JanZerebecki added a comment. >>! In T76373#799504, @Smalyshev wrote: > # How to handle qualifiers? As an example, such ones as point-in-time and > start-date/end-date. Most queries we'd do probably would be interested in > current values, but some may need past values.
For queries not explicitly asking for something besides the best, current consensus ranks are used to select the relevant statements. I.e. every time one selects a statement only use those that are rank preferred if a one of that rank exist for that property otherwise those of rank normal. Statements ranked as deprecated should be ignored unless specified otherwise in the query. See https://www.wikidata.org/wiki/Help:Ranking . But yes for e.g. for a query like who was the major of Berlin in 1984 you would need to look at statements that are 1) ranked preferred or normal and 2) that have a qualifier start-date which is <=1984 and 3) that do not have a qualifier end-date or whose end-date is >= 1984 . > ## With properties, it's easier because we can just stuff everything into a > vertex, but we need some standard format of storing non-current data. > ## With edges we have a potential to pollute the graph with myriad of data > that we will have to filter out on each query. We may distinguish them by > edge properties, but that could complicate many queries. Alternatively, we > could make special "past" edges - named with prefix, etc. - so that queries > interested in non-current data All this should not be decided based on ranks not qualifiers. However if you evaluate which statements (based on rank) are to be considered by default on import then if you still want to answer questions like the above historic one you still need a solution to be able to specifically query for preferred/normal/deprecated statements ignoring default. > ## Any such processing would require ad-hoc handing of specific properties, > such as point-in-time and start/end, to identify which data is current and > which is not. We may need to be careful with this since wrong handling of > this on initial import would require re-scan of the whole data set. No, as above. Queries not explicitly asking for something besides the best, current consensus do not need to understand qualifiers, only ranks. That should remove that possibility of incorrect interpretation during import. TASK DETAIL https://phabricator.wikimedia.org/T76373 REPLY HANDLER ACTIONS Reply to comment or attach files, or !close, !claim, !unsubscribe or !assign <username>. To: Smalyshev, JanZerebecki Cc: Smalyshev, Manybubbles, GWicke, JanZerebecki, jkroll, Wikidata-bugs, aude, daniel _______________________________________________ Wikidata-bugs mailing list [email protected] https://lists.wikimedia.org/mailman/listinfo/wikidata-bugs
