| CPC G06Q 30/0643 (2013.01) [G06N 20/00 (2019.01)] | 20 Claims |
|
1. A system comprising:
a processor; and
a non-transitory computer-readable medium storing computing instructions that, when executed on the processor, cause the processor to perform operations comprising:
training a visual similarity model to output one or more vector representations in a vector space of an item based on a training dataset, wherein:
the visual similarity model comprises an image similarity model and a textual similarity model;
the image similarity model is configured to be retrained, using a convolutional neural network, on a repository of digital images of items stored in a data store, wherein a last shared layer of the convolutional neural network outputs an image embedding of the item;
the textual similarity model is configured to use labels from at least non-image features of the item in a recurrent neural network to output a textual embedding of the item; and
the image embedding and the textual embedding comprise embeddings;
determining one or more nearest neighbors based on the embeddings nearest to a respective embedding in the vector space for the item; and
facilitating displaying the item in an order based upon the one or more nearest neighbors.
|