I would say that GND is a “good enough” answer.
Most named entities are persons, organizations, events, creative works and places and these are all mutually exclusive. There ought to be a system interlock to prevent confusion between them.
“Organism Classification” or whatever you call it should also be on the list, because of prevalence.
One thing I’d add to that is fictional character because there are a (1) lot of them and (2) they can be ontologized more-or-less in parallel with people, and (3) you’ll get cleaner people if you keep fictional characters out. (On the other hand, there are fictional events, places, etc. too, though these are not so well documented.) Is it easy to add a new GND type?
I think you’re calling the “wastebin” category term, which is reasonable (I’d call it a “concept”.) Going much further than this you’ll run into Borges encyclopedia style risks, but aren’t the categories named in GND upwards of 80% of the topics? Can you run a report on this?
From: Sven Manguard Sent: Sunday, June 30, 2013 2:19 PM To: Discussion list for the Wikidata project. Subject: [Wikidata-l] A solution with finality is needed for P107 - maintype (GND)
I have just closed a second deletion discussion for Property:P107 - main type (GND).
As with the first discussion, it is clear that there is a broad sense that main type (GND) is not an ideal solution, however as it stands now, a large enough portion of the community does not want to get rid of it unless/until a replacement system is found or developed. For this reason, I closed the discussion as no consensus and opened up a request for comment on the matter of finding a replacement for P107.
I have gone to the unusual step of emailing the mailing list for three reasons. First, P107 is the most used property on the project, and it or its replacement will (most likely) remain the most used property on the project forever. Second, the GND has evolved into a component of how Wikidata is structured; our lists of properties are sorted by GND type, and that has a real impact on what properties are used on what pages. The third reason is that, as a general statement, participation levels in requests for comment have been downright sad. Three or four people participating in an RfC is, for a project of this size, unhealthy, and most RfCs don't get more than that many people participating in them. For something this important, we need at least a dozen people, preferably at least twice that.</rant>
Anyways, the RfC is at https://www.wikidata.org/wiki/Wikidata:Requests_for_comment/Primary_sorting_... and I hope that, with broad participation, we can finally resolve this issue.
Yours, Sven
-------------------------------------------------------------------------------- _______________________________________________ Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Hoi, The reason why it is NOT good enough and what you fail to understand is that this is NOT an attribute that we should morph into something else. Its name makes it clear: "main type (GND)" this implies that the definition and its values are external to Wikidata; they are the definition as per the GND.
For me it means that when a specific value of this main type makes sense... ie it is about a person, I use it. I do not use it for any other value. The added value for using it is some of the tools that INSIST on its use. From a theoretical point of view, "instance of" serves us equally well without relying on external values and systems.
The reason why I proposed the removal of p107 is that people give it a value that they do not support by providing arguments and guidance on how to ensure that data entered is valid. So far I have noticed that Wikidata is seen as secondary to whatever Wikipedia. In my opinion Wikidata is a project in its own right and many artefacts of Wikipedia just do not belong in Wikidata. P107 is one such artefact. Thanks, GerardM
On 1 July 2013 17:03, Paul A. Houle paul@ontology2.com wrote:
I would say that GND is a “good enough” answer. Most named entities are persons, organizations, events, creativeworks and places and these are all mutually exclusive. There ought to be a system interlock to prevent confusion between them.
“Organism Classification” or whatever you call it should also be onthe list, because of prevalence.
One thing I’d add to that is fictional character because there are a(1) lot of them and (2) they can be ontologized more-or-less in parallel with people, and (3) you’ll get cleaner people if you keep fictional characters out. (On the other hand, there are fictional events, places, etc. too, though these are not so well documented.) Is it easy to add a new GND type?
I think you’re calling the “wastebin” category term, which is reasonable (I’d call it a “concept”.) Going much further than this you’ll run into Borges encyclopedia style risks, but aren’t the categories named in GND upwards of 80% of the topics? Can you run a report on this?
*From:* Sven Manguard svenmanguard@gmail.com *Sent:* Sunday, June 30, 2013 2:19 PM *To:* Discussion list for the Wikidata project.wikidata-l@lists.wikimedia.org *Subject:* [Wikidata-l] A solution with finality is needed for P107 - maintype (GND)
I have just closed a second deletion discussion for Property:P107 - main type (GND).
As with the first discussion, it is clear that there is a broad sense that main type (GND) is not an ideal solution, however as it stands now, a large enough portion of the community does not want to get rid of it unless/until a replacement system is found or developed. For this reason, I closed the discussion as no consensus and opened up a request for comment on the matter of finding a replacement for P107.
I have gone to the unusual step of emailing the mailing list for three reasons. First, P107 is the most used property on the project, and it or its replacement will (most likely) remain the most used property on the project forever. Second, the GND has evolved into a component of how Wikidata is structured; our lists of properties are sorted by GND type, and that has a real impact on what properties are used on what pages. The third reason is that, as a general statement, participation levels in requests for comment have been downright sad. Three or four people participating in an RfC is, for a project of this size, unhealthy, and most RfCs don't get more than that many people participating in them. For something this important, we need at least a dozen people, preferably at least twice that.
</rant>
Anyways, the RfC is at https://www.wikidata.org/wiki/Wikidata:Requests_for_comment/Primary_sorting_... I hope that, with broad participation, we can finally resolve this issue.
Yours, Sven
Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
We really need to keep everything in one forum. Can you two please copy your comments to https://www.wikidata.org/wiki/Wikidata:Requests_for_comment/Primary_sorting_... continue the discussion there? I worry that the person closing the discussion might not be on the mailing list and might not see your points.
S On Jul 1, 2013 12:16 PM, "Gerard Meijssen" gerard.meijssen@gmail.com wrote:
Hoi, The reason why it is NOT good enough and what you fail to understand is that this is NOT an attribute that we should morph into something else. Its name makes it clear: "main type (GND)" this implies that the definition and its values are external to Wikidata; they are the definition as per the GND.
For me it means that when a specific value of this main type makes sense... ie it is about a person, I use it. I do not use it for any other value. The added value for using it is some of the tools that INSIST on its use. From a theoretical point of view, "instance of" serves us equally well without relying on external values and systems.
The reason why I proposed the removal of p107 is that people give it a value that they do not support by providing arguments and guidance on how to ensure that data entered is valid. So far I have noticed that Wikidata is seen as secondary to whatever Wikipedia. In my opinion Wikidata is a project in its own right and many artefacts of Wikipedia just do not belong in Wikidata. P107 is one such artefact. Thanks, GerardM
On 1 July 2013 17:03, Paul A. Houle paul@ontology2.com wrote:
I would say that GND is a “good enough” answer. Most named entities are persons, organizations, events, creativeworks and places and these are all mutually exclusive. There ought to be a system interlock to prevent confusion between them.
“Organism Classification” or whatever you call it should also be onthe list, because of prevalence.
One thing I’d add to that is fictional character because there are a(1) lot of them and (2) they can be ontologized more-or-less in parallel with people, and (3) you’ll get cleaner people if you keep fictional characters out. (On the other hand, there are fictional events, places, etc. too, though these are not so well documented.) Is it easy to add a new GND type?
I think you’re calling the “wastebin” category term, which is reasonable (I’d call it a “concept”.) Going much further than this you’ll run into Borges encyclopedia style risks, but aren’t the categories named in GND upwards of 80% of the topics? Can you run a report on this?
*From:* Sven Manguard svenmanguard@gmail.com *Sent:* Sunday, June 30, 2013 2:19 PM *To:* Discussion list for the Wikidata project.wikidata-l@lists.wikimedia.org *Subject:* [Wikidata-l] A solution with finality is needed for P107 - maintype (GND)
I have just closed a second deletion discussion for Property:P107 - main type (GND).
As with the first discussion, it is clear that there is a broad sense that main type (GND) is not an ideal solution, however as it stands now, a large enough portion of the community does not want to get rid of it unless/until a replacement system is found or developed. For this reason, I closed the discussion as no consensus and opened up a request for comment on the matter of finding a replacement for P107.
I have gone to the unusual step of emailing the mailing list for three reasons. First, P107 is the most used property on the project, and it or its replacement will (most likely) remain the most used property on the project forever. Second, the GND has evolved into a component of how Wikidata is structured; our lists of properties are sorted by GND type, and that has a real impact on what properties are used on what pages. The third reason is that, as a general statement, participation levels in requests for comment have been downright sad. Three or four people participating in an RfC is, for a project of this size, unhealthy, and most RfCs don't get more than that many people participating in them. For something this important, we need at least a dozen people, preferably at least twice that.
</rant>
Anyways, the RfC is at https://www.wikidata.org/wiki/Wikidata:Requests_for_comment/Primary_sorting_... I hope that, with broad participation, we can finally resolve this issue.
Yours, Sven
Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l
Wikidata-l mailing list Wikidata-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-l