| CPC G06Q 30/0253 (2013.01) | 18 Claims |
|
1. A system comprising:
one or more processors; and
one or more non-transitory computer-readable media storing computing instructions that, when executed on the one or more processors, perform:
receiving a taxonomy identifier corresponding to a taxonomy for a product;
determining one or more taxonomy embeddings based on the taxonomy identifier by:
collecting one or more taxonomy identifiers corresponding to one or more taxonomies for one or more products;
applying one or more respective filtering thresholds to each of the one or more taxonomy identifiers to create a modified set of the one or more taxonomy identifiers;
creating a training data set comprising the modified set of the one or more taxonomy identifiers; and
training a machine learning model in a first stage using the training data set to estimate internal parameters of the machine learning model to determine the one or more taxonomy embeddings, the one or more taxonomy embeddings representing at least a first level of the taxonomy and a second level of the taxonomy, the machine learning model to determine the one or more taxonomy embeddings based on training features to enable the machine learning model to determine words or patterns around a center word, the training features including a window size, a dictionary size, a context word vector, a center word vector, and a probability, the probability corresponding to a function between the context word vector and the center word vector;
reducing a number of the taxonomy embeddings in subsequent processing by the machine learning model by removing taxonomies that are below a threshold, wherein the threshold is a number of aggregate page views; and
mapping the one or more taxonomies, as reduced, to publisher placements to display the product within the one or more taxonomies, as reduced, on a graphical user interface (GUI).
|