Re: Owning URIs (Was: Yet Another LOD cloud browser)
Kingsley Idehen wrote: David Huynh wrote: Sherman Monroe wrote: To be more specific, these days a news reporter can say foobar.com http://foobar.com on TV and expect that to mean something to most of the audience. That's a marvel. Something more than just the string foobar.com http://foobar.com is transfered. It's the expectation that if anyone in the audience were to type foobar.com http://foobar.com into any web browser, then they would be seeing information served by the authority associated with some topic or entity called foobar as socially defined. And 99% of the audience would be seeing the same information. What's the equivalent or analogous of that on the SW? I just want to make sure the analogies are aligned properly and are salient. The WWW contains only nouns (no sentences). If I have an interest or service I want to share with others, then I post a webpage and /share its URL/ with you. In the SW, things are centered around the crowd, if I have something to say about the an interest, service, place, person, etc, then I /reference its URL/ in my statements. So the SW contains sentences that can be browsed. Type the URL in the WWW browser, you get /the thing /being shared. Type the URI in the SW browser, you get the /things people say about the thing/. I didn't quite express myself clearly. If you were to take the previous sentence (I didn't quite express myself clearly), and encode it in RDF, what would you get? It certainly is something that I said about the thing, the thing being vaguely what I tried to explain before (how do you mint a URI for that?). The point is that using RDF or whatever other non-natural language structured data representation, you cannot practically represent the things people say about the thing in the majority of real-life cases. You can only express a very tiny subset of what can be said in natural language. This affects how people conceptualize and use this medium. If I hear a URI on TV, would I be motivated enough to type it into some browser when what I get back looks like an engineering spec sheet, but worse--with different rows from different sources, forcing me to derive the big picture myself, urn:sdajfdadjfai324829083742983:sherman_monroe name: Sherman Monroe (according to foo.com) age: __ (according to bar.com) age: ___ (according to bar2.com) nationality: __ (according to baz.com) ... rather than, say, a natural language essay that conveys a coherent opinion, or a funny video? David David, When you see a URI (a URL is a URI to me) on the TV, or hear one mentioned on the TV or Radio, you now have the option to interact with a variety of representations associated with the aforementioned Thing identified by the URI. You have representational choices that didn't exist until now. Choice is inherently optional :-) Beware the paradox of choices :-) http://www.amazon.com/Paradox-Choice-Why-More-Less/dp/0060005696/ref=sr_1_2?ie=UTF8qid=1242800143sr=8-2 A URI by definition cannot presuppose representation. This is the heart of the matter. The Semantic Web Project isn't about a new Web distinct from the ubiquitous World Wide Web. I think that sentiment and thinking faded a long time ago. If you are used to seeing a nice looking HTML based Web Page when you place URIs in a browser or click on them, then there's nothing wrong with that, always interact with a Web resource using the representation that best suits the kind of interaction at hand. Thus, someone else may want to know what data was contextualized by the nice looking HTML representation (the data behind and around the page), and on that basis seek a different representation via the same URI that unveils the kind descriptive granularity delivered by an Entity-Attribute-Value graph (e.g., RDF). The revolution is about choice via negotiated representations in a manner that's unobtrusive to the Web in its current form. Nobody has to change how they use the Web, we are just adding options to an evolving medium. You've forced my hand, I need to make a movie once and for all :-) It's not forcing, just nudging :) It'll be a win for all. David
Re: RDF: a suitable NLP KB representation (Was: Owning URIs (Was: Yet Another LOD cloud browser))
On 20/5/09 07:44, David Huynh wrote: Sherman Monroe wrote: That's when I was turned on to Frame Semantics, which I immediately praised, it is by far the most expressive and elegant knowledge representation framework for NL I have come across (although, it's been 3 or 4 years since I really looked). In short, frame semantics sees all sentences as a scene (like a movie scene) and the nouns all play roles in that scene. E.g. a boy eating is involved in a ConsumeFood scene, and the actors are the boy, the utensil he uses, the food, the chair he sits in. So I choose framesemantics as the KB model for Cypher grammar parser output. Thanks, Sherman, for your story. I had a history with Semantic Web technologies, too, since 2001. Data on the Web is inevitable. I just want to figure out ahead of time what it will actually be like. This sent off lightbulbs for me, I went back to RDF, and saw that, low and behold, frames can be represented as RDF, the scene types being classes, a scene instance (i.e. the thing representing a complete sentence) being the subject, the property is the role, and the object is the thing playing that role, e.g: EatFrame023 rdf:type mlo:EatFrame EatFrame023 mlo:eater someschema:URIForJohn EatFrame023 utensil someschema:JohnFavoriteSpoon EatFrame023 mlo:seatedAt _:anonChair EatFrame023 foaf:location someschema:JohnsLivingRoom EatFrame023 someschema:time _:01122 EatFrame023 truthval false^booleanValueType dbpedia:Heroes(Series) rdf:type dbpedia:TVShow dbpedia:Heroes(Series) dbpedia:showtime _:01122 _:01122 rdf:type types:TimeSpan _:01122 types:startHour 20^num:PositiveInteger _:01122 types:startMinutes 00^num:PositiveInteger _:01122 types:endHour 21^num:PositiveInteger _:01122 types:endMinutes 00^num:PositiveInteger _:01122 types:timezone EST This says: /No, John didn't eat in a sandwich in a chair in his living room using his favorite spoon, during the TV show Heroes/. Do you still believe RDF is incapable of expressing complex NL statements? Yes, I still believe. :) Skeptical? Me too, here. You have to be pretty careful with negations expressed over representations that are shipped around in RDF triples. EatFrame023 rdf:type mlo:EatFrame EatFrame023 mlo:eater someschema:URIForJohn EatFrame023 utensil someschema:JohnFavoriteSpoon EatFrame023 mlo:seatedAt _:anonChair EatFrame023 foaf:location someschema:JohnsLivingRoom EatFrame023 someschema:time _:01122 EatFrame023 truthval false^booleanValueType What happens when you add properties to EatFrame023, or when you remove properties from the description? If the above is true, can I take it, omit EatFrame023 utensil someschema:JohnFavoriteSpoon and pass it along to some downstream system in a context that suggests it remains a true description? Detail aside: yes, event-centric descriptions are pretty seductive. In a world where everything is constantly changing, at least a true description of an event seems something that is timelessly true. But they can be super-slippery to compute with, especially in a world of partial, fragmented descriptions. They're also hard to manage w.r.t. to identification: given two descriptions of an event or frame, how do you know whether they refer to different ones? etc. My dabbling with event modelling came through http://www.ilrt.bris.ac.uk/discovery/harmony/docs/abc/abc_draft.html where we tried to explore the resolution of some tensions between Dublin Core and other metadata vocabularies by articulating everything in terms of events. It is very appealing, but ultimately I think the problem with describing everything in terms of a giant what happened? log is that it doesn't directly tell you what the state of the world is at any point. And that's what many information consumers (human or mechanical) more often need. /What happens if all the worlds databases (e.g. Oracle, Mysql, etc databases out there) could be directly connected to one another in a large global network, all sharing one massive, distributed schema, and people were able to send queries to that network using a Esperanto for SQL?/ The ability of RDF to represent (not sentences but) rows and columns of any database schema imaginable means it can deliver this vision, and the value tied to it. And look what happened to Esperanto... After one century, 2 million speakers, or 0.025% of the world population. [...] Media are notoriously hard to understand, from what I can understand. If we were to say that television was radio but just with images, then we would be missing something huge. Or that printing was writing but just much faster. Or that writing was speech just recorded on paper. Or that SPARQL is just Esperanto for computers? The metaphor is nice but ... just that. As you point out, even non-metaphorical conceptual shortcuts can mislead us. Sometimes metaphors can inspire us and show a glimpse of the bigger picture. But I don't think they so often help us predict what'll actually
Re: ANN: GoodRelations - E-Commerce on the Web of Data - New Datasets and Applications
Hi Libby, That's rather fabulous! Can you give some information about how often this dataset is updated, and what's its geographical and product type reach? Thanks! This particular data set is a rather static collection and has a bias towards US products. It will soon be complemented by a more dynamic and European-centric second data set. In the long run, we will have to convince professional providers of commodity master data (e.g. GS1) to release their data following our structure. Currently, this is not possible due to licensing restrictions (there are look-up services like GEPIR, but none of them allows redistribution of the data). The upcoming second data set will be based on a community process, i.e., shop owners enter labels for EAN/UPCs in a Wiki. Since EAN/UPCs must (theoretically) not be reused, the current data set should be pretty reliable, though not necessarily very complete. I see the main benefit of the current data set in - using it as a showcase how small businesses can fetch product master data from the Semantic Web and - showing how data on the same commodity from multiple sources can be easily linked on the basis of having the same http://purl.org/goodrelations/v1.html#hasEAN_UCC-13 property value. Individual commodity descriptions can be retrieved as follows: http://openean.kaufkauf.net/id/EanUpc_UPC/EAN Example: http://openean.kaufkauf.net/id/EanUpc_0001067792600 This seems to give me multiple product descriptions - am I misunderstanding? The whole data set is divided in currently 100 (will be changed to 1000 soon) RDF files, which are being served via a bit complicated .htaccess configuration. The reason is that the large number of instance data would otherwise require 1 million very small files (a few triples each), which may cause problems with several file systems. Also, since we want as much of our data as possible to stay within OWL DL (I know not everybody in the community shares that), this would cause a lot of redundancy due to ontology imports / header data in each single file. But as far as I can see, the current approach should not have major side effects - you get back additional triples, but the size of the files being served is limited. Currently, we serve 4 MB file chunks. We will shortly reduce that to 400 - 800 KB. That seems reasonable to me. Best Martin Libby begin:vcard fn:Martin Hepp n:Hepp;Martin org:Bundeswehr University Munich;E-Business and Web Science Research Group adr:;;Werner-Heisenberg-Web 39;Neubiberg;;D-85577;Germany email;internet:mh...@computer.org tel;work:+49 89 6004 4217 tel;pager:skype: mfhepp url:http://www.heppnetz.de version:2.1 end:vcard
Re: Owning URIs (Was: Yet Another LOD cloud browser)
David Huynh wrote: Kingsley Idehen wrote: David Huynh wrote: Sherman Monroe wrote: To be more specific, these days a news reporter can say foobar.com http://foobar.com on TV and expect that to mean something to most of the audience. That's a marvel. Something more than just the string foobar.com http://foobar.com is transfered. It's the expectation that if anyone in the audience were to type foobar.com http://foobar.com into any web browser, then they would be seeing information served by the authority associated with some topic or entity called foobar as socially defined. And 99% of the audience would be seeing the same information. What's the equivalent or analogous of that on the SW? I just want to make sure the analogies are aligned properly and are salient. The WWW contains only nouns (no sentences). If I have an interest or service I want to share with others, then I post a webpage and /share its URL/ with you. In the SW, things are centered around the crowd, if I have something to say about the an interest, service, place, person, etc, then I /reference its URL/ in my statements. So the SW contains sentences that can be browsed. Type the URL in the WWW browser, you get /the thing /being shared. Type the URI in the SW browser, you get the /things people say about the thing/. I didn't quite express myself clearly. If you were to take the previous sentence (I didn't quite express myself clearly), and encode it in RDF, what would you get? It certainly is something that I said about the thing, the thing being vaguely what I tried to explain before (how do you mint a URI for that?). The point is that using RDF or whatever other non-natural language structured data representation, you cannot practically represent the things people say about the thing in the majority of real-life cases. You can only express a very tiny subset of what can be said in natural language. This affects how people conceptualize and use this medium. If I hear a URI on TV, would I be motivated enough to type it into some browser when what I get back looks like an engineering spec sheet, but worse--with different rows from different sources, forcing me to derive the big picture myself, urn:sdajfdadjfai324829083742983:sherman_monroe name: Sherman Monroe (according to foo.com) age: __ (according to bar.com) age: ___ (according to bar2.com) nationality: __ (according to baz.com) ... rather than, say, a natural language essay that conveys a coherent opinion, or a funny video? David David, When you see a URI (a URL is a URI to me) on the TV, or hear one mentioned on the TV or Radio, you now have the option to interact with a variety of representations associated with the aforementioned Thing identified by the URI. You have representational choices that didn't exist until now. Choice is inherently optional :-) Beware the paradox of choices :-) http://www.amazon.com/Paradox-Choice-Why-More-Less/dp/0060005696/ref=sr_1_2?ie=UTF8qid=1242800143sr=8-2 A URI by definition cannot presuppose representation. This is the heart of the matter. The Semantic Web Project isn't about a new Web distinct from the ubiquitous World Wide Web. I think that sentiment and thinking faded a long time ago. If you are used to seeing a nice looking HTML based Web Page when you place URIs in a browser or click on them, then there's nothing wrong with that, always interact with a Web resource using the representation that best suits the kind of interaction at hand. Thus, someone else may want to know what data was contextualized by the nice looking HTML representation (the data behind and around the page), and on that basis seek a different representation via the same URI that unveils the kind descriptive granularity delivered by an Entity-Attribute-Value graph (e.g., RDF). The revolution is about choice via negotiated representations in a manner that's unobtrusive to the Web in its current form. Nobody has to change how they use the Web, we are just adding options to an evolving medium. You've forced my hand, I need to make a movie once and for all :-) It's not forcing, just nudging :) It'll be a win for all. David David, Okay, so you've successfully nudged me :-) Here is the first cut (others will follow as this was done in haste, but demonstrates the essence of the matter). 1. YouTube -- http://www.youtube.com/watch?v=CweYtyw7fnY 2. Vimeo -- http://vimeo.com/4736569 -- Regards, Kingsley Idehen Weblog: http://www.openlinksw.com/blog/~kidehen President CEO OpenLink Software Web: http://www.openlinksw.com
Re: DBpedia user, who are you?
Georgi Kobilarov wrote: Hi all, Dear folks, I'm Davide Palmisano, an Asemantics[1] senior researcher and I'm very pleased to reply to Georgi's questions. I'm currently doing some planning for the future roadmap of DBpedia, and therefore gathering requirements and use cases. So I'm wondering: - Who is using DBpedia today or has evaluated it in the past, Currently we are using DBPedia within two disjoint main scenarios. the first one is related to the EU project called NoTube[2] where we are planning to use DBpedia as a main knowledge core to build semantic web based user profiles in order to make personalized TV content recommendation. This is a research project mainly aimed to produce innovative algorithm for the content discovery. The second one, partly covered by an NDA so I cannot be more precise, is an ambitious project that we will present to the next SemWeb09 called 99ways[3] where we are planning to make an intensive use of DBpedia. For example, we are currently making an autocompletion service that taking as input a substring it returns a list of DBpedia URIs grouped by their most representative skos:subject. The way we are calculating the most representative skos:subject for each URI is the key point within the overall algorithm. - What are you doing with it or how would you like to use it, oops, as the precedent one :) - How would you like to see it evolve? Grow, grow and grow! Jokes apart, the first real and important evolution that comes up in my mind is partially related to the uptime and to the scalability of the system. Improving the scalability of the SPARQL end point backend would be the key task to allow the resolution of very frequent and complex SPARQL queries. Especially interested in usage of DBpedia (and Linked Data) within organizations or even commercial scenarios. Please let me know, either on-list of off-list (and state in case you don't want that information to be disclosed). Thanks, Georgi all the best, Davide -- Georgi Kobilarov Freie Universität Berlin www.georgikobilarov.com -- Davide Palmisano Head of Research and Development Asemantics Srl - www.asemantics.com Circonvallazione Trionfale 27 00195 ROMA Italy skype id: davidepalmisano mobile: +393396101142
Re: DBpedia user, who are you?
On Wed, 2009-05-20 at 13:04 +0200, Georgi Kobilarov wrote: - What are you doing with it or how would you like to use it, Linked railway data project - right now, I'm just querying dbpedia to find an alternative URI for each station, so that I can provide ovterms:similarTo links from one URI for the station to another. But in the future I'd hope to also pull data about each station from dbpedia. libre.fm - not using dbpedia yet, but it's being considered as a place to find album art, artist biographies, etc. - How would you like to see it evolve? Batch queries would be nice. Imagine that I want to find a particular piece of information about multiple resources - e.g. the number of platforms a train station has, for each station in a list I've got of 1000. Right now I have two options, a single SPARQL query like: SELECT ?s ?o WHERE { ?s dbpedia:platforms ?o } which might return hundreds of results for subjects I'm not interested in; or querying each subject individually. The former is not the best use of bandwidth, especially for properties like rdfs:label where there are likely to be far more irrelevant results than relevant; and the latter can't be especially efficient in terms of dbpedia's CPU time. The other thing I'd like to see is some sort of better reasoning about redirects. I don't know quite how, but there must be something better than what we have now. -- Toby Inkster t...@g5n.co.uk
Re: DBpedia user, who are you?
We use dbpedia as part of the linked data world when computing networks and service details of things that we know have dbpedia entries; we also use the sameAs information. Eg For example see the ³Description² bit of http://www.rkbexplorer.com/detail/?uri=http://southampton.rkbexplorer.com/id /person-02686 or http://www.rkbexplorer.com/explorer/#display=person-{http://southampton.rkbe xplorer.com/id/person-02686} So, as with the nature of this world, it is not obvious at all that it is being used (unless you look at the raw data), because it is simply (like other sites) that there is a sameAs to a Uri there. Future? Just want to be able to continue using it? Best Hugh On 20/05/2009 12:04, Georgi Kobilarov georgi.kobila...@gmx.de wrote: Hi all, I'm currently doing some planning for the future roadmap of DBpedia, and therefore gathering requirements and use cases. So I'm wondering: - Who is using DBpedia today or has evaluated it in the past, - What are you doing with it or how would you like to use it, - How would you like to see it evolve? Especially interested in usage of DBpedia (and Linked Data) within organizations or even commercial scenarios. Please let me know, either on-list of off-list (and state in case you don't want that information to be disclosed). Thanks, Georgi -- Georgi Kobilarov Freie Universität Berlin www.georgikobilarov.com
Re: DBpedia user, who are you?
Hello! On Wed, May 20, 2009 at 12:04 PM, Georgi Kobilarov georgi.kobila...@gmx.de wrote: Hi all, I'm currently doing some planning for the future roadmap of DBpedia, and therefore gathering requirements and use cases. So I'm wondering: - Who is using DBpedia today or has evaluated it in the past, - What are you doing with it or how would you like to use it, - How would you like to see it evolve? Especially interested in usage of DBpedia (and Linked Data) within organizations or even commercial scenarios. Please let me know, either on-list of off-list (and state in case you don't want that information to be disclosed). Glad to contribute to that :-) We are using DBpedia in quite a lot of services at the BBC, as detailed in our ESWC paper [1]. I am also using it in almost all the services hosted at dbtune.org. Wrt. future plans, here are a couple of things that would be very great to have in future versions of dbpedia: 1) Query by example. You submit a bunch of DBpedia resources, and it returns a SPARQL query selecting them and resources with similar properties. 2) Live update from Wikipedia (but it seems quite close to being real, now) 3) An interface for submitting out-going links, instead of having to ping the dbpedia list each time Cheers, y [1] http://www.georgikobilarov.com/publications/2009/eswc2009-bbc-dbpedia.pdf Thanks, Georgi -- Georgi Kobilarov Freie Universität Berlin www.georgikobilarov.com
Re: ANN: GoodRelations - E-Commerce on the Web of Data - New Datasets and Applications
Hi Steve, as I replied to Libby (but did not include all mailing lists): The whole data set is served from currently 100 smaller files, which will be broken down to 1000 files shortly. For various reasons however, we don't want to serve one file per element, because that will create a huge overhead - the individual data sets are rather small (a few triples per item). Having one million micro-files is hard to manage. Also, since we want to stay within OWL DL, we would have to duplicate proper ontology header meta-data a million times. Thus, we use a (rather large) set of rules in the .htaccess file to serve that part of the data set that contains the element you are actually looking for. You will receive a few more triples than you need, but simply discard those ;-) Martin Steve Harris wrote: Very cool resource. On 20 May 2009, at 10:18, Libby Miller wrote: Individual commodity descriptions can be retrieved as follows: http://openean.kaufkauf.net/id/EanUpc_UPC/EAN Example: http://openean.kaufkauf.net/id/EanUpc_0001067792600 This seems to give me multiple product descriptions - am I misunderstanding? Yeah, looks like it returns the entire document that the particular EAN appears in. Not very linked data friendly (you'll end up with a large proportion of repeated triples in identical graphs, with different graph URIS), but certainly better than nothing. - Steve -- -- martin hepp e-business web science research group universitaet der bundeswehr muenchen e-mail: mh...@computer.org phone: +49-(0)89-6004-4217 fax:+49-(0)89-6004-4620 www:http://www.unibw.de/ebusiness/ (group) http://www.heppnetz.de/ (personal) skype: mfhepp Check out the GoodRelations vocabulary for E-Commerce on the Web of Data! Webcast explaining the Web of Data for E-Commerce: - http://www.heppnetz.de/projects/goodrelations/webcast/ Tool for registering your business: -- http://www.ebusiness-unibw.org/tools/goodrelations-annotator/ Overview article on Semantic Universe: - http://www.semanticuniverse.com/articles-semantic-web-based-e-commerce-webmasters-get-ready.html Project page and resources for developers: - http://purl.org/goodrelations/ Upcoming events: --- Full-day tutorial at ESWC 2009: The Web of Data for E-Commerce in One Day: A Hands-on Introduction to the GoodRelations Ontology, RDFa, and Yahoo! SearchMonkey http://www.eswc2009.org/program-menu/tutorials/70 Talk at the Semantic Technology Conference 2009: Semantic Web-based E-Commerce: The GoodRelations Ontology http://www.semantic-conference.com/session/1881/ begin:vcard fn:Martin Hepp n:Hepp;Martin org:Bundeswehr University Munich;E-Business and Web Science Research Group adr:;;Werner-Heisenberg-Web 39;Neubiberg;;D-85577;Germany email;internet:mh...@computer.org tel;work:+49 89 6004 4217 tel;pager:skype: mfhepp url:http://www.heppnetz.de version:2.1 end:vcard
Re: ANN: GoodRelations - E-Commerce on the Web of Data - New Datasets and Applications
Steve Harris wrote: Very cool resource. On 20 May 2009, at 10:18, Libby Miller wrote: Individual commodity descriptions can be retrieved as follows: http://openean.kaufkauf.net/id/EanUpc_UPC/EAN Example: http://openean.kaufkauf.net/id/EanUpc_0001067792600 This seems to give me multiple product descriptions - am I misunderstanding? Yeah, looks like it returns the entire document that the particular EAN appears in. Not very linked data friendly (you'll end up with a large proportion of repeated triples in identical graphs, with different graph URIS), but certainly better than nothing. thought entertained=minimal If the location header was set in the response I guess that might help. /thought Damian
Re: ANN: GoodRelations - E-Commerce on the Web of Data - New Datasets and Applications
Hello! Not very linked data friendly (you'll end up with a large proportion of repeated triples in identical graphs, with different graph URIS), but certainly better than nothing. Just jumping on that - is that an issue? I would think not, as you may want to repeat information across different views. For example (slightly biased :-) ), you may want to repeat broadcast information in a schedule view, instead of asking the user agent to manually get a hundred of URIs. Cheers, y
Re: ANN: GoodRelations - E-Commerce on the Web of Data - New Datasets and Applications
Alternatively you could put that data in a RDF store, and just serve up the fragments using a wrapped CONSTRUCT query. That's what we do for qdos.com, eg http://qdos.com/user/Steve-Harris/18b6f60b41e05aaa418565ebfe901d6b/rdfxml and it's pretty efficient, more efficient that storing 1000 separate files as XML. The downside is that the RDF is not very pretty to look at, but it could be with a better RDF/XML serialiser. - Steve On 20 May 2009, at 14:59, Martin Hepp (UniBW) wrote: Hi Steve, as I replied to Libby (but did not include all mailing lists): The whole data set is served from currently 100 smaller files, which will be broken down to 1000 files shortly. For various reasons however, we don't want to serve one file per element, because that will create a huge overhead - the individual data sets are rather small (a few triples per item). Having one million micro-files is hard to manage. Also, since we want to stay within OWL DL, we would have to duplicate proper ontology header meta-data a million times. Thus, we use a (rather large) set of rules in the .htaccess file to serve that part of the data set that contains the element you are actually looking for. You will receive a few more triples than you need, but simply discard those ;-) Martin Steve Harris wrote: Very cool resource. On 20 May 2009, at 10:18, Libby Miller wrote: Individual commodity descriptions can be retrieved as follows: http://openean.kaufkauf.net/id/EanUpc_UPC/EAN Example: http://openean.kaufkauf.net/id/EanUpc_0001067792600 This seems to give me multiple product descriptions - am I misunderstanding? Yeah, looks like it returns the entire document that the particular EAN appears in. Not very linked data friendly (you'll end up with a large proportion of repeated triples in identical graphs, with different graph URIS), but certainly better than nothing. - Steve -- -- martin hepp e-business web science research group universitaet der bundeswehr muenchen e-mail: mh...@computer.org phone: +49-(0)89-6004-4217 fax:+49-(0)89-6004-4620 www:http://www.unibw.de/ebusiness/ (group) http://www.heppnetz.de/ (personal) skype: mfhepp Check out the GoodRelations vocabulary for E-Commerce on the Web of Data! = = == Webcast explaining the Web of Data for E-Commerce: - http://www.heppnetz.de/projects/goodrelations/webcast/ Tool for registering your business: -- http://www.ebusiness-unibw.org/tools/goodrelations-annotator/ Overview article on Semantic Universe: - http://www.semanticuniverse.com/articles-semantic-web-based-e-commerce-webmasters-get-ready.html Project page and resources for developers: - http://purl.org/goodrelations/ Upcoming events: --- Full-day tutorial at ESWC 2009: The Web of Data for E-Commerce in One Day: A Hands-on Introduction to the GoodRelations Ontology, RDFa, and Yahoo! SearchMonkey http://www.eswc2009.org/program-menu/tutorials/70 Talk at the Semantic Technology Conference 2009: Semantic Web-based E-Commerce: The GoodRelations Ontology http://www.semantic-conference.com/session/1881/ martin_hepp.vcf -- Steve Harris Garlik Limited, 2 Sheen Road, Richmond, TW9 1AE, UK +44(0)20 8973 2465 http://www.garlik.com/ Registered in England and Wales 535 7233 VAT # 849 0517 11 Registered office: Thames House, Portsmouth Road, Esher, Surrey, KT10 9AD
Re: ANN: GoodRelations - E-Commerce on the Web of Data - New Datasets and Applications
On 20 May 2009, at 15:48, Yves Raimond wrote: Hello! Not very linked data friendly (you'll end up with a large proportion of repeated triples in identical graphs, with different graph URIS), but certainly better than nothing. Just jumping on that - is that an issue? I would think not, as you may want to repeat information across different views. For example (slightly biased :-) ), you may want to repeat broadcast information in a schedule view, instead of asking the user agent to manually get a hundred of URIs. It's not ideal if you're crawling that data - you'd end up with many thousands of copies of an (almost?) identical document. But with no obvious clue that they're the same. I didn't look at what was going on in the HTTP, but using the right 40x forwards it could make it clear to the client what's happening. - Steve -- Steve Harris Garlik Limited, 2 Sheen Road, Richmond, TW9 1AE, UK +44(0)20 8973 2465 http://www.garlik.com/ Registered in England and Wales 535 7233 VAT # 849 0517 11 Registered office: Thames House, Portsmouth Road, Esher, Surrey, KT10 9AD
Re: DBpedia user, who are you?
On Wed, 2009-05-20 at 14:57 +0100, Yves Raimond wrote: 3) An interface for submitting out-going links, instead of having to ping the dbpedia list each time Ooh!! This can be done?! Ping! http://ontologi.es/rail/links_dbpedia.ttl -- Toby Inkster t...@g5n.co.uk
Re: DBpedia user, who are you?
Nice work. However :-) It should be @prefix : http://ontologi.es/rail/stations/gb/ . not @prefix : http://ontologi.es/rail/station/gb/ . Cheers Hugh On 20/05/2009 16:15, Toby Inkster t...@g5n.co.uk wrote: On Wed, 2009-05-20 at 14:57 +0100, Yves Raimond wrote: 3) An interface for submitting out-going links, instead of having to ping the dbpedia list each time Ooh!! This can be done?! Ping! http://ontologi.es/rail/links_dbpedia.ttl -- Toby Inkster t...@g5n.co.uk
Re: [semanticweb] ANN: GoodRelations Service Update 2009-05-05 - Please refresh your caches!
Hi, Martin, I found your works on business ontologies as having some useful commercial prospects. I suggest to post your Good Relations ontology under the heading of Ontology Standards and Industry Standards, a special committee of the emerging non-profit international research organization: International Body for Ontology and Semantics Standards (IBOSS), http://www.standardontology.org (In the sage of devevlopment) Any researchers, developers, institutions, or legal entities in possession of high quality ontology content or semantic applications are also welcomed. Azamat Abdoullaev IBOSS Group - Original Message - From: Martin Hepp (UniBW) To: semantic-web at W3C ; semantic...@yahoogroups.com ; public-lod@w3.org Sent: Wednesday, May 20, 2009 11:36 AM Subject: [semanticweb] ANN: GoodRelations Service Update 2009-05-05 - Please refresh your caches! Dear all: We just released a service update of the GoodRelations ontology for e-commerce. The ontology is available at http://purl.org/goodrelations/v1 If you want to explicitly fetch the OWL file or the HTML documentation, you may also use http://purl.org/goodrelations/v1.owl or http://purl.org/goodrelations/v1.html Please replace all local copies of GoodRelations by this new file. The service update is designed to be fully backwards-compatible. Only a few changes may require small modifications of current applications or data sets. Those few potentially incompatible changes are as follows: - gr:description is now deprecated; we suggest using rdfs:comment instead. - gr:isListPrice is now deprecated and replaced by the more powerful gr:priceType property. The remaining changes are vocabulary extensions or improvements that help increase the compatibility with Semantic Web applications. For example, we - changed cardinality recommendations for opening hours and business functions, which simplifies the usage of GoodRelations in RDFa; - added PayPal as a payment method, and - added labels to all elements. A complete change log is at http://tinyurl.com/q8yln9 A big thank you for the many valuable suggestions for improvement! Best Martin -- martin hepp e-business web science research group universitaet der bundeswehr muenchen e-mail: mh...@computer.org phone: +49-(0)89-6004-4217 fax: +49-(0)89-6004-4620 www: http://www.unibw.de/ebusiness/ (group) http://www.heppnetz.de/ (personal) skype: mfhepp Check out the GoodRelations vocabulary for E-Commerce on the Web of Data! Webcast explaining the Web of Data for E-Commerce: - http://www.heppnetz.de/projects/goodrelations/webcast/ Tool for registering your business: -- http://www.ebusiness-unibw.org/tools/goodrelations-annotator/ Overview article on Semantic Universe: - http://www.semanticuniverse.com/articles-semantic-web-based-e-commerce-webmasters-get-ready.html Project page and resources for developers: - http://purl.org/goodrelations/ Upcoming events: --- Full-day tutorial at ESWC 2009: The Web of Data for E-Commerce in One Day: A Hands-on Introduction to the GoodRelations Ontology, RDFa, and Yahoo! SearchMonkey http://www.eswc2009.org/program-menu/tutorials/70 Talk at the Semantic Technology Conference 2009: Semantic Web-based E-Commerce: The GoodRelations Ontology http://www.semantic-conference.com/session/1881/ -- -- martin hepp e-business web science research group universitaet der bundeswehr muenchen e-mail: mh...@computer.org phone: +49-(0)89-6004-4217 fax: +49-(0)89-6004-4620 www: http://www.unibw.de/ebusiness/ (group) http://www.heppnetz.de/ (personal) skype: mfhepp Check out the GoodRelations vocabulary for E-Commerce on the Web of Data! Webcast explaining the Web of Data for E-Commerce: - http://www.heppnetz.de/projects/goodrelations/webcast/ Tool for registering your business: -- http://www.ebusiness-unibw.org/tools/goodrelations-annotator/ Overview article on Semantic Universe: - http://www.semanticuniverse.com/articles-semantic-web-based-e-commerce-webmasters-get-ready.html Project page and resources for developers: - http://purl.org/goodrelations/ Upcoming events: --- Full-day tutorial at ESWC 2009: The Web of Data for E-Commerce in One Day: A Hands-on Introduction to the
Re: ANN: GoodRelations - E-Commerce on the Web of Data - New Datasets and Applications
Steve Harris wrote: Alternatively you could put that data in a RDF store, and just serve up the fragments using a wrapped CONSTRUCT query. That's what we do for qdos.com, eg http://qdos.com/user/Steve-Harris/18b6f60b41e05aaa418565ebfe901d6b/rdfxml and it's pretty efficient, more efficient that storing 1000 separate files as XML. The downside is that the RDF is not very pretty to look at, but it could be with a better RDF/XML serialiser. - Steve Steve, The data is already in an RDF Store. Of course, you can add yours etc. :-) Martin should have sent links like: 1. http://tr.im/lThV -- the whole ontology 2. http://tr.im/lTiC -- sampling of Products or Service Model instance data Kingsley On 20 May 2009, at 14:59, Martin Hepp (UniBW) wrote: Hi Steve, as I replied to Libby (but did not include all mailing lists): The whole data set is served from currently 100 smaller files, which will be broken down to 1000 files shortly. For various reasons however, we don't want to serve one file per element, because that will create a huge overhead - the individual data sets are rather small (a few triples per item). Having one million micro-files is hard to manage. Also, since we want to stay within OWL DL, we would have to duplicate proper ontology header meta-data a million times. Thus, we use a (rather large) set of rules in the .htaccess file to serve that part of the data set that contains the element you are actually looking for. You will receive a few more triples than you need, but simply discard those ;-) Martin Steve Harris wrote: Very cool resource. On 20 May 2009, at 10:18, Libby Miller wrote: Individual commodity descriptions can be retrieved as follows: http://openean.kaufkauf.net/id/EanUpc_UPC/EAN Example: http://openean.kaufkauf.net/id/EanUpc_0001067792600 This seems to give me multiple product descriptions - am I misunderstanding? Yeah, looks like it returns the entire document that the particular EAN appears in. Not very linked data friendly (you'll end up with a large proportion of repeated triples in identical graphs, with different graph URIS), but certainly better than nothing. - Steve -- -- martin hepp e-business web science research group universitaet der bundeswehr muenchen e-mail: mh...@computer.org phone: +49-(0)89-6004-4217 fax:+49-(0)89-6004-4620 www:http://www.unibw.de/ebusiness/ (group) http://www.heppnetz.de/ (personal) skype: mfhepp Check out the GoodRelations vocabulary for E-Commerce on the Web of Data! Webcast explaining the Web of Data for E-Commerce: - http://www.heppnetz.de/projects/goodrelations/webcast/ Tool for registering your business: -- http://www.ebusiness-unibw.org/tools/goodrelations-annotator/ Overview article on Semantic Universe: - http://www.semanticuniverse.com/articles-semantic-web-based-e-commerce-webmasters-get-ready.html Project page and resources for developers: - http://purl.org/goodrelations/ Upcoming events: --- Full-day tutorial at ESWC 2009: The Web of Data for E-Commerce in One Day: A Hands-on Introduction to the GoodRelations Ontology, RDFa, and Yahoo! SearchMonkey http://www.eswc2009.org/program-menu/tutorials/70 Talk at the Semantic Technology Conference 2009: Semantic Web-based E-Commerce: The GoodRelations Ontology http://www.semantic-conference.com/session/1881/ martin_hepp.vcf -- Regards, Kingsley Idehen Weblog: http://www.openlinksw.com/blog/~kidehen President CEO OpenLink Software Web: http://www.openlinksw.com
Re: URI lifecycle (Was: Owning URIs)
David Booth wrote: Hi John, Re: The URI Lifecycle in Semantic Web Architecture: http://dbooth.org/2009/lifecycle/ On Tue, 2009-05-19 at 10:46 -0700, John Graybeal wrote: *Very* interesting paper, for the content and for the links. Addresses many a topic I've been trying to sort out. If I may ask for a clarification on a few key points at the beginning: 1) At what point does 'minting' occur? (a) When I think of the URI, (b) when I first write it down as a string in some file, (c) when I 'serve' it in some formal way, (d) when I make a statement that references it, or (e) ...? You define it as 'establishing the association between the URI and the resource it denotes', but how does the process of establishing that association occur, exactly? It all seems a little imprecise with respect to real-world resources. The simplest answer is that the URI is minted when the URI owner publishes its URI declaration, since it is the URI declaration that establishes the association between the URI and the resource it denotes. 2) Am I correct in thinking the URI owner is just the person who has the authority to create a URI (and optionally provide an initial set of statements about it)? In the SW, the idea of someone having the authority to link their URI to the actual resource -- Earth's moon for example -- is confusing, since many people will mint URIs meant to refer to the Earth's moon; I think they all have that authority, in some sense. (AWWW focused more on the actual URI and information resources, where there is an implicit association, often through deferencing.) In simple terms, the URI owner is the owner of the domain from which the URI is allocated, or the owner's delegate. For example, if John owns domain foo.example.com then John is the owner of all URIs allocated within that domain, such as http://foo.example.com/bar/whiz/bang . However, John could delegate minting authority to all or part of his URI space. For example, John could delegate minting authority for all URIs matching http://foo.example.com/lucinda/* to Lucinda. David, What about describing this in terms of: Data Space or URI Space ownership? You are describing functionality that should be integral to any Data Space or URI Space platform that plugs into the Internet ? Kingsley The notion of URI ownership is defined in the AWWW section 2.2.2.1: http://www.w3.org/TR/webarch/#uri-ownership 3) Can you define a core assertion? If I can improve my assertions to clarify that I meant the Earth moon we all know about, as opposed to some other 'Earth moon', is that not allowed per R1? How do we know when an improvement makes the original concept more useful, as opposed to erroneous for some users? (Note your suggestion later that it's OK when expectations are properly set, a la SKOS.) The core assertions are merely those that are provided in the URI declaration and serve to define the association between the URI and a resource. They do so by constraining the permissible interpretations for that URI. (An interpretation in RDF semantics lingo maps URIs to resources.) In the end the question of whether a change in a URI declaration will be helpful or harmful to your users is a judgement call. In theory, any change to the core assertions has the potential of invalidating some user's code. However, in practice some changes are far less likely to cause problems than others, because they don't affect the set of permissible interpretations -- at least not in a way that matters. For example, in the moon example at http://dbooth.org/2007/uri-decl/#example changing the rdfs:seeAlso assertion is unlikely to break users' code because it doesn't really constrain the resource identity of the URI http://dbooth.org/2007/moon/ . One can think of the core assertions as constraining the set of permissible interpretations for that URI. There will always be some ambiguity about what resource the URI denotes -- this is inescapable -- but the core assertions clearly delineate that ambiguity. This is further explained in a companion paper, Denotation as a Two-Step Mapping in Semantic Web Architecture: http://dbooth.org/2009/denotation/ The paper is a nice encapsulation of many of the idiosyncrasies of the current state of the social practice. Thanks You're welcome. And thanks very much for your comments! -- Regards, Kingsley Idehen Weblog: http://www.openlinksw.com/blog/~kidehen President CEO OpenLink Software Web: http://www.openlinksw.com
Re: DBpedia user, who are you?
Yves Raimond wrote: Hello! On Wed, May 20, 2009 at 12:04 PM, Georgi Kobilarov georgi.kobila...@gmx.de wrote: Hi all, I'm currently doing some planning for the future roadmap of DBpedia, and therefore gathering requirements and use cases. So I'm wondering: - Who is using DBpedia today or has evaluated it in the past, - What are you doing with it or how would you like to use it, - How would you like to see it evolve? Especially interested in usage of DBpedia (and Linked Data) within organizations or even commercial scenarios. Please let me know, either on-list of off-list (and state in case you don't want that information to be disclosed). Glad to contribute to that :-) We are using DBpedia in quite a lot of services at the BBC, as detailed in our ESWC paper [1]. I am also using it in almost all the services hosted at dbtune.org. Wrt. future plans, here are a couple of things that would be very great to have in future versions of dbpedia: 1) Query by example. You submit a bunch of DBpedia resources, and it returns a SPARQL query selecting them and resources with similar properties. 2) Live update from Wikipedia (but it seems quite close to being real, now) 3) An interface for submitting out-going links, instead of having to ping the dbpedia list each time Cheers, y [1] http://www.georgikobilarov.com/publications/2009/eswc2009-bbc-dbpedia.pdf Re. pinger services for SPARUL type effects, the availability of a FOAF+SSL based DBpedia SPARQL endpoint will make this feasible. And for those that don't have WebIDs (URIs), OAuth based SPARQL endpoint will do. Kingsley Thanks, Georgi -- Georgi Kobilarov Freie Universität Berlin www.georgikobilarov.com -- Regards, Kingsley Idehen Weblog: http://www.openlinksw.com/blog/~kidehen President CEO OpenLink Software Web: http://www.openlinksw.com
Re: ANN: GoodRelations - E-Commerce on the Web of Data - New Datasets and Applications
Steve Harris wrote: On 20 May 2009, at 16:38, Kingsley Idehen wrote: Steve Harris wrote: Alternatively you could put that data in a RDF store, and just serve up the fragments using a wrapped CONSTRUCT query. That's what we do for qdos.com, eg http://qdos.com/user/Steve-Harris/18b6f60b41e05aaa418565ebfe901d6b/rdfxml and it's pretty efficient, more efficient that storing 1000 separate files as XML. The downside is that the RDF is not very pretty to look at, but it could be with a better RDF/XML serialiser. - Steve Steve, The data is already in an RDF Store. Of course, you can add yours etc. :-) Martin should have sent links like: 1. http://tr.im/lThV -- the whole ontology 2. http://tr.im/lTiC -- sampling of Products or Service Model instance data Cool, how do I for eg. get the immediate data around http://openean.kaufkauf.net/id/EanUpc_0001067792600 out? It's not obvious how you request it. Do I just issue SPARQL to somewhere? The SPARQL link at the bottom goes to a page about SPARQL. I poked around for a bit and tried http://lod.openlinksw.com/sparql?query=DESCRIBE+%3Chttp%3A%2F%2Fopenean.kaufkauf.net%2Fid%2FEanUpc_0001067792600%3Eoutput=n3 but it gives an error, so I'm not sure what I did wrong. Steve, Still hot staging a few things. Nothing wrong with your command, we just need to complete some data re-organization work on this particular instance. Check back in a day or so :-) Also, we are adding this Linked Commerce Data to the LOD cloud, so do expect a published dump for the static data from this emerging space. Re. new data about new business entities, check PingTheSemantic Web . Kingsley - Steve On 20 May 2009, at 14:59, Martin Hepp (UniBW) wrote: Hi Steve, as I replied to Libby (but did not include all mailing lists): The whole data set is served from currently 100 smaller files, which will be broken down to 1000 files shortly. For various reasons however, we don't want to serve one file per element, because that will create a huge overhead - the individual data sets are rather small (a few triples per item). Having one million micro-files is hard to manage. Also, since we want to stay within OWL DL, we would have to duplicate proper ontology header meta-data a million times. Thus, we use a (rather large) set of rules in the .htaccess file to serve that part of the data set that contains the element you are actually looking for. You will receive a few more triples than you need, but simply discard those ;-) Martin Steve Harris wrote: Very cool resource. On 20 May 2009, at 10:18, Libby Miller wrote: Individual commodity descriptions can be retrieved as follows: http://openean.kaufkauf.net/id/EanUpc_UPC/EAN Example: http://openean.kaufkauf.net/id/EanUpc_0001067792600 This seems to give me multiple product descriptions - am I misunderstanding? Yeah, looks like it returns the entire document that the particular EAN appears in. Not very linked data friendly (you'll end up with a large proportion of repeated triples in identical graphs, with different graph URIS), but certainly better than nothing. - Steve -- -- martin hepp e-business web science research group universitaet der bundeswehr muenchen e-mail: mh...@computer.org phone: +49-(0)89-6004-4217 fax:+49-(0)89-6004-4620 www:http://www.unibw.de/ebusiness/ (group) http://www.heppnetz.de/ (personal) skype: mfhepp Check out the GoodRelations vocabulary for E-Commerce on the Web of Data! Webcast explaining the Web of Data for E-Commerce: - http://www.heppnetz.de/projects/goodrelations/webcast/ Tool for registering your business: -- http://www.ebusiness-unibw.org/tools/goodrelations-annotator/ Overview article on Semantic Universe: - http://www.semanticuniverse.com/articles-semantic-web-based-e-commerce-webmasters-get-ready.html Project page and resources for developers: - http://purl.org/goodrelations/ Upcoming events: --- Full-day tutorial at ESWC 2009: The Web of Data for E-Commerce in One Day: A Hands-on Introduction to the GoodRelations Ontology, RDFa, and Yahoo! SearchMonkey http://www.eswc2009.org/program-menu/tutorials/70 Talk at the Semantic Technology Conference 2009: Semantic Web-based E-Commerce: The GoodRelations Ontology http://www.semantic-conference.com/session/1881/ martin_hepp.vcf -- Regards, Kingsley Idehen Weblog: http://www.openlinksw.com/blog/~kidehen President CEO OpenLink Software Web: http://www.openlinksw.com -- Regards, Kingsley Idehen Weblog: http://www.openlinksw.com/blog/~kidehen President CEO OpenLink Software Web: http://www.openlinksw.com
VoCamp Sunnyvale 2009
(Apologies if you receive multiple copies of this message) === VoCamp Sunnyvale 2009 June 18-19, 2009, Sunnyvale, California http://vocamp.org/wiki/VoCampSunnyvale2009 === VoCamps are informal events for people interested in solving the practical problems of the Semantic Web in a social setting, with a particular focus on topics related to vocabularies and semantic interoperability in general. Following on the success of the previous four events, the next VoCamp will take place on the week of the Semantic Technologies conference (SemTech) at Yahoo in Sunnyvale, California. VoCamps are free for participants. Sign up by going to the website at vocamp.org. Attendance is limited to 30 people at the moment so be quick. Co-organizers: Peter Mika (Yahoo!) Melinda Chung (Yahoo!) Evan Goer (Yahoo!) Links: http://vocamp.org/wiki/VoCampSunnyvale2009 http://semtech2009.com/session/2134/
Re: DBpedia user, who are you?
At Turn2Live.com we will start (soon) to consume music data about artists from LOD and obviously from DBpedia. We are at a very initial phase. Hope to have demos soon! Juan Sequeda, Ph.D Student Dept. of Computer Sciences The University of Texas at Austin www.juansequeda.com www.semanticwebaustin.org On Wed, May 20, 2009 at 1:04 PM, Georgi Kobilarov georgi.kobila...@gmx.dewrote: Hi all, I'm currently doing some planning for the future roadmap of DBpedia, and therefore gathering requirements and use cases. So I'm wondering: - Who is using DBpedia today or has evaluated it in the past, - What are you doing with it or how would you like to use it, - How would you like to see it evolve? Especially interested in usage of DBpedia (and Linked Data) within organizations or even commercial scenarios. Please let me know, either on-list of off-list (and state in case you don't want that information to be disclosed). Thanks, Georgi -- Georgi Kobilarov Freie Universität Berlin www.georgikobilarov.com
Re: URI lifecycle (Was: Owning URIs)
Hi John, Re: The URI Lifecycle in Semantic Web Architecture: http://dbooth.org/2009/lifecycle/ On Tue, 2009-05-19 at 10:46 -0700, John Graybeal wrote: *Very* interesting paper, for the content and for the links. Addresses many a topic I've been trying to sort out. If I may ask for a clarification on a few key points at the beginning: 1) At what point does 'minting' occur? (a) When I think of the URI, (b) when I first write it down as a string in some file, (c) when I 'serve' it in some formal way, (d) when I make a statement that references it, or (e) ...? You define it as 'establishing the association between the URI and the resource it denotes', but how does the process of establishing that association occur, exactly? It all seems a little imprecise with respect to real-world resources. The simplest answer is that the URI is minted when the URI owner publishes its URI declaration, since it is the URI declaration that establishes the association between the URI and the resource it denotes. 2) Am I correct in thinking the URI owner is just the person who has the authority to create a URI (and optionally provide an initial set of statements about it)? In the SW, the idea of someone having the authority to link their URI to the actual resource -- Earth's moon for example -- is confusing, since many people will mint URIs meant to refer to the Earth's moon; I think they all have that authority, in some sense. (AWWW focused more on the actual URI and information resources, where there is an implicit association, often through deferencing.) In simple terms, the URI owner is the owner of the domain from which the URI is allocated, or the owner's delegate. For example, if John owns domain foo.example.com then John is the owner of all URIs allocated within that domain, such as http://foo.example.com/bar/whiz/bang . However, John could delegate minting authority to all or part of his URI space. For example, John could delegate minting authority for all URIs matching http://foo.example.com/lucinda/* to Lucinda. The notion of URI ownership is defined in the AWWW section 2.2.2.1: http://www.w3.org/TR/webarch/#uri-ownership 3) Can you define a core assertion? If I can improve my assertions to clarify that I meant the Earth moon we all know about, as opposed to some other 'Earth moon', is that not allowed per R1? How do we know when an improvement makes the original concept more useful, as opposed to erroneous for some users? (Note your suggestion later that it's OK when expectations are properly set, a la SKOS.) The core assertions are merely those that are provided in the URI declaration and serve to define the association between the URI and a resource. They do so by constraining the permissible interpretations for that URI. (An interpretation in RDF semantics lingo maps URIs to resources.) In the end the question of whether a change in a URI declaration will be helpful or harmful to your users is a judgement call. In theory, any change to the core assertions has the potential of invalidating some user's code. However, in practice some changes are far less likely to cause problems than others, because they don't affect the set of permissible interpretations -- at least not in a way that matters. For example, in the moon example at http://dbooth.org/2007/uri-decl/#example changing the rdfs:seeAlso assertion is unlikely to break users' code because it doesn't really constrain the resource identity of the URI http://dbooth.org/2007/moon/ . One can think of the core assertions as constraining the set of permissible interpretations for that URI. There will always be some ambiguity about what resource the URI denotes -- this is inescapable -- but the core assertions clearly delineate that ambiguity. This is further explained in a companion paper, Denotation as a Two-Step Mapping in Semantic Web Architecture: http://dbooth.org/2009/denotation/ The paper is a nice encapsulation of many of the idiosyncrasies of the current state of the social practice. Thanks You're welcome. And thanks very much for your comments! -- David Booth, Ph.D. Cleveland Clinic (contractor) Opinions expressed herein are those of the author and do not necessarily reflect those of Cleveland Clinic.
Re: Dereferencing a URI vs querying a SPARQL endpoint
I think sitemap may already have what you want. We use slicing=subject-object For our sets such as: (See http://acm.rkbexplorer.com/sitemap.xml). http://sw.deri.org/2007/07/sitemapextension/ Says: The sc:linkedDataPrefix and sc:sparqlEndpointLocation tags can have an optional slicing attribute that takes a value from the list of slicing methods below. slicing=subject The description of a resource X includes the triples whose subject is X. slicing=subject-object The description of a resource X includes the triples whose subject or object is X. slicing=cbd The description of a resource X includes its Concise Bounded Description. slicing=scbd The description of a resource X includes its Symmetric Concise Bounded Description. slicing=msgs The description of a resource X includes all the Minimal Self-Contained Graphs involving X. On 20/05/2009 17:59, Daniel Schwabe dschw...@inf.puc-rio.br wrote: Dear all, while designing Explorator [1], where one can explore one or more triple repositories that provide SPARQL enpoints (as well as direct URI dereferencing), I found the following question, to which I don't really know the answer... For the sake of this discussion, I'm considering only such sites, i.e., those that provide SPRQL enpoints. For a given URI r, is there any relation between the triples I get when I dereference it directly, as opposed to querying the SPARQL enpoint for all triples r, ?p, ?o ? Should there be (I could also get ?s, ?p, r, for example) ? For sites such as dbpedia I believe that I get the same set of triples. But I believe this is not a general behavior. Should there be a good practice about this for LoD sites that provide SPARQL endpoints? At the very least, perhaps this could also be described in the semantic sitemap.xml, no? Cheers D [1] http://www.tecweb.inf.puc-rio.br/explorator
Re: Dereferencing a URI vs querying a SPARQL endpoint
Sorry, I'll try harder :-) I understand that what you are asking is something like this. For some sites (including rkbexplorer), when you resolve a URI, it constructs a SPARQL query and returns the result of the query. This might be all the triples with the subject, or object, or both, or something more complex that takes into account b-nodes. So it might be nice if somewhere, such as the sitemap.xml, this query was documented. I think this is exactly what the slicing is trying to do, but instead of publishing the actual query, it names the common (obvious) ones to use. So a slicing of subject would tell you that you could do the query you say below on the appropriate SPARQL endpoint and get exactly the same thing you would get by resolving the URI. So I still think that answers your question, but I'm sure you can tell me if it doesn't :-) And others will say if I am wrong. Best Hugh On 20/05/2009 20:47, Daniel Schwabe dschw...@inf.puc-rio.br wrote: Dan and Hugh, let me be more specific. I'm not really advocating that only *one* direction should be returned (or even both directions). I am asking a more general question (to which I don't think Hugh really gave an answer either) which is, is there any query that returns the same triples as the ones you get when you dereference a URI, in a site that also provides a SPARQL endpoint? In the affirmative case, I am suggesting that the corresponding query be documented in the sitemap.xml document. Does this make sense? Cheers D On 20/05/2009 14:15, Dan Brickley wrote: On 20/5/09 18:59, Daniel Schwabe wrote: Dear all, while designing Explorator [1], where one can explore one or more triple repositories that provide SPARQL enpoints (as well as direct URI dereferencing), I found the following question, to which I don't really know the answer... For the sake of this discussion, I'm considering only such sites, i.e., those that provide SPRQL enpoints. For a given URI r, is there any relation between the triples I get when I dereference it directly, as opposed to querying the SPARQL enpoint for all triples r, ?p, ?o ? Should there be (I could also get ?s, ?p, r, for example) ? For sites such as dbpedia I believe that I get the same set of triples. But I believe this is not a general behavior. Should there be a good practice about this for LoD sites that provide SPARQL endpoints? At the very least, perhaps this could also be described in the semantic sitemap.xml, no? In general, I'd be wary of doing anything that assumes the direction a property is named in is important. Taking the old MCF example, http://www.w3.org/TR/NOTE-MCF-XML-970624/#sec2.1 the_songlines eg:author bruce_chatwin . where eg:author has a domain of Document and a range of Person. Exactly the same information could be conveyed in data where the property naming direction was reversed. And case by case, different natural languages and application environments will favour slightly one direction over the other. Here we could as well have had bruce_chatwin eg:wrote the_songlines . or eg:book or eg:pub or eg:xyz, with domain Person, range Document. As it happens in English, the word author doesn't have a natural and obvious inverse here but that's incidental. The point is that both forms tell you just as much about the person as about the document, regardless of property naming and direction. The form using eg:author seems to be document-centric, but in fact it should equally support UI layers that are concerned with the person or the document. It would be dissapointing if a UI that was presenting info about Bruce Chatwin was to miss out that he was the author of the_songlines, simply because somewhere along the line a schema writer chose to deploy a property author rather than wrote... cheers, Dan [1] http://www.tecweb.inf.puc-rio.br/explorator
Re: Dereferencing a URI vs querying a SPARQL endpoint
I would expect that a DESCRIBE query to the SPARQL endpoint return what I get when dereferencing the URI. pa Daniel Schwabe a écrit : Dan and Hugh, let me be more specific. I'm not really advocating that only *one* direction should be returned (or even both directions). I am asking a more general question (to which I don't think Hugh really gave an answer either) which is, is there any query that returns the same triples as the ones you get when you dereference a URI, in a site that also provides a SPARQL endpoint? In the affirmative case, I am suggesting that the corresponding query be documented in the sitemap.xml document. Does this make sense? Cheers D On 20/05/2009 14:15, Dan Brickley wrote: On 20/5/09 18:59, Daniel Schwabe wrote: Dear all, while designing Explorator [1], where one can explore one or more triple repositories that provide SPARQL enpoints (as well as direct URI dereferencing), I found the following question, to which I don't really know the answer... For the sake of this discussion, I'm considering only such sites, i.e., those that provide SPRQL enpoints. For a given URI r, is there any relation between the triples I get when I dereference it directly, as opposed to querying the SPARQL enpoint for all triples r, ?p, ?o ? Should there be (I could also get ?s, ?p, r, for example) ? For sites such as dbpedia I believe that I get the same set of triples. But I believe this is not a general behavior. Should there be a good practice about this for LoD sites that provide SPARQL endpoints? At the very least, perhaps this could also be described in the semantic sitemap.xml, no? In general, I'd be wary of doing anything that assumes the direction a property is named in is important. Taking the old MCF example, http://www.w3.org/TR/NOTE-MCF-XML-970624/#sec2.1 the_songlines eg:author bruce_chatwin . where eg:author has a domain of Document and a range of Person. Exactly the same information could be conveyed in data where the property naming direction was reversed. And case by case, different natural languages and application environments will favour slightly one direction over the other. Here we could as well have had bruce_chatwin eg:wrote the_songlines . or eg:book or eg:pub or eg:xyz, with domain Person, range Document. As it happens in English, the word author doesn't have a natural and obvious inverse here but that's incidental. The point is that both forms tell you just as much about the person as about the document, regardless of property naming and direction. The form using eg:author seems to be document-centric, but in fact it should equally support UI layers that are concerned with the person or the document. It would be dissapointing if a UI that was presenting info about Bruce Chatwin was to miss out that he was the author of the_songlines, simply because somewhere along the line a schema writer chose to deploy a property author rather than wrote... cheers, Dan [1] http://www.tecweb.inf.puc-rio.br/explorator
Re: Dereferencing a URI vs querying a SPARQL endpoint
Hi, If you have a dataset that is very large and highly interlinked on particular URI's, the DESCRIBE response may be too large to reasonably transmit to a user over the internet (and to expect a sparql endpoint to give out in one chunk). This is assuming the typical DESCRIBE behaviour that sparql vendors implement which picks out r ?p1 ?o (forward) and ?s ?p2 r (reverse) . If you know that you want both forward and reverse behaviour then to be you should probably utilise a SPARQL endpoint and page through the possible results with OFFSET and LIMIT until you don't get anymore results. In relation to the Bio2RDF results, the URI that you dereference with the federated queries is a mixture of what you could get at a particular set of endpoints, with some forward and some reverse relations, configured so that the system won't go down just from the weight of someone trying to effectively do DESCRIBE http://bio2rdf.org/taxon:9606. That would be linked to in a few hundred thousand places, but still only has a few forward construct triples that come out of the taxonomy database. In this case, the direction of the relationship is important in real world terms because it the size of the relationship. Insisting that whenever someone wants to get information about a taxonomy identifier (or some other classification method) that they have to also get everything else possibly related to it would cause a mountain of information. This is why [1] [2] [3] etc. are available for people wanting to get more related links. (although there may be slow endpoints that make each of those quite long operations) Admittedly, the results for resolving Bio2RDF URI's come from multiple endpoints, so if you just focused on a single Bio2RDF SPARQL endpoint you would get reasonable results from DESCRIBE most of the time. Cheers, Peter [1] http://qut.bio2rdf.org/pageoffset1/links/taxon:9606 [2] http://qut.bio2rdf.org/pageoffset2/links/taxon:9606 [3] http://qut.bio2rdf.org/pageoffset3/links/taxon:9606 2009/5/21 Pierre-Antoine Champin swlists-040...@champin.net: I would expect that a DESCRIBE query to the SPARQL endpoint return what I get when dereferencing the URI. pa
Re: URI lifecycle (Was: Owning URIs)
Hi David, On 20/05/2009 06:01, David Booth da...@dbooth.org wrote: A last comment, which I know we have discussed, and you possibly disagree: Community expropriation of a URI Might have meant something else. One of the problems is that many authors will not discharge their Statement Author Responsibilities, but will assume that the URI is the one they want. Over time, this may mean that the general SW uses a URI in a way other than the URI owner intends, to the extent that it becomes irrelevant what was the original meaning (there are many parallels for this in natural language, and indeed it is the social process that causes language to change). [ . . . ] Yes, that's a great topic for discussion. It is clear that semantic drift is a natural part of natural language: a word that meant one thing years ago may mean something quite different now. As humans we can usually deal with this semantic drift by knowing the context in which a word is used, though it can cause real life misunderstandings sometimes. However, I think our use of URIs in RDF is different from our use of words in natural language, in two important ways: - RDF is designed for machine processing -- not just human communication -- and machines are not so good at understanding context and resolving ambiguity; and - with URI declarations there is a simple, feasible, low-cost mechanism available that can be used to anchor the semantics of a URI. In short, although semantic web architecture could be designed to permit unrestricted semantic drift, I think it is a better design -- better serving the semantic web community as a whole -- to adopt an architecture that permits the semantics of each URI to be anchored, by use of a URI declaration. Absolutement. But your paper is not about architecture. The architecture, as you say, permits the semantics of each URI to be anchored. The (one of the?) good thing about your paper is that it is about the stuff that is not enforced by the architecture, but rather addresses what might be called the social processes and what responsibilities might be. And works hard to avoid confusion between them. So if one was to envisage ways in which the consequences of failure to adhere to the responsibilities might have a significant impact, and how that impact might be accommodated or challenged, then I think it can be useful to study it. I happen to think that people and hence agents will simply assume they know what URIs mean without checking the anchor, in the same way they use words without checking the dictionary. If I was marking this email up in RDFa, I would be much more likely to guess, or simply go and use the URIs you had used to mark up your email, rather than check each one back at base - I would never be able to do anything if I checked every word in the dictionary. In fact, how much of all the RDFa that is now being generated gets checked? I do take your point that a lot of this is happening with machines, but even they will make the same mistake when choosing a URI. Best Hugh For more explanation see: Why URI Declarations? A comparison of architectural approaches http://dbooth.org/2008/irsw/ -- David Booth, Ph.D. Cleveland Clinic (contractor) Opinions expressed herein are those of the author and do not necessarily reflect those of Cleveland Clinic.
Re: Dereferencing a URI vs querying a SPARQL endpoint
Pierre-Antoine Champin wrote: I would expect that a DESCRIBE query to the SPARQL endpoint return what I get when dereferencing the URI. pa Daniel, Is this your problem: Linked Data Servers publish URIs. The mechanism that delivers these URIs tends to vary since they are the product of URL-rewrite rules that may or may not be associated with SPARQL queries, and when SPARQL Query based you may be dealing with a CONSTRUCT or a DESCRIBE. Ideally, you would like to be able to discern via SPARQL, what SPARQL query patterns sits behind the re-write rule for a given de-referencable URI. Please confirm yay or nay. Kingsley Daniel Schwabe a écrit : Dan and Hugh, let me be more specific. I'm not really advocating that only *one* direction should be returned (or even both directions). I am asking a more general question (to which I don't think Hugh really gave an answer either) which is, is there any query that returns the same triples as the ones you get when you dereference a URI, in a site that also provides a SPARQL endpoint? In the affirmative case, I am suggesting that the corresponding query be documented in the sitemap.xml document. Does this make sense? Cheers D On 20/05/2009 14:15, Dan Brickley wrote: On 20/5/09 18:59, Daniel Schwabe wrote: Dear all, while designing Explorator [1], where one can explore one or more triple repositories that provide SPARQL enpoints (as well as direct URI dereferencing), I found the following question, to which I don't really know the answer... For the sake of this discussion, I'm considering only such sites, i.e., those that provide SPRQL enpoints. For a given URI r, is there any relation between the triples I get when I dereference it directly, as opposed to querying the SPARQL enpoint for all triples r, ?p, ?o ? Should there be (I could also get ?s, ?p, r, for example) ? For sites such as dbpedia I believe that I get the same set of triples. But I believe this is not a general behavior. Should there be a good practice about this for LoD sites that provide SPARQL endpoints? At the very least, perhaps this could also be described in the semantic sitemap.xml, no? In general, I'd be wary of doing anything that assumes the direction a property is named in is important. Taking the old MCF example, http://www.w3.org/TR/NOTE-MCF-XML-970624/#sec2.1 the_songlines eg:author bruce_chatwin . where eg:author has a domain of Document and a range of Person. Exactly the same information could be conveyed in data where the property naming direction was reversed. And case by case, different natural languages and application environments will favour slightly one direction over the other. Here we could as well have had bruce_chatwin eg:wrote the_songlines . or eg:book or eg:pub or eg:xyz, with domain Person, range Document. As it happens in English, the word author doesn't have a natural and obvious inverse here but that's incidental. The point is that both forms tell you just as much about the person as about the document, regardless of property naming and direction. The form using eg:author seems to be document-centric, but in fact it should equally support UI layers that are concerned with the person or the document. It would be dissapointing if a UI that was presenting info about Bruce Chatwin was to miss out that he was the author of the_songlines, simply because somewhere along the line a schema writer chose to deploy a property author rather than wrote... cheers, Dan [1] http://www.tecweb.inf.puc-rio.br/explorator -- Regards, Kingsley Idehen Weblog: http://www.openlinksw.com/blog/~kidehen President CEO OpenLink Software Web: http://www.openlinksw.com
Re: Dereferencing a URI vs querying a SPARQL endpoint
Kingsley Idehen wrote: Pierre-Antoine Champin wrote: I would expect that a DESCRIBE query to the SPARQL endpoint return what I get when dereferencing the URI. pa Daniel, Is this your problem: Linked Data Servers publish URIs. The mechanism that delivers these URIs tends to vary since they are the product of URL-rewrite rules that may or may not be associated with SPARQL queries, and when SPARQL Query based you may be dealing with a CONSTRUCT or a DESCRIBE. If a URI actually refers to an RDF document, I would imagine there is no URL rewriting involved; it resolves to the document itself. And for SPARQL based, who knows what I may be dealing with? Peter already exemplified that you may get something that is neither a CONSTRUCT nor a DESCRIBE... Ideally, you would like to be able to discern via SPARQL, what SPARQL query patterns sits behind the re-write rule for a given de-referencable URI. Basically yes, although I'm not even requiring being able to do it directly via SPARQL (that would be actually nice)... Also, I'd be curious to know what is more efficient - dereferencing or issuing the query through the endpoint. Cheers D