dbpedia / mappings-tracker

This project is used for tracking mapping issues in mappings.dbpedia.org
9 stars 6 forks source link

statistics - properties incorrectly shown as not in template or not used #74

Open CaptSolo opened 8 years ago

CaptSolo commented 8 years ago

Statistics shows incorrect information about some infobox properties:

Example:

It was suggested on dbpedia-discuss that this status is displayed because the properties are not actually used in Wikipedia pages. That is not the case for the example above - it can be verified that properties shown with this status are used in instances of Rakstnieka_infokaste infobox (using a Wikipedia dump from some months ago to avoid the possibility that they might be added after DBPedia stats were generated):

> bunzip2 -c lvwiki-20150702-pages-meta-current.xml.bz2 |  awk '/{{Rakstnieka infokaste/,/}}/' - | egrep "alma_mater\s*=\s*(\w+|[\['])" | wc
      15     112    1087

All 3 properties shown with this status (= blue color code) are used in infobox instances:

Related mailing list discussion: http://sourceforge.net/p/dbpedia/mailman/message/34541728/