US 12,136,489 B2
Predictive recommendation systems using compliance profile data objects
Karim Mahmoud Mohamed Moustafa, Dublin (IE); Harutyun Shahumyan, Dublin (IE); Gevorg Poghosyan, Dublin (IE); Kieran M. Cooney, Dublin (IE); and Lisa E. Walsh, Dublin (IE)
Assigned to Optum Services (Ireland) Limited, Dublin (IE)
Filed by Optum Services (Ireland) Limited, Dublin (IE)
Filed on Nov. 16, 2021, as Appl. No. 17/528,041.
Prior Publication US 2023/0154596 A1, May 18, 2023
Int. Cl. G06Q 30/02 (2023.01); G16H 20/00 (2018.01); G16H 40/20 (2018.01); G16H 50/70 (2018.01)
CPC G16H 40/20 (2018.01) [G16H 20/00 (2018.01); G16H 50/70 (2018.01)] 19 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
generating, by one or more processors and using a natural language processing (NLP) model, a guideline vector for a service need condition from reference data associated with the service need condition, wherein the service need condition is one of a plurality of service need conditions;
generating, by the one or more processors, a compliance profile vector for a provider entity, wherein the compliance profile vector comprises a plurality of compliance scores for the provider entity with respect to the plurality of service need conditions, and wherein generating the compliance profile vector comprises:
for the service need condition: (i) retrieving one or more procedural record data objects each describing a treatment provided by the provider entity for the service need condition, and (ii) determining a compliance score from the plurality of compliance scores for the service need condition using the one or more procedural record data objects and the guideline vector, and
determining the compliance profile vector based at least in part on the compliance score;
selecting, by the one or more processors, the provider entity from a plurality of provider entities based on a vector comparison between the compliance profile vector and a plurality of compliance profile vectors respectively corresponding to the plurality of provider entities; and
performing an automated recommendation-based action based at least in part on the selected provider entity.