Hoi, Much of the content of DBpedia and Wikidata have the same origin; harvesting data from a Wikipedia. There is a lot of discussion going on about quality and one point that I make is that comparing "Sources" and concentrating on the differences particularly where statements differ is where it is easiest to make a quality difference.
So given that DBpedia harvests both Wikipedia and Wikidata, can it provide us with a view where a Wikipedia statement and a Wikidata statement differ. To make it useful, it is important to subset this data. I will not start with 500.000 differences but I will begin when they are about a subset that I care about.
When I care about entries for alumni of a university, I will consider curating the information in question. Particularly when I know the language of the Wikipedia.
When we can do this, another thing that will promote the use of a tool like this is when regularly (say once a month) numbers are stored and trends are published.
How difficult is it to come up with something like this. I know this tool would be based on DBpedia but there are several reasons why this is good. First it gives added relevance to DBpedia (without detracting from Wikidata) and secondly as DBpedia updates on RSS changes for several Wikipedias, the effect of these changes is quickly noticed when a new set of data is requested.
Please let us know what the issues are and what it takes to move forward with this, Does this make sense? Thanks, GerardM
http://ultimategerardm.blogspot.nl/2017/03/quality-dbpedia-and-kappa-alpha-p...
Hi
I don't have an idea about how to develop this, but it seems like an interesting project!
Best, Reem
On 30 Mar 2017 10:17, "Gerard Meijssen" gerard.meijssen@gmail.com wrote:
Hoi, Much of the content of DBpedia and Wikidata have the same origin; harvesting data from a Wikipedia. There is a lot of discussion going on about quality and one point that I make is that comparing "Sources" and concentrating on the differences particularly where statements differ is where it is easiest to make a quality difference.
So given that DBpedia harvests both Wikipedia and Wikidata, can it provide us with a view where a Wikipedia statement and a Wikidata statement differ. To make it useful, it is important to subset this data. I will not start with 500.000 differences but I will begin when they are about a subset that I care about.
When I care about entries for alumni of a university, I will consider curating the information in question. Particularly when I know the language of the Wikipedia.
When we can do this, another thing that will promote the use of a tool like this is when regularly (say once a month) numbers are stored and trends are published.
How difficult is it to come up with something like this. I know this tool would be based on DBpedia but there are several reasons why this is good. First it gives added relevance to DBpedia (without detracting from Wikidata) and secondly as DBpedia updates on RSS changes for several Wikipedias, the effect of these changes is quickly noticed when a new set of data is requested.
Please let us know what the issues are and what it takes to move forward with this, Does this make sense? Thanks, GerardM
http://ultimategerardm.blogspot.nl/2017/03/quality- dbpedia-and-kappa-alpha-psi.html
Wikidata mailing list Wikidata@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata
Hoi, I was asked by one of the DBpedia people to write a project plan.. I gave it a try [1].
The idea is to first compare DBpedia with Wikidata where a comparison is possible. When it is not (differences in their classes for instance) it is at first not what we focus on.
Please comment on the talk page and when there are things missing in the plan, please help it improve. Thanks, GerardM
[1] https://www.wikidata.org/wiki/User:GerardM/DBpedia_for_Quality
On 1 April 2017 at 10:44, Reem Al-Kashif reemalkashif@gmail.com wrote:
Hi
I don't have an idea about how to develop this, but it seems like an interesting project!
Best, Reem
On 30 Mar 2017 10:17, "Gerard Meijssen" gerard.meijssen@gmail.com wrote:
Hoi, Much of the content of DBpedia and Wikidata have the same origin; harvesting data from a Wikipedia. There is a lot of discussion going on about quality and one point that I make is that comparing "Sources" and concentrating on the differences particularly where statements differ is where it is easiest to make a quality difference.
So given that DBpedia harvests both Wikipedia and Wikidata, can it provide us with a view where a Wikipedia statement and a Wikidata statement differ. To make it useful, it is important to subset this data. I will not start with 500.000 differences but I will begin when they are about a subset that I care about.
When I care about entries for alumni of a university, I will consider curating the information in question. Particularly when I know the language of the Wikipedia.
When we can do this, another thing that will promote the use of a tool like this is when regularly (say once a month) numbers are stored and trends are published.
How difficult is it to come up with something like this. I know this tool would be based on DBpedia but there are several reasons why this is good. First it gives added relevance to DBpedia (without detracting from Wikidata) and secondly as DBpedia updates on RSS changes for several Wikipedias, the effect of these changes is quickly noticed when a new set of data is requested.
Please let us know what the issues are and what it takes to move forward with this, Does this make sense? Thanks, GerardM
http://ultimategerardm.blogspot.nl/2017/03/quality-dbpedia- and-kappa-alpha-psi.html
Wikidata mailing list Wikidata@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata
Wikidata mailing list Wikidata@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata
Hoi, With the recent introduction of federation for DBpedia, it is possible to have queries for the DBpedias for a specific language and Wikidata. I have blogged how we can make use for this [1].
It makes it much easier to compare Wikidata and DBpedia and when we take this serious and apply some effort we can make a tool like the one by Pasleim [2] for Wikipedias that do not have a category for people who died in a given year. Thanks, GerardM
[1] http://ultimategerardm.blogspot.nl/2017/04/wikidata-user-story-dbpedia-death... [2] http://tools.wmflabs.org/pltools/recentdeaths/
On 1 April 2017 at 11:34, Gerard Meijssen gerard.meijssen@gmail.com wrote:
Hoi, I was asked by one of the DBpedia people to write a project plan.. I gave it a try [1].
The idea is to first compare DBpedia with Wikidata where a comparison is possible. When it is not (differences in their classes for instance) it is at first not what we focus on.
Please comment on the talk page and when there are things missing in the plan, please help it improve. Thanks, GerardM
[1] https://www.wikidata.org/wiki/User:GerardM/DBpedia_for_Quality
On 1 April 2017 at 10:44, Reem Al-Kashif reemalkashif@gmail.com wrote:
Hi
I don't have an idea about how to develop this, but it seems like an interesting project!
Best, Reem
On 30 Mar 2017 10:17, "Gerard Meijssen" gerard.meijssen@gmail.com wrote:
Hoi, Much of the content of DBpedia and Wikidata have the same origin; harvesting data from a Wikipedia. There is a lot of discussion going on about quality and one point that I make is that comparing "Sources" and concentrating on the differences particularly where statements differ is where it is easiest to make a quality difference.
So given that DBpedia harvests both Wikipedia and Wikidata, can it provide us with a view where a Wikipedia statement and a Wikidata statement differ. To make it useful, it is important to subset this data. I will not start with 500.000 differences but I will begin when they are about a subset that I care about.
When I care about entries for alumni of a university, I will consider curating the information in question. Particularly when I know the language of the Wikipedia.
When we can do this, another thing that will promote the use of a tool like this is when regularly (say once a month) numbers are stored and trends are published.
How difficult is it to come up with something like this. I know this tool would be based on DBpedia but there are several reasons why this is good. First it gives added relevance to DBpedia (without detracting from Wikidata) and secondly as DBpedia updates on RSS changes for several Wikipedias, the effect of these changes is quickly noticed when a new set of data is requested.
Please let us know what the issues are and what it takes to move forward with this, Does this make sense? Thanks, GerardM
http://ultimategerardm.blogspot.nl/2017/03/quality-dbpedia-a nd-kappa-alpha-psi.html
Wikidata mailing list Wikidata@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata
Wikidata mailing list Wikidata@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata
Hi, a few clarifications from my side A more up-to-date link for the data is here: https://gist.github.com/ jimkont/01f6add8527939c39192bcb3f840eca0 and this dump was not generated with federated queries, as this was not possible at the time of creation but with a simple script
it is meant only as a proof-of-concept project that showcases differences in birthdates between Wikidata, Dutch Wikipedia and Greek Wikipedia as extracted in the DBpedia 2016-04 release, (which also means that it is based on ~1y old data)
Best, Dimitris
On Mon, Apr 17, 2017 at 4:14 PM, Gerard Meijssen gerard.meijssen@gmail.com wrote:
Hoi, With the recent introduction of federation for DBpedia, it is possible to have queries for the DBpedias for a specific language and Wikidata. I have blogged how we can make use for this [1].
It makes it much easier to compare Wikidata and DBpedia and when we take this serious and apply some effort we can make a tool like the one by Pasleim [2] for Wikipedias that do not have a category for people who died in a given year. Thanks, GerardM
[1] http://ultimategerardm.blogspot.nl/2017/04/wikidata-user -story-dbpedia-death-and.html [2] http://tools.wmflabs.org/pltools/recentdeaths/
On 1 April 2017 at 11:34, Gerard Meijssen gerard.meijssen@gmail.com wrote:
Hoi, I was asked by one of the DBpedia people to write a project plan.. I gave it a try [1].
The idea is to first compare DBpedia with Wikidata where a comparison is possible. When it is not (differences in their classes for instance) it is at first not what we focus on.
Please comment on the talk page and when there are things missing in the plan, please help it improve. Thanks, GerardM
[1] https://www.wikidata.org/wiki/User:GerardM/DBpedia_for_Quality
On 1 April 2017 at 10:44, Reem Al-Kashif reemalkashif@gmail.com wrote:
Hi
I don't have an idea about how to develop this, but it seems like an interesting project!
Best, Reem
On 30 Mar 2017 10:17, "Gerard Meijssen" gerard.meijssen@gmail.com wrote:
Hoi, Much of the content of DBpedia and Wikidata have the same origin; harvesting data from a Wikipedia. There is a lot of discussion going on about quality and one point that I make is that comparing "Sources" and concentrating on the differences particularly where statements differ is where it is easiest to make a quality difference.
So given that DBpedia harvests both Wikipedia and Wikidata, can it provide us with a view where a Wikipedia statement and a Wikidata statement differ. To make it useful, it is important to subset this data. I will not start with 500.000 differences but I will begin when they are about a subset that I care about.
When I care about entries for alumni of a university, I will consider curating the information in question. Particularly when I know the language of the Wikipedia.
When we can do this, another thing that will promote the use of a tool like this is when regularly (say once a month) numbers are stored and trends are published.
How difficult is it to come up with something like this. I know this tool would be based on DBpedia but there are several reasons why this is good. First it gives added relevance to DBpedia (without detracting from Wikidata) and secondly as DBpedia updates on RSS changes for several Wikipedias, the effect of these changes is quickly noticed when a new set of data is requested.
Please let us know what the issues are and what it takes to move forward with this, Does this make sense? Thanks, GerardM
http://ultimategerardm.blogspot.nl/2017/03/quality-dbpedia-a nd-kappa-alpha-psi.html
Wikidata mailing list Wikidata@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata
Wikidata mailing list Wikidata@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata
Wikidata mailing list Wikidata@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata
On 4/18/17 5:04 AM, Dimitris Kontokostas wrote:
Hi, a few clarifications from my side A more up-to-date link for the data is here: https://gist.github.com/jimkont/01f6add8527939c39192bcb3f840eca0 https://gist.github.com/jimkont/01f6add8527939c39192bcb3f840eca0 and this dump was not generated with federated queries, as this was not possible at the time of creation but with a simple script
it is meant only as a proof-of-concept project that showcases differences in birthdates between Wikidata, Dutch Wikipedia and Greek Wikipedia as extracted in the DBpedia 2016-04 release, (which also means that it is based on ~1y old data)
Best, Dimitris
Dimitris,
Here are links to SPARQL-FED queries that hit both DBpedia and Wikidata endpoints. Basically, you have a federated query over 3 SPARQL Endpoints.
[1] https://tinyurl.com/mdkgg7h -- SPARQL-FED Results Page
[2] https://medium.com/virtuoso-blog/on-the-mutually-beneficial-nature-of-dbpedi... -- Post about both projects using variant of SPARQL-FED example above.
Kingsley
On Mon, Apr 17, 2017 at 4:14 PM, Gerard Meijssen <gerard.meijssen@gmail.com mailto:gerard.meijssen@gmail.com> wrote:
Hoi, With the recent introduction of federation for DBpedia, it is possible to have queries for the DBpedias for a specific language and Wikidata. I have blogged how we can make use for this [1]. It makes it much easier to compare Wikidata and DBpedia and when we take this serious and apply some effort we can make a tool like the one by Pasleim [2] for Wikipedias that do not have a category for people who died in a given year. Thanks, GerardM [1] http://ultimategerardm.blogspot.nl/2017/04/wikidata-user-story-dbpedia-death-and.html <http://ultimategerardm.blogspot.nl/2017/04/wikidata-user-story-dbpedia-death-and.html> [2] http://tools.wmflabs.org/pltools/recentdeaths/ <http://tools.wmflabs.org/pltools/recentdeaths/> On 1 April 2017 at 11:34, Gerard Meijssen <gerard.meijssen@gmail.com <mailto:gerard.meijssen@gmail.com>> wrote: Hoi, I was asked by one of the DBpedia people to write a project plan.. I gave it a try [1]. The idea is to first compare DBpedia with Wikidata where a comparison is possible. When it is not (differences in their classes for instance) it is at first not what we focus on. Please comment on the talk page and when there are things missing in the plan, please help it improve. Thanks, GerardM [1] https://www.wikidata.org/wiki/User:GerardM/DBpedia_for_Quality <https://www.wikidata.org/wiki/User:GerardM/DBpedia_for_Quality> On 1 April 2017 at 10:44, Reem Al-Kashif <reemalkashif@gmail.com <mailto:reemalkashif@gmail.com>> wrote: Hi I don't have an idea about how to develop this, but it seems like an interesting project! Best, Reem On 30 Mar 2017 10:17, "Gerard Meijssen" <gerard.meijssen@gmail.com <mailto:gerard.meijssen@gmail.com>> wrote: Hoi, Much of the content of DBpedia and Wikidata have the same origin; harvesting data from a Wikipedia. There is a lot of discussion going on about quality and one point that I make is that comparing "Sources" and concentrating on the differences particularly where statements differ is where it is easiest to make a quality difference. So given that DBpedia harvests both Wikipedia and Wikidata, can it provide us with a view where a Wikipedia statement and a Wikidata statement differ. To make it useful, it is important to subset this data. I will not start with 500.000 differences but I will begin when they are about a subset that I care about. When I care about entries for alumni of a university, I will consider curating the information in question. Particularly when I know the language of the Wikipedia. When we can do this, another thing that will promote the use of a tool like this is when regularly (say once a month) numbers are stored and trends are published. How difficult is it to come up with something like this. I know this tool would be based on DBpedia but there are several reasons why this is good. First it gives added relevance to DBpedia (without detracting from Wikidata) and secondly as DBpedia updates on RSS changes for several Wikipedias, the effect of these changes is quickly noticed when a new set of data is requested. Please let us know what the issues are and what it takes to move forward with this, Does this make sense? Thanks, GerardM http://ultimategerardm.blogspot.nl/2017/03/quality-dbpedia-and-kappa-alpha-psi.html <http://ultimategerardm.blogspot.nl/2017/03/quality-dbpedia-and-kappa-alpha-psi.html> _______________________________________________ Wikidata mailing list Wikidata@lists.wikimedia.org <mailto:Wikidata@lists.wikimedia.org> https://lists.wikimedia.org/mailman/listinfo/wikidata <https://lists.wikimedia.org/mailman/listinfo/wikidata> _______________________________________________ Wikidata mailing list Wikidata@lists.wikimedia.org <mailto:Wikidata@lists.wikimedia.org> https://lists.wikimedia.org/mailman/listinfo/wikidata <https://lists.wikimedia.org/mailman/listinfo/wikidata> _______________________________________________ Wikidata mailing list Wikidata@lists.wikimedia.org <mailto:Wikidata@lists.wikimedia.org> https://lists.wikimedia.org/mailman/listinfo/wikidata <https://lists.wikimedia.org/mailman/listinfo/wikidata>-- Kontokostas Dimitris
Wikidata mailing list Wikidata@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata