5 ms·

Related, see this older attempt to create a structured dataset from Wikipedia rooting back to the semantic web: https://en.wikipedia.org/wiki/DBpedia https://en

Related, see this older attempt to create a structured dataset from Wikipedia rooting back to the semantic web: https://en.wikipedia.org/wiki/DBpedia https://en.wikipedia.org/wiki/DBpedia "DBpedia (from "DB" for "database") is a project aiming to extract structured content from the information created in the Wikipedia project."

d--b 5y ago
This is different though. This is from wikipedia, and aims at structuring data that is used in wikipedia. I am guessing this is because if something changes in the world, then in wikipedia, the information needs to be changed in all articles (all languages, all connex articles etc.). This would add a level of indirection for facts.
vinni2 5y ago
Ya but I haven’t seen a new release of DBpedia in years.
- geraltofrivia 5y ago
  DBPedia board has been faffing about regarding the new hyper-relational data model [0] and has come up with a bunch of hack-ish solutions (read: esoteric data models) to incorporate it. As far as I remember, there were dumps from 2018 with some hyper-relational facts but by and large the go-to dumps is from 2016-10 [1]. I think they acknowledge that info-box information is no longer good enough (both in coverage and completeness) and have handed over the moniker of the open-access KG to Wikidata. To their credit, I think its the right call. src: I worked in a lab lead by DBPedia's founder (and one of the contributors to the original infobox to KB code). [0] https://blog.liu.se/olafhartig/2019/01/10/position-statement-rdf-star-and-sparql-star/ https://blog.liu.se/olafhartig/2019/01/10/position-statement... ^^This enables, amongst other things, fact's validity to be quantified. Instead of saying <barack obama> <president> <united states>, you can now say {<barack obama> <president> <united states>} <from> <DD-MM-YYYY>; <to> <DD-MM-YYYY> [1] http://downloads.dbpedia.org/current/ http://downloads.dbpedia.org/current/ (from 2018); http://downloads.dbpedia.org/ http://downloads.dbpedia.org/ (see the folder titled 2016-10)
mmarx 5y ago
Unlike DBpedia, Wikidata is not structured content extracted by parsing Wikipedia articles, but instead a structured collection of factual knowledge created by human editors.
- tshaddox 5y ago
  Human editors, and many many bots.
  - mmarx 5y ago
    True, but then there are also quite a few bots editing the Wikipedias. In the end, bots are still run by humans, and are subject to prior community approval.

d--b 5y ago

This is different though. This is from wikipedia, and aims at structuring data that is used in wikipedia. I am guessing this is because if something changes in the world, then in wikipedia, the information needs to be changed in all articles (all languages, all connex articles etc.). This would add a level of indirection for facts.

vinni2 5y ago

Ya but I haven’t seen a new release of DBpedia in years.

geraltofrivia 5y ago
DBPedia board has been faffing about regarding the new hyper-relational data model [0] and has come up with a bunch of hack-ish solutions (read: esoteric data models) to incorporate it. As far as I remember, there were dumps from 2018 with some hyper-relational facts but by and large the go-to dumps is from 2016-10 [1]. I think they acknowledge that info-box information is no longer good enough (both in coverage and completeness) and have handed over the moniker of the open-access KG to Wikidata. To their credit, I think its the right call. src: I worked in a lab lead by DBPedia's founder (and one of the contributors to the original infobox to KB code). [0] https://blog.liu.se/olafhartig/2019/01/10/position-statement-rdf-star-and-sparql-star/ https://blog.liu.se/olafhartig/2019/01/10/position-statement... ^^This enables, amongst other things, fact's validity to be quantified. Instead of saying <barack obama> <president> <united states>, you can now say {<barack obama> <president> <united states>} <from> <DD-MM-YYYY>; <to> <DD-MM-YYYY> [1] http://downloads.dbpedia.org/current/ http://downloads.dbpedia.org/current/ (from 2018); http://downloads.dbpedia.org/ http://downloads.dbpedia.org/ (see the folder titled 2016-10)

mmarx 5y ago

Unlike DBpedia, Wikidata is not structured content extracted by parsing Wikipedia articles, but instead a structured collection of factual knowledge created by human editors.

tshaddox 5y ago
Human editors, and many many bots.
- mmarx 5y ago
  True, but then there are also quite a few bots editing the Wikipedias. In the end, bots are still run by humans, and are subject to prior community approval.