Given a set of 1 to 3 (maybe 4) words return a map/dict of tuples with city,region and country information (each with label and Wikidata-Q ID) and the probability of matching (ordering is good enough for a start). For the time being the probability may be calculated by the population later we'll use the probability distribution of conference corpus entries that have been successfully matched.
Comparison of population is done on the "lowest" level. E.g. Athens, Greece is preferred to Athens, Georgia, USA since Athens, Greece has a pop of 600 thousand while Athens Georgia population is 5 times lower.
Given a set of 1 to 3 (maybe 4) words return a map/dict of tuples with city,region and country information (each with label and Wikidata-Q ID) and the probability of matching (ordering is good enough for a start). For the time being the probability may be calculated by the population later we'll use the probability distribution of conference corpus entries that have been successfully matched.
Comparison of population is done on the "lowest" level. E.g. Athens, Greece is preferred to Athens, Georgia, USA since Athens, Greece has a pop of 600 thousand while Athens Georgia population is 5 times lower.