US 12,299,039 B2
Graph based predictive inferences for domain taxonomy
Savindra Singh, Uttar Pradesh (IN); Neelabh Mishra, New Delhi (IN); Sanchit Kumar, Uttar Pradesh (IN); and Sana Zehra, New Delhi (IN)
Assigned to OPTUM, INC., Minnetonka, MN (US)
Filed by Optum, Inc., Minnetonka, MN (US)
Filed on Mar. 8, 2023, as Appl. No. 18/180,376.
Prior Publication US 2024/0303276 A1, Sep. 12, 2024
Int. Cl. G06F 17/00 (2019.01); G06F 7/00 (2006.01); G06F 16/901 (2019.01); G06F 16/906 (2019.01)
CPC G06F 16/9024 (2019.01) [G06F 16/906 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
generating, by one or more processors, a network graph for an entity class based on a plurality of interaction data objects corresponding to a plurality of entities of the entity class, wherein:
(i) the plurality of interaction data objects comprises one or more interaction codes,
(ii) the network graph comprises a plurality of nodes and a plurality of edges,
(iii) a first node of the plurality of nodes corresponds to a first interaction code of at least one of the plurality of interaction data objects,
(iv) a first edge of the plurality of edges connects a node pair comprising the first node and a second node of the plurality of nodes that is associated with a first interaction data object of the plurality of interaction data objects,
(v) the first node is associated with a node weight indicative of a code frequency of the first interaction code in the plurality of interaction data objects, and
(vi) the first edge is associated with an edge weight indicative of a code pair frequency for two interaction codes corresponding to the node pair in the plurality of interaction data objects;
generating, by the one or more processors, a relevant network graph by removing one or more outlier nodes of the plurality of nodes;
generating, by the one or more processors, one or more node clusters based on the relevant network graph, the node weight, and the edge weight;
generating, by the one or more processors, a class taxonomy for the entity class based on the one or more node clusters, wherein the class taxonomy is indicative of a plurality of homogenous subclasses of the entity class; and
training, by the one or more processors, one or more machine learning models based at least in part on the relevant network graph and the class taxonomy.