| CPC G06F 16/285 (2019.01) [G06F 9/4881 (2013.01); G06F 11/3409 (2013.01)] | 20 Claims |
|
1. A computer-implemented method for matching entities, the computer-implemented method comprising:
receiving information identifying a plurality of entities comprising source entities and target entities, wherein an entity is associated with one or more resources;
for each of the plurality of entities, generating a vector representation;
storing vector representations of entities in a vector database, wherein the vector database computes vector distances between pairs of entities to identify similar entities using cosine similarity;
clustering, using the vector database, target entities to generate a plurality of clusters based on vector distances between pairs of target entities, each cluster comprising target entities that are close to other target entities within the cluster compared to target entities of remaining clusters, wherein each cluster has a representative target entity; and
optimizing an aggregate score across the plurality of entities using a plurality of queues, each queue associated with a source entity and including one or more target entities, comprising:
ranking representative target entities of clusters based on vector distances between a particular source entity and the representative target entities,
identifying a representative target entity based on the ranking,
identifying a cluster having the representative target entity, of target entities based on a distance between the source entities and entities of the cluster of target entities,
selecting a subset of target entities from the identified cluster using a machine learning based model trained to determine a score between an input source entity and target entity, and
modifying one or more queues from the plurality of queues to optimize the aggregate score.
|