The Bundestag (the German parliament) published basic information („Stammdaten“) of all its members, and I’m trying to import it into Wikidata.
This import is also tracked and discussed on Wikidata – see Wikidata:Dataset Imports/Basic data about the members of the Bundestag.
I converted the data into a more useful format (JSON instead of XML) and started to extract useful bits (names, biographical information) into separate files.
The next step was reconciling the MdBs with Wikidata items. I imported the names and dates/places of birth/death into OpenRefine and had it automatically match them to Wikidata items; as it turned out, the most useful results were obtained when matching just on the names and ignoring the other data, since otherwise the extra data would lower OpenRefine’s confidence in the match below the auto-match threshold. 3197 entries could be matched automatically, and the remaining 876 were matched manually to Wikidata items by @akaes and me. The result was the file IDs containing a mapping from the ID in the Bundestag’s database to the Wikidata item ID.
After that, the dates of birth and dates of death of all MdBs were imported into Wikidata, using QuickStatements’ CSV mode.
(More detailed descriptions of these steps are available in the commit messages – see the log.)
We have a tab-separated file from Bundestag ID to Wikidata item ID, plus some supplementary information. Dates of birth and death have been imported, everything else remains to be done.
Decide what to export and how to do it. Matching places of birth/death or other biographical information sounds like a lot of hassle, so I think I’d prefer importing the parliamentary terms next.