at https://groups.google.com/group/opencog. > To view this discussion on the web visit https://groups.google.com/d/ms > gid/opencog/58ac80b7-cb99-f30e-fd5e-6b13169b1389%40gmail.com. > > For more options, visit https://groups.google.com/d/optout. > -- *"The problem is not that artificial intelligence will get too smart and take over the world," computer scientist Pedro Domingos writes, "the problem is that it's too stupid and already has." * -- You received this message because you are subscribed to the Google Groups "opencog" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/opencog. To view this discussion on the web visit https://groups.google.com/d/msgid/opencog/CAHrUA36E_mKYJKO2284QUt099z2%3DuAVPBQ1pMVeRUWPeW4kEaw%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
