US 11,726,980 B2
Auto detection of matching fields in entity resolution systems
Neeraj Ramkrishna Singh, Bangalore (IN); Abhishek Seth, Deoband (IN); Soma Shekar Naganna, Bangalore (IN); and Shettigar Parkala Srinivas, Bangalore (IN)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Jul. 14, 2020, as Appl. No. 16/928,361.
Prior Publication US 2022/0019571 A1, Jan. 20, 2022
Int. Cl. G06F 16/23 (2019.01); G06F 16/28 (2019.01)
CPC G06F 16/2365 (2019.01) [G06F 16/288 (2019.01)] 18 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
prior to undergoing matching of any data among future payload data in an entity resolution system:
determining a potential new attribute field for use in matching data among the future payload data in the entity resolution system;
determining a matching function for the potential new attribute field, wherein the matching function corresponds to particular match criteria associated with the potential new attribute field;
obtaining a score list for a reference data set that is distinct from the future payload data, wherein the score list for the reference data set includes expected matching outcomes and associated rates of false positives and false negatives when using the potential new attribute field in a matching process of known matched pairs in the reference data set;
determining a correlation of an attribute score for the potential new attribute field with the score list for the reference data set; and
selecting the potential new attribute field for use in matching the future payload data in the entity resolution system based, at least in part, on the correlation of the attribute score with the score list for the reference data set.