The parallel evolution of the XBRL and the Semantic Web is one of the more puzzling of technology misnomers. The Semantic Web expresses a vision about a Web of Data connected by formal meaning (Context). Congruently, XBRL espouses a vision whereby by formally defined Financial Data is accessible via the Web (and other networks). The Semantic Web uses Schemas and Ontologies for defining Data Domains while XBRL uses Taxonomies that are XML Schema Based. The Semantic Web uses XML as one of its Data Interchange formats (i.e RDF/XML) while XBRL is based on XML at all levels (model and instance data).
It is the goal of the XBRL Ontology project that we mesh the XBRL and Semantic Web realms by producing OWL based Ontologies for XBRL's XML Schema based Taxonomies as this is a critical prerequisite for generating RDF Instance Data (Triples) from XBRL Data Sources (e.g. XBRL Documents). This effort is not intended to supercede the use of XML Schemas in XBRL in any way. It simply provides a mechanism for exposing XBRL based Financial Data to the Semantic Data Web.
What are the anticipated deliverables:
1. OWL Ontologies for XBRL Taxonomies such as the XBRL GL (and others) 2. Sample RDF instance data for the XBRL Ontologies 3. SPARQL (Semantic Web Query Language) based Access Points for sample XBRL Instance Data
Benefits:
1. Transparent integration of data produced by disparate financial systems by: - Mapping of application data (e.g. SQL) to relevant XBRL Ontologies which are then exposed to WAN (Web) or LAN (Intranet) via SPARQL access points - Production of RDF Instance data from existing XBRL documents on the Web
2. An easy mechanism for plugging XBRL data into burgeoning Semantic Data Web
I would like to point you to the MUSING project launched by several EU companies and Eupropean research institutes. I have attended the creation of several XBRL ontologies last summer. If you would like, I can introduce you and your project to them, and we may have some cooperation on this common interest.
> The parallel evolution of the XBRL and the Semantic Web is one of the > more puzzling of technology misnomers. The Semantic Web expresses a > vision about a Web of Data connected by formal meaning (Context). > Congruently, XBRL espouses a vision whereby by formally defined > Financial Data is accessible via the Web (and other networks). The > Semantic Web uses Schemas and Ontologies for defining Data Domains > while XBRL uses Taxonomies that are XML Schema Based. The Semantic Web > uses XML as one of its Data Interchange formats (i.e RDF/XML) while > XBRL is based on XML at all levels (model and instance data).
> It is the goal of the XBRL Ontology project that we mesh the XBRL and > Semantic Web realms by producing OWL based Ontologies for XBRL's XML > Schema based Taxonomies as this is a critical prerequisite for > generating RDF Instance Data (Triples) from XBRL Data Sources (e.g. > XBRL Documents). This effort is not intended to supercede the use of > XML Schemas in XBRL in any way. It simply provides a mechanism for > exposing XBRL based Financial Data to the Semantic Data Web.
> What are the anticipated deliverables:
> 1. OWL Ontologies for XBRL Taxonomies such as the XBRL GL (and others) > 2. Sample RDF instance data for the XBRL Ontologies > 3. SPARQL (Semantic Web Query Language) based Access Points for > sample XBRL Instance Data
> Benefits:
> 1. Transparent integration of data produced by disparate financial > systems by: > - Mapping of application data (e.g. SQL) to relevant XBRL > Ontologies which are then exposed to WAN (Web) or LAN (Intranet) via > SPARQL access points > - Production of RDF Instance data from existing XBRL documents on > the Web
> 2. An easy mechanism for plugging XBRL data into burgeoning Semantic > Data Web
> I would like to point you to the MUSING project launched by several EU > companies and Eupropean research institutes. I have attended the > creation of several XBRL ontologies last summer. If you would like, I > can introduce you and your project to them, and we may have some > cooperation on this common interest.
Well yeah, it would be more than appreciated. Is there any references, links, etc. to hese resources and the MUSING project?
On Apr 21, 7:08 pm, Frederick Giasson <f...@fgiasson.com> wrote:
> Hi Yihong,
> > I would like to point you to the MUSING project launched by several EU > > companies and Eupropean research institutes. I have attended the > > creation of several XBRL ontologies last summer. If you would like, I > > can introduce you and your project to them, and we may have some > > cooperation on this common interest.
> Well yeah, it would be more than appreciated. Is there any references, links, > etc. to hese resources and the MUSING project?
> Salutations,
> Fred
As indicated by Fred already, we are absolutely about collaboration. The last thing we want to do is reinvent the wheel here, so any work in this area that as been done should be exposed via a URI :-)
Here is the web site of the MUSING project: http://www.musing.eu/. It is a multi-million-euro project in four years. In particular, the OWL version XBRL taxonomy conversion research is conducted by Martin Hepp, who is affiliated in DERI Innsbruck. Last summer, I was with his team and we together studied some fundamental research on the methods to convert XBRL to OWL. Moreover, his team has already successfully converted several existing XBRL standards to OWL ontologies (especially some European taxonomies, which could be a well complement to the US and Canada specifications).
Martin is a very brilliant and kind scientist. He would be glad to share with you people, I believe. Moreover, personally I would like to make some contributions to this project too. Based on my research last summer in Innsbruck, I believe this is a very valuable project, and I would like to continue contributing to this effort.
Certainly. I have already replied Fred's mail about the MUSING project. I highly value the worthiness of this effort about converting XBRL taxonomies to OWL. Financial reports are a chunk of very much valuable resources that could and should be prompted by the semantic web technologies. More importantly, financial reports usually demand rigid and even stiff rules to protect their preciseness. This property could help us avoid many ambiguities, which are often the most difficult issue when handling real-world semantic web applications. I am glad to know this new launched effort and willing to do my part of contribution to this great event. Let's see how we can cooperate better in the future.
best,
Yihong
On Apr 21, 5:14 pm, kidehen <kide...@openlinksw.com> wrote:
> On Apr 21, 7:08 pm, Frederick Giasson <f...@fgiasson.com> wrote:
> > Hi Yihong,
> > > I would like to point you to the MUSING project launched by several EU > > > companies and Eupropean research institutes. I have attended the > > > creation of several XBRL ontologies last summer. If you would like, I > > > can introduce you and your project to them, and we may have some > > > cooperation on this common interest.
> > Well yeah, it would be more than appreciated. Is there any references, links, > > etc. to hese resources and the MUSING project?
> > Salutations,
> > Fred
> As indicated by Fred already, we are absolutely about collaboration. > The last thing we want to do is reinvent the wheel here, so any work > in this area that as been done should be exposed via a URI :-)
> Here is the web site of the MUSING project: http://www.musing.eu/. It > is a multi-million-euro project in four years. In particular, the OWL > version XBRL taxonomy conversion research is conducted by Martin Hepp, > who is affiliated in DERI Innsbruck. Last summer, I was with his team
Is that effort is the same as the one outlined in the article "Translating XBRL Into Description Logic. An Approach Using protege, Sesame & OWL"?
Because I contacted these researchers and got only one answer about 1 month ago, and nothing since then (and I cannot see Mr. Hepp's name in the paper).
Also, what is the purpose of this XBRL taxonomy conversion research? The thing is that I can't find anything vis-a-vis this effort on MUSING's website. Would be nice to get some conversations examples, etc :)
> and we together studied some fundamental research on the methods to > convert XBRL to OWL. Moreover, his team has already successfully > converted several existing XBRL standards to OWL ontologies > (especially some European taxonomies, which could be a well complement > to the US and Canada specifications).
Would be great to have more information on their process, goals, etc. Would be great to know if this work could be included in the work we are intended to do with that initiative, etc.
> Martin is a very brilliant and kind scientist. He would be glad to
I will contact him right away.
> share with you people, I believe. Moreover, personally I would like to > make some contributions to this project too. Based on my research > last summer in Innsbruck, I believe this is a very valuable project, > and I would like to continue contributing to this effort.
Well you are more than welcome! you expertise in the domain is more than welcome.
The team is growing up quickly, we should start to talk about how we plan to proceed for the first steps.
it would be very useful to deal with the possibilities of interconnections between XBRL taxos and ontologies.I hope that the results of the MUSING project as well as other scientifc results (as far as I can see, at least one member of this group is invilved in such a project) could be used here. Perhaps I can meet some of you at the forthcoming international XBRL meeting in June in Munich.
I did some embedding work more intuitively at the times of XBRL 2.0 (when definiton linkbases existed, I used the to define the parent- child relation in an ontology).
Elmar Drewitz
On 22 Apr., 16:26, Frederick Giasson <f...@fgiasson.com> wrote:
> Hi Yihong> Here is the web site of the MUSING project:http://www.musing.eu/. It > > is a multi-million-euro project in four years. In particular, the OWL > > version XBRL taxonomy conversion research is conducted by Martin Hepp, > > who is affiliated in DERI Innsbruck. Last summer, I was with his team
> Is that effort is the same as the one outlined in the article > "Translating XBRL Into Description Logic. An Approach Using protege, > Sesame & OWL"?
> Because I contacted these researchers and got only one answer about 1 > month ago, and nothing since then (and I cannot see Mr. Hepp's name in > the paper).
> Also, what is the purpose of this XBRL taxonomy conversion research? The > thing is that I can't find anything vis-a-vis this effort on MUSING's > website. Would be nice to get some conversations examples, etc :)
> > and we together studied some fundamental research on the methods to > > convert XBRL to OWL. Moreover, his team has already successfully > > converted several existing XBRL standards to OWL ontologies > > (especially some European taxonomies, which could be a well complement > > to the US and Canada specifications).
> Would be great to have more information on their process, goals, etc. > Would be great to know if this work could be included in the work we are > intended to do with that initiative, etc.
> > Martin is a very brilliant and kind scientist. He would be glad to
> I will contact him right away.
> > share with you people, I believe. Moreover, personally I would like to > > make some contributions to this project too. Based on my research > > last summer in Innsbruck, I believe this is a very valuable project, > > and I would like to continue contributing to this effort.
> Well you are more than welcome! you expertise in the domain is more than > welcome.
> The team is growing up quickly, we should start to talk about how we > plan to proceed for the first steps.
I just replied and the post got lost somewhere. So here is the brief answers to some of your questions.
> Is that effort is the same as the one outlined in the article > "Translating XBRL Into Description Logic. An Approach Using protege, > Sesame & OWL"?
> Because I contacted these researchers and got only one answer about 1 > month ago, and nothing since then (and I cannot see Mr. Hepp's name in > the paper).
> Also, what is the purpose of this XBRL taxonomy conversion research? The > thing is that I can't find anything vis-a-vis this effort on MUSING's > website. Would be nice to get some conversations examples, etc :)
The paper you mentioned is written by Thierry and one of his colleague. They are from DFKI, which is another sub-contractor in the MUSING project. Thought Thierry and Martin are good friends, they are in two different institutes. Besides them, Ruben Lara and others also published a paper on this topic, "XBRL taxonomies and OWL ontologies for investment funds." It was on a workshop co-located with ER-2006 (Martin is the workshop organizer).
At that time, Martin and I had studied some fundamental methods on automating ontology conversion between XBRL and OWL. Due to the short term I stayed with DERI, we had not experimented much on our methodologies. But I know that they had converted several XBRL taxonomies to OWL (possibly by hand on most of them) for the MUSING project after my leaving. There is a repository of these converted OWL- version XBRL taxonomies. I prefer to not releasing it to the public by me since I am not the owner. I would suggest asking for Martin to release a public access to their work since there might be some legal issues on using these results for the purposes other than personal research (I am not sure though). But anyway, I can help if this is really a general interest to this group.
> The paper you mentioned is written by Thierry and one of his > colleague. They are from DFKI, which is another sub-contractor in the > MUSING project. Thought Thierry and Martin are good friends, they are > in two different institutes. Besides them, Ruben Lara and others also
Okay great, thanks for the precision
> published a paper on this topic, "XBRL taxonomies and OWL ontologies > for investment funds." It was on a workshop co-located with ER-2006 > (Martin is the workshop organizer).
Well yeah, I contacted them too.
> At that time, Martin and I had studied some fundamental methods on > automating ontology conversion between XBRL and OWL. Due to the short > term I stayed with DERI, we had not experimented much on our > methodologies. But I know that they had converted several XBRL > taxonomies to OWL (possibly by hand on most of them) for the MUSING
Yeah, this is what I read in these papers.
> project after my leaving. There is a repository of these converted OWL- > version XBRL taxonomies. I prefer to not releasing it to the public by > me since I am not the owner. I would suggest asking for Martin to > release a public access to their work since there might be some legal > issues on using these results for the purposes other than personal > research (I am not sure though). But anyway, I can help if this is > really a general interest to this group.
Okay, as you know, I sent him an email about that, so we will see.
Hope that we start collaborating on that initiative. We will check what we can do vis-a-vis the rights on their work, etc. Anyway, I have really happy to have you on board to start the development of that ontology.
> it would be very useful to deal with the possibilities of > interconnections between XBRL taxos and ontologies.I hope that the > results of the MUSING project as well as other scientifc results (as > far as I can see, at least one member of this group is invilved in > such a project) could be used here.
Well yeah, the first step is to make sure to be aware of all the projects that already working on the issue, and to get feedbacks and opinions from the larger number of people that as an interest in the project. Vis-a-vis the MUSING project, I contacted Mr. Hepp, so we will see what happen with that.
> Perhaps I can meet some of you at the forthcoming international XBRL > meeting in June in Munich.
We should see, it depends on people's availability, but the word is out now.
> I did some embedding work more intuitively at the times of XBRL 2.0 > (when definiton linkbases existed, I used the to define the parent- > child relation in an ontology).
I there any place where we could get in touch with the work you have done?
> I just replied and the post got lost somewhere. So here is the brief > answers to some of your questions.
> > Is that effort is the same as the one outlined in the article > > "Translating XBRL Into Description Logic. An Approach Using protege, > > Sesame & OWL"?
> > Because I contacted these researchers and got only one answer about 1 > > month ago, and nothing since then (and I cannot see Mr. Hepp's name in > > the paper).
> > Also, what is the purpose of this XBRL taxonomy conversion research? The > > thing is that I can't find anything vis-a-vis this effort on MUSING's > > website. Would be nice to get some conversations examples, etc :)
> The paper you mentioned is written by Thierry and one of his > colleague. They are from DFKI, which is another sub-contractor in the > MUSING project. Thought Thierry and Martin are good friends, they are > in two different institutes. Besides them, Ruben Lara and others also > published a paper on this topic, "XBRL taxonomies and OWL ontologies > for investment funds." It was on a workshop co-located with ER-2006 > (Martin is the workshop organizer).
> At that time, Martin and I had studied some fundamental methods on > automating ontology conversion between XBRL and OWL. Due to the short > term I stayed with DERI, we had not experimented much on our > methodologies. But I know that they had converted several XBRL > taxonomies to OWL (possibly by hand on most of them) for the MUSING > project after my leaving. There is a repository of these converted OWL- > version XBRL taxonomies. I prefer to not releasing it to the public by > me since I am not the owner. I would suggest asking for Martin to > release a public access to their work since there might be some legal > issues on using these results for the purposes other than personal > research (I am not sure though). But anyway, I can help if this is > really a general interest to this group.
exactly parallel to the interntl XBRL meeting in Munich, the ESWC meeting takes place at DERI in Innsbruck I just learned. Martin Hepp is one of the driving persons there.
Concerning the general aims of our group, I would firstly contribute some of my starting points:
- to define the transformation of the XBRL taxo elements (as concepts or classes, straightforward to my opinion) - define or make selectable for a transformation which linkbase defines the subclass hierarchy - the role of the other linkbases (as relations/attributes ...) as well as the multidimensional features of XBRL (used in the COREP/ FINREP taxos e. g.) - to deal with the notion of "context" in the XBRL sense (not contained in the schema files). - other XBRL features like the rule language used in XBRL (which should be substituted by a general web rule language to my unfortunately privat opinion, perhaps somebody in this group can convince the "pope of XBRL" to follow this idea...)
Other ideas will follow when some starting point is set.
But of course I hope to learn more about the transformation principles followed by the researcher groups known till now (M. Hepp and the Spanish group. By googling, I have the impression that on Taiwanesian universities, researchers deal with such questions too but the findings are denoted only in Chinese letters ...).
As I told, my work done in the past was based on XBRL 2.0 , where the definition linkbase was used to define the subclass hierarchy, so this is no longer usable for a general solution. I dealt only with GAAP taxos (German, IFRS, UK, older US-GAAP-taxos). One problem in this area is the "principle of modularization" of many GAAP taxos (like the newer IFRS taxo(s), the US GAAP taxos, not the German one which is a smooth compact taxo to my opinion).
If it makes sense, I can post some screenshots here.
Best,
Elmar
On Apr 23, 1:07 am, Frederick Giasson <f...@fgiasson.com> wrote:
> Hi Elmar,> it would be very useful to deal with the possibilities of > > interconnections between XBRL taxos and ontologies.I hope that the > > results of the MUSING project as well as other scientifc results (as > > far as I can see, at least one member of this group is invilved in > > such a project) could be used here.
> Well yeah, the first step is to make sure to be aware of all the > projects that already working on the issue, and to get feedbacks and > opinions from the larger number of people that as an interest in the > project. Vis-a-vis the MUSING project, I contacted Mr. Hepp, so we will > see what happen with that.
> > Perhaps I can meet some of you at the forthcoming international XBRL > > meeting in June in Munich.
> We should see, it depends on people's availability, but the word is out now.
> > I did some embedding work more intuitively at the times of XBRL 2.0 > > (when definiton linkbases existed, I used the to define the parent- > > child relation in an ontology).
> I there any place where we could get in touch with the work you have done?
> exactly parallel to the interntl XBRL meeting in Munich, the ESWC > meeting takes place at DERI in Innsbruck I just learned. Martin Hepp > is one of the driving persons there.
This is great, would be interesting if someone could meet him to check if it would be possible to have more information about their work.
> Concerning the general aims of our group, I would firstly contribute > some of my starting points:
> - to define the transformation of the XBRL taxo elements (as concepts > or classes, straightforward to my opinion)
Yeah, and how to create an ontology that has the power to describe all these different taxonomies. I think this is the real challenge here: describing all current taxonomies as long as all the taxonomies that doesn't even exist yet).
> - define or make selectable for a transformation which linkbase > defines the subclass hierarchy > - the role of the other linkbases (as relations/attributes ...) as > well as the multidimensional features of XBRL (used in the COREP/ > FINREP taxos e. g.)
Exact
> - to deal with the notion of "context" in the XBRL sense (not > contained in the schema files). > - other XBRL features like the rule language used in XBRL (which > should be substituted by a general web rule language to my > unfortunately privat opinion, perhaps somebody in this group can > convince the "pope of XBRL" to follow this idea...)
> Other ideas will follow when some starting point is set.
Good, what I would like, if it is possible for you, is to put these bullets on the wiki (wiki.xbrlontology.com) so that people could extend them with their personal opinions, etc. So this could be a starting point and a way to define the scope of this ontology.
> But of course I hope to learn more about the transformation principles > followed by the researcher groups known till now (M. Hepp and the > Spanish group. By googling, I have the impression that on Taiwanesian > universities, researchers deal with such questions too but the > findings are denoted only in Chinese letters ...).
Yeah, I hope too, and what I hope is not re-doing what other people have already done in the past: a wheel should keep rolling, not re- invented each time :)
> As I told, my work done in the past was based on XBRL 2.0 , where the > definition linkbase was used to define the subclass hierarchy, so this > is no longer usable for a general solution. I dealt only with GAAP > taxos (German, IFRS, UK, older US-GAAP-taxos). One problem in this > area is the "principle of modularization" of many GAAP taxos (like the > newer IFRS taxo(s), the US GAAP taxos, not the German one which is a > smooth compact taxo to my opinion).
> If it makes sense, I can post some screenshots here.
Yeah sure, go ahead. Here and/or on the wiki so that we can keep a track of these screenshots.
> > exactly parallel to the interntl XBRL meeting in Munich, the ESWC > > meeting takes place at DERI in Innsbruck I just learned. Martin Hepp > > is one of the driving persons there.
> This is great, would be interesting if someone could meet him to check > if it would be possible to have more information about their work.
> > Concerning the general aims of our group, I would firstly contribute > > some of my starting points:
> > - to define the transformation of the XBRL taxo elements (as concepts > > or classes, straightforward to my opinion)
> Yeah, and how to create an ontology that has the power to describe all > these different taxonomies. I think this is the real challenge here: > describing all current taxonomies as long as all the taxonomies that > doesn't even exist yet).
Well, an ontology that describes the metamodel of XBRL with rules of creation could be the point. This is a nice scientific task I think, perhaps a bachelor thesis we could propose to the academic scientists here or outside this group. I could play the role of a practical advisor ...
> > - define or make selectable for a transformation which linkbase > > defines the subclass hierarchy > > - the role of the other linkbases (as relations/attributes ...) as > > well as the multidimensional features of XBRL (used in the COREP/ > > FINREP taxos e. g.)
> Exact
> > - to deal with the notion of "context" in the XBRL sense (not > > contained in the schema files). > > - other XBRL features like the rule language used in XBRL (which > > should be substituted by a general web rule language to my > > unfortunately privat opinion, perhaps somebody in this group can > > convince the "pope of XBRL" to follow this idea...)
> > Other ideas will follow when some starting point is set.
> Good, what I would like, if it is possible for you, is to put these > bullets on the wiki (wiki.xbrlontology.com) so that people could > extend them with their personal opinions, etc. So this could be a > starting point and a way to define the scope of this ontology.
> > But of course I hope to learn more about the transformation principles > > followed by the researcher groups known till now (M. Hepp and the > > Spanish group. By googling, I have the impression that on Taiwanesian > > universities, researchers deal with such questions too but the > > findings are denoted only in Chinese letters ...).
> Yeah, I hope too, and what I hope is not re-doing what other people > have already done in the past: a wheel should keep rolling, not re- > invented each time :)
> > As I told, my work done in the past was based on XBRL 2.0 , where the > > definition linkbase was used to define the subclass hierarchy, so this > > is no longer usable for a general solution. I dealt only with GAAP > > taxos (German, IFRS, UK, older US-GAAP-taxos). One problem in this > > area is the "principle of modularization" of many GAAP taxos (like the > > newer IFRS taxo(s), the US GAAP taxos, not the German one which is a > > smooth compact taxo to my opinion).
> > If it makes sense, I can post some screenshots here.
> Yeah sure, go ahead. Here and/or on the wiki so that we can keep a > track of these screenshots.
> Well, an ontology that describes the metamodel of XBRL with rules of > creation could be the point. This is a nice scientific task I think, > perhaps a bachelor thesis we could propose to the academic scientists > here or outside this group. I could play the role of a practical > advisor ...
Yeah it is sure that it could be a great way to get the first step done for the elaboration of this ontology. In fact, the first goal was to create this community to do exactly that. However, if you know anyone interested in that, then we could proceed that way and help him to do that work.
So, go ahead with that idea and check if someone could be interested and keep me updated :)
|
