Hi all!
I really try not to spam the chat too much with pointers to my work on the Abstract Wikipedia, but this one is probably also interesting for Wikidata contributors. It is the draft for a chapter submitted to Koerner and Reagle's Wikipedia@20 book, and talks about knowledge diversity under the light of centralisation through projects such as Wikidata.
Public commenting phase is open until July 19, and very welcome: "Collaborating on the sum of all knowledge across languages"
About the book: https://meta.wikimedia.org/wiki/Wikipedia@20 Link to chapter: https://wikipedia20.pubpub.org/pub/vyf7ksah
Cheers, Denny
Hoi, I do not have a profile there and I am nowadays reluctant to add profiles.
Having said that. The data that will be in an abstract Wikipedia allows for the generation of text in any language. What it does is enable the generation of texts on the fly to be used as a complement to any Wikipedia. Best is not to save generated texts but to cache them. We could even seek for the routines and data used to generate articles for for instance the Cebuano Wikipedia, remove the completely generated articles and keep the same level of service. In this way it is for Wikipedians to write text and have generated text as a basis to improve upon.
Hidden in Reasonator there is the "concept cloud". This is just one example [1]. It shows the items that are common in the Wikipedias for an item. With an abstract Wikipedia we can compare and indicate what articles do not have links to concepts that are known to be good. There may be all kinds of reasons for that but it is easy to find the wrong reasons when you compare it with what others include. You will find the problematic, the erroneous and the bigoted links. When we do we could override the delivery of that article with a choice to the "abstract" for comparison. Thanks, GerardM
On Sat, 6 Jul 2019 at 23:53, Denny Vrandečić [email protected] wrote:
Hi all!
I really try not to spam the chat too much with pointers to my work on the Abstract Wikipedia, but this one is probably also interesting for Wikidata contributors. It is the draft for a chapter submitted to Koerner and Reagle's Wikipedia@20 book, and talks about knowledge diversity under the light of centralisation through projects such as Wikidata.
Public commenting phase is open until July 19, and very welcome: "Collaborating on the sum of all knowledge across languages"
About the book: https://meta.wikimedia.org/wiki/Wikipedia@20 Link to chapter: https://wikipedia20.pubpub.org/pub/vyf7ksah
Cheers, Denny _______________________________________________ Wikidata mailing list [email protected] https://lists.wikimedia.org/mailman/listinfo/wikidata
Hi Gerard,
thank you for your comments. I agree with them - the generated text shouldn't be stored in the local Wikipedias, but merely cached. I have updated the text accordingly to make it explicit.
Thanks! Cheers, Denny
On Sat, Jul 6, 2019 at 11:09 PM Gerard Meijssen [email protected] wrote:
Hoi, I do not have a profile there and I am nowadays reluctant to add profiles.
Having said that. The data that will be in an abstract Wikipedia allows for the generation of text in any language. What it does is enable the generation of texts on the fly to be used as a complement to any Wikipedia. Best is not to save generated texts but to cache them. We could even seek for the routines and data used to generate articles for for instance the Cebuano Wikipedia, remove the completely generated articles and keep the same level of service. In this way it is for Wikipedians to write text and have generated text as a basis to improve upon.
Hidden in Reasonator there is the "concept cloud". This is just one example [1]. It shows the items that are common in the Wikipedias for an item. With an abstract Wikipedia we can compare and indicate what articles do not have links to concepts that are known to be good. There may be all kinds of reasons for that but it is easy to find the wrong reasons when you compare it with what others include. You will find the problematic, the erroneous and the bigoted links. When we do we could override the delivery of that article with a choice to the "abstract" for comparison. Thanks, GerardM
On Sat, 6 Jul 2019 at 23:53, Denny Vrandečić [email protected] wrote:
Hi all!
I really try not to spam the chat too much with pointers to my work on the Abstract Wikipedia, but this one is probably also interesting for Wikidata contributors. It is the draft for a chapter submitted to Koerner and Reagle's Wikipedia@20 book, and talks about knowledge diversity under the light of centralisation through projects such as Wikidata.
Public commenting phase is open until July 19, and very welcome: "Collaborating on the sum of all knowledge across languages"
About the book: https://meta.wikimedia.org/wiki/Wikipedia@20 Link to chapter: https://wikipedia20.pubpub.org/pub/vyf7ksah
Cheers, Denny _______________________________________________ Wikidata mailing list [email protected] https://lists.wikimedia.org/mailman/listinfo/wikidata
Wikidata mailing list [email protected] https://lists.wikimedia.org/mailman/listinfo/wikidata