A statement by statement 'diff' would be great to have because it is 'diffing' graphs (not text) and would be a key component in the process
I will have to 1) do alias detection first to resolve the URI's across the models, because the URI of the same resource is different in each model and there is variation in properties, 2) then re-construct the models with the resolved URI's, 3) then use your 'diff' to analyze the differences, 4) construct derivation tree/lattice. thanks Arthur On May 29, 2009, at 11:07 AM, Holger Knublauch wrote: > > It can compare two arbitrary files in your workspace, but it does this > triple by triple (with special handling of blank node structures). > This means that if you have changed a namespace in all triples of a > file, then the system will report the whole file as different. We can > add such features after we get some real-world feedback. > > The next release is planned for mid June, before SemTech. > > Holger > > > > On May 29, 2009, at 8:55 AM, Arthur wrote: > >> >> That is great news! Would I have to put both models in the same >> namespace to do the diff? >> Thanks >> Arthur >> >> On May 29, 10:44 am, Holger Knublauch <[email protected]> wrote: >>> The upcoming version of TBC-ME will have a diff tool that allows you >>> to compare arbitrary files. The output of this diff will be another >>> RDF model which can then be browsed in TBC or TBE or otherwise >>> further >>> processed with SPARQL and SPARQLMotion. >>> >>> This may not address all your needs out of the box, but the approach >>> is very extensible and would allow you to add your own diff rules >>> and >>> algorithms. >>> >>> Holger >>> >>> On May 29, 2009, at 8:41 AM, Arthur wrote: >>> >>> >>> >>> >>> >>>> There are two parts to this question 1) how best compute >>>> differences >>>> between models, and 2) how best to represent differences between >>>> models? >>> >>>> Context: >>>> I have a number of (instance) models, each sharing a common schema >>>> (import), each in its own namespace, each representing one of >>>> multiple >>>> simulation runs in a project analyzing a common situation. Each >>>> model represents a different experiment ( i.e., different >>>> parameterizations) of a simulation model with results. I am >>>> translating existing simulations (proprietary format) into OWL to >>>> do >>>> this, so this is post-hoc analysis (the end goal is to handle model >>>> variation when it happens in the application using owl, but I need >>>> to >>>> deal with legacy simulations as well). The difference between >>>> models is very small (a handful of statements), because the >>>> engineers >>>> are simply varying input parameters for risk analysis. This is >>>> like >>>> an SCM problem in the sense that I would like to create a model >>>> derivation tree, but different because it is post-hoc, and I need >>>> easy >>>> access to all the models for statistical aggregations (i.e., risk >>>> analysis), else I could use an SCM. The objective is is to to save >>>> space and gain referential integrity/normalization by for example, >>>> re- >>>> organizing the models into a derivation tree . >>> >>>> Possible Solutions: >>>> Are there tools within TBC that I could use to do this? I will >>>> need >>>> to identify the common substructure between the models, so that the >>>> largest common submodel is the root of the derivation tree, and so >>>> on. What is the best way to 'diff' them, or find what is common >>>> between them? I could use imports to structure the derivation >>>> tree >>>> or I could use SPARQL construct queries and Sparql Motion to >>>> represent >>>> the derivation tree more implicitly and derive the models on >>>> demand, >>>> or should I simply use SVN in TBC? >>> >>>> Any ideas would be most welcome >>>> Thanks >>>> Arthur >>> > > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "TopBraid Composer Users" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/topbraid-composer-users?hl=en -~----------~----~----~----~------~----~------~--~---
