Thanks for the pointer. I believe it would be good to link these ontologies with corresponding dbpedia pages - assuming wikipedia has descriptions of these algorithms. dbpedia is becoming a de-facto hub for all kinds of knowledge integration tasks.
Anyway, this discussion clearly shows that SPARQL needs a well-defined mechanism for defining extensions such as custom mathematical operators. The SPIN functions framework might be one possible starting point, and as I said earlier we are enhancing SPIN with more flexible executable languages such as JavaScript. The idea is that SPIN functions would provide metadata that point to an executable algorithm (e.g. by pointing to a .js file on the Web and giving a JavaScript function name). Then any SPARQL engine could look these functions up and run them. For the kind of complex mathematical operations that require a lot of number crunching, it may be necessary to take this a step further, and allow JavaScript code to access the triple store as well. We may feed our observations into the newly formed W3C SPARQL 2 working group. In the meantime, the usual Jena/TBC extension points can be used to create those functions in Java. Holger On Feb 3, 2009, at 6:48 AM, Dave butlerdi wrote: > There is an Openmath symbols ontology that may make it easier to > actually have the meaning in your ontology but with outside > processing. > > http://monet.nag.co.uk/monet/publicdocs/monet_ontologies.html > > On Tue, Feb 3, 2009 at 3:39 PM, AndrewB <[email protected]> wrote: > > I will look into semantic JSP thing, in the end it might be better > sticking to jenna/java because it would be easier to distribute. > > Thanks, > Andrew > > > In particular, Semantic JSP may be an option worth looking into. > This > > gives you server-side JSP processing with embedded SPARQL. This > means > > your stat package can programmed in Java and run on the server. > There > > is a JSP example in the Help pages. > > > > This also finesses the problem from one of ontology-editors-as-the- > > interface to applications using ontologies in the back-end. > > > > From that point there are many options using TBC-ME/Live. The > > TopBraid Suite videos (http://www.topquadrant.com/topbraid/composer/ > > videos.html) give a taste, with more coming in TBS 3.0. > > > > In the end, the stat functions do not have to be "in" the ontology. > > Personally, I would be wary of doing all of your processing from > > within a query. SPARQLMotion different, as data processing is its > > provenience, but overloading a query language (SPARQL or SQL) with a > > lot of processing seems like a mismatch that could lead to complex > > aggregates with performance issues. Again, it's all dependent on > what > > kinds of stats you want to produce. > > > > -- Scott > > > > On Feb 2, 8:15 pm, Holger Knublauch <[email protected]> wrote: > > > > > Hi Andrew, > > > > > On Feb 2, 2009, at 4:50 PM, AndrewB wrote: > > > > > > Hello Mr. Knublauch > > > > The company I work for owns a TBCME and a TBL license. We > will > > > > be using Jenna for much of our math but If I ask my higher ups > to > > > > submit something like a feature request do you think there > would be > > > > any way we could imbed statistial funcitons into the ontology/ > sparql, > > > > or possibly in sparql motion? > > > > > Sure. Paying customers get preferred treatment when it comes to > new > > > feature requests. What I would need though is a precise list of > > > additional functions that you would need. > > > > > > Also I came across some pdf on the web where a guy references a > > > > outside function and it somehow connects to a javascript file > on his > > > > web server and does the math there. Cool but ugly solution. > > > > > The next beta will include an extension of SPIN that allows > anyone to > > > define new SPARQL functions using JavaScript. However, for the > time > > > being this would not help you very much because the JavaScript > code > > > would not be allowed to query the triple store (Jena Graph) at > > > execution time. So for complex operations it would need to get all > > > values as arguments and then do the math on them. I would be > > > interested to hear why you would consider such an interpreted > (e.g. > > > JavaScript) solution ugly - it would make it possible for anyone > to > > > create and share SPARQL functions on the semantic web. > > > > > Holger > > > > > -- > Regards > > Dave Butler > butlerdi-at-pharm2phork-dot-org > > Also on Skype as butlerdi > > Get Skype here http://www.skype.com/download.html > > > ********************************************************************** > This email and any files transmitted with it are confidential and > intended solely for the use of the individual or entity to whom they > are addressed. If you have received this email in error please notify > the system manager. > > This footnote also confirms that this email message has been swept by > MIMEsweeper for the presence of computer viruses. > > www.mimesweeper.com > ********************************************************************** > > > --~--~---------~--~----~------------~-------~--~----~ 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 -~----------~----~----~----~------~----~------~--~---
