US 12,608,738 B1Granted

Generating vector-based recommendations

AssigneeOracle International Corporation

InventorsVenkatakrishnan Gopalakrishnan, May Bich Nhi Lam, Diego Ceferino Torres Dho, Jan Sterba

Life of this patent

Estimating the term…

Loading patent…

Abstract

Techniques are described herein for performing generation of vector-based recommendations for entities. One or more embodiments address the challenge of generating recommendations for entities lacking sufficient recommendation generation input data. Unlike traditional collaborative filtering methods, various embodiments leverage content filtering techniques to make tailored and relevant recommendations for entities. By analyzing entity attributes and generating entity attribute vectors in an N-dimensional space, this system identifies the most suitable transactions by determining the closest neighbors in this space, thereby enhancing the accuracy and effectiveness of recommendations for entities associated with insufficient recommendation generation input data within the platform.

Claims20

Dependent claims

One or more non-transitory computer readable media comprising instructions which, when executed by one or more hardware processors, causes performance of operations comprising:

obtaining a plurality of sets of entity attributes, wherein each set of entity attributes, of the plurality of sets of entity attributes, corresponds to a respective entity of a plurality of entities;
aggregating the entity attributes in the plurality of the sets of entity attributes to generate a list of n entity attributes;
generating a plurality of entity attribute vectors for the plurality of entities, respectively, at least by:
for each particular entity of the plurality of entities, generating a particular entity attribute vector representing information for one or more particular entity attributes in a particular set of entity attributes associated with the particular entity, wherein the information for the one or more particular entity attributes comprises one or more of:
(a) locations of the one or more particular entity attributes within the list of n entity attributes,
(b) a term frequency (tf) of the one or more particular entity attributes in the particular set of entity attributes, or
(c) an inverse document frequency (idf) of the one or more particular entity attributes in the list of n entity attributes;
determining a target entity is associated with insufficient recommendation data for generating a recommendation for the target entity;
responsive to determining the target entity is associated with insufficient recommendation data, generating a recommendation for the target entity based on recommendation data of one or more additional entities at least by:
identifying an entity attribute vector for the target entity;
identifying k entity attribute vectors, of the plurality of entity attribute vectors, that are closest to the entity attribute vector;
and
identifying k entities, of the plurality of entities, that correspond to the k entity attribute vectors;
and
generating the recommendation for the target entity based on a set of recommendation data associated with the k entities.

12 more elements

A method comprising:

obtaining a plurality of sets of entity attributes, wherein each set of entity attributes, of the plurality of sets of entity attributes, corresponds to a respective entity of a plurality of entities;
aggregating the entity attributes in the plurality of the sets of entity attributes to generate a list of n entity attributes;
generating a plurality of entity attribute vectors for the plurality of entities, respectively, at least by:
for each particular entity of the plurality of entities, generating a particular entity attribute vector representing information for one or more particular entity attributes in a particular set of entity attributes associated with the particular entity, wherein the information for the one or more particular entity attributes comprises one or more of:
(a) locations of the one or more particular entity attributes within the list of n entity attributes,
(b) a term frequency (tf) of the one or more particular entity attributes in the particular set of entity attributes, or
(c) an inverse document frequency (idf) of the one or more particular entity attributes in the list of n entity attributes;
determining a target entity is associated with insufficient recommendation data for generating a recommendation for the target entity;
responsive to determining the target entity is associated with insufficient recommendation data, generating a recommendation for the target entity based on recommendation data of one or more additional entities at least by:
identifying an entity attribute vector for the target entity;
identifying k entity attribute vectors, of the plurality of entity attribute vectors, that are closest to the entity attribute vector;
and
identifying k entities, of the plurality of entities, that correspond to the k entity attribute vectors;
and
generating the recommendation for the target entity based on a set of recommendation data associated with the k entities.

12 more elements

A system comprising:

one or more processors;
and
memory storing instructions that, when executed by the one or more processors, cause the system to perform:
obtaining a plurality of sets of entity attributes,
wherein each set of entity attributes, of the plurality of sets of entity attributes, corresponds to a respective entity of a plurality of entities;
aggregating the entity attributes in the plurality of the sets of entity attributes to generate a list of n entity attributes;
generating a plurality of entity attribute vectors for the plurality of entities, respectively, at least by:
for each particular entity of the plurality of entities, generating a particular entity attribute vector representing information for one or more particular entity attributes in a particular set of entity attributes associated with the particular entity, wherein the information for the one or more particular entity attributes comprises one or more of:
(a) locations of the one or more particular entity attributes within the list of n entity attributes,
(b) a term frequency (tf) of the one or more particular entity attributes in the particular set of entity attributes, or
(c) an inverse document frequency (idf) of the one or more particular entity attributes in the list of n entity attributes;
determining a target entity is associated with insufficient recommendation data for generating a recommendation for the target entity;
responsive to determining the target entity is associated with insufficient recommendation data, generating a recommendation for the target entity based on recommendation data of one or more additional entities at least by:
identifying an entity attribute vector for the target entity;
identifying k entity attribute vectors, of the plurality of entity attribute vectors, that are closest to the entity attribute vector;
and
identifying k entities, of the plurality of entities, that correspond to the k entity attribute vectors;
and
generating the recommendation for the target entity based on a set of recommendation data associated with the k entities.

16 more elements

Dependencies are inferred from the claim text.