Table 2 Real world datasets

Property sources

a. Data extracted using OXPATH [42]. 40 real estate sources, 6 to 16 attributes each, an average of 11 attributes per source, 7.8k tuples in total. Used as source data

b. Wadar [43]. 10 sources, 18 varying attributes each, 3.2k tuples. Used as source data

c. Price Paid Data (www.gov.uk). Property sales in England and Wales for 2017. 16 attributes, 130 properties sold in Oxford and Manchester, sampled from 461.2k tuples. Used as example data

English indices of deprivation

6 attributes, 62.3k tuples with indices for Manchester and Oxford, from http://www.gov.uk. Used as source data

Open addresses UK data

11 attributes, 41.7k tuples for Manchester and Oxford (postcode sectors M and O), from openaddressesuk.org. Used as reference data