Milano housenumber import preparation is being preformed thru steps descripted here. Basically: refining input dataset, conflating and auditing. Below steps has been accomplished thru Openrefine (Cleaning to exporting), OSM_conflator tool and audit online tool.
Openrefine OSM_conflator
Visit https://dati.comune.milano.it/dataset/ds634-numeri-civici-coordinate and load CSV version in Openrefine
In Openrefine, apply file "operations"
In Openrefine, "export templating" using "export" file formatting.
Dataset features 60.000+ nodes, so you can define subsets thru Openrefine "facet" on MUNICIPIO field.
conflate -i .json -v -c preview_test.json -o test.osm profile-py
Here you can findf a sample audit for subset "MUNICIPIO 5" (approx 5.000 nodes) http://audit.osmz.ru/project/MI-M5/
Please, use http://overpass-turbo.eu/s/Ufx as reference, changing Municipio accoprdingly, then wget using Export "raw data directly from Overpass API" URL
To identify potential unmaches between addr and street http://overpass-turbo.eu/s/Dc1 wget -O OSM-MiHighWayNames.csv "http://overpass-api.de/api/interpreter?data=%5Bout%3Acsv%28%22name%22%29%5D%3Barea%5B%22name%22%3D%22Milano%22%5D%2D%3E%2Ea%3B%28way%5B%22highway%22%5D%28area%2Ea%29 %3B%29%3Bout%3B%0A"