Patent
Atlas literature
Patent
US 12,175,504 B2Patent drawings and their descriptions. Click a drawing to enlarge it.
FIG. 1 is a diagram of one embodiment of a card payment processing system in which the disclosed embodiments may be implemented.
FIG. 2 illustrates a flow diagram of a process for unsu- pervised embeddings disentanglement using a GAN to gen- erate merchant recommendations to according to …
FIG. 3 is a diagram illustrating processing of the raw merchant embeddings and the raw user embeddings.
FIG. 4B is a diagram illustrating a principal component analysis (PCA) view of 200-dimension merchant embed- dings of Subway® and McDonalds® restaurants in San …
FIG. 5C illustrates an implementation of the discriminator according to one example embodiment.
FIG. 6 illustrates a mapping transition into 3D space from newly generated modified merchant embeddings of west states (lighter points) to target merchant …
FIG. 7 shows the results of the classification using the GAN without feature enhancement.
FIG. 8 shows the results of the classification using the GAN with feature enhancement.
FIG. 9 illustrates results of a location classification with the different alpha (a) values.
FIG. 10 illustrates results of a category classification with different a values.
FIG. 11 illustrates a Table showing results on nearest neighbor of the SF and NY dataset.
FIG. 12 illustrates a Table showing results on nearest neighbor of the East and West dataset.
Claims define the patent's legal scope. Independent claims stand alone; dependent claims (nested) narrow them. Click a claim to expand its dependents.
A computer-implemented method for training a recom-mendation system to provide merchant recommendations to a target user, comprising: receiving, by one or more processors, historical payment transaction records of transactions made between users and merchants; generating, by the one or more processors, from the historical payment transaction records a training set of raw merchant embeddings that represent the merchants and raw user embeddings that represent the users within an embedding space using a word embedding process, wherein the raw merchant embeddings include a plurality of embedded features from the historical payment transaction records including a location fea-ture, a price feature, and a number of transactions feature, wherein the raw merchant embeddings include source merchant embeddings from a home location of the target user and target merchant embeddings from a target location; training, by the one or more processors, a generative adversarial network (GAN) to generate modified mer-chant embeddings from the raw merchant embeddings, where the modified merchant embeddings have an effect of the location feature removed, the GAN com-prising a generator and a discriminator, the GAN operational for: receiving, by the generator, the source merchant embeddings as a set X of n embeddings X=x1,..., xn from the home location; receiving, by the discriminator, the target merchant embeddings as a set Y of m embeddings Y=y1,..., ym from the target location; and generating, by the generator, the modified merchant embeddings by applying a mapping function F to the set X of the source merchant embeddings, F (X)=F (x1),..., F (xn), wherein the generator uses F (X) to map the modified merchant embeddings to an embeddings space such that the discriminator cannot tell a difference between F (x1),..., F (xn) and the set Y of the target merchant embeddings; and discriminating, by the discriminator, between elements randomly sampled from F (X)=F (x1),..., F (xn) and Y to identify an origin of the modified merchant embeddings; continue, by the one or more processors, training the GAN until the discriminator cannot tell a difference between the target merchant embeddings and the modi-fied merchant embeddings, indicating the location fea-ture in the source merchant embeddings is mapped to the target merchant embeddings; and subsequent to the training and responsive to receiving a request for merchant recommendations in the target location for the target user, generating by the GAN a list of merchant rankings based on a new set of modi-fied merchant embeddings generated from a new set of payment transaction records made between the users and merchants, past preferences of the target user, and the target location to recommend merchants in the target location.
The method of claim 1, wherein the plurality of embedded features include a cuisine type feature.
The method of claim 1, wherein generating a list of merchant rankings based on the new set of modified mer-chant embeddings further comprises: receiving, by the one or more processors, the new set of payment transaction records of transactions made between the users and the merchants; generating, by the one or more processors, from the new set of payment transaction records new raw merchant embeddings and new raw user embeddings; performing, by the one or more processors, a domain-adversarial approach to map the plurality of embedded features from the source merchant embeddings to the target location together to a common embedding space using the GAN that performs a disentanglement pro-cess on the new raw merchant embeddings to remove the effect of the location feature by generating a new set of modified merchant embeddings that are free of the location feature; using, by the one or more processors, a preference model to determine new past preferences of the target user for the merchants based on the new modified merchant embeddings and the new raw user embeddings; automatically generating, by the one or more processors, the list of merchant rankings based on the new set of modified merchant embeddings, the new past prefer-ences of the target user, and the target location to recommend merchants in the target location; and providing, by the one or more processors, the list of merchant rankings to the target user.
The method of claim 1 wherein automatically gener-ating the list of merchant rankings further comprises: generating merchant preferences for the target user based on the modified merchant embeddings and the raw user embeddings; searching for local merchants located within proximity of a current location of the target user; and ranking the local merchants found within proximity of the target user based on the merchant preferences for the target user.
A merchant recommendation system, comprising: a transaction database storing historical payment transac-tion records; one or more processors having access to the transaction database; and a software component executed by the processor that is configured to: receive the historical payment transaction records made between users and merchants; generate from the historical payment transaction records a training set of raw merchant embeddings that repre-sent the merchants and raw user embeddings that represent the users within an embedding space using a word embedding process, wherein the raw merchant embeddings include a plurality of embedded features from the historical payment transaction records includ-ing a location feature, a price feature, and a number of transactions feature, wherein the raw merchant embed-dings include source merchant embeddings from a home location of a target user and target merchant embeddings from a target location; train a generative adversarial network (GAN) to generate modified merchant embeddings from the raw merchant embeddings, where the modified merchant embeddings have an effect of the location feature removed, the GAN comprising a generator and a discriminator, the GAN operational to: receive, by the generator, the source merchant embed-dings as a set X of n embeddings X=x1,..., xn from the home location; receive, by the discriminator, the target merchant embeddings as a set Y of m embeddings Y=y1,..., ym from the target location; and generate, by the generator, the modified merchant embeddings by applying a mapping function F to the set X of the source merchant embeddings, F (X)=F (x1),..., F (xn), wherein the generator uses F (X) to map the modified merchant embeddings to an embeddings space such that the discriminator cannot tell a difference between F (x1),..., F (xn) and the set Y of the target merchant embeddings; and discriminate, by the discriminator, between elements randomly sampled from F (X)=F (x1),..., F (xn) and Y to identify an origin of the modified merchant embeddings; continue, by the one or more processors, to train the GAN until the discriminator cannot tell a difference between the target merchant embeddings and the modified mer-chant embeddings, indicating the location feature in the source merchant embeddings is mapped to the target merchant embeddings; and subsequent to training and responsive to receiving a request for merchant recommendations in the target location for the target user, generating by, the GAN a list of merchant rankings based on a new set of modi-fied merchant embeddings generated from a new set of payment transaction records made between the users and merchants, and the target location to recommend merchants in the target location.
The system of claim 9, wherein the plurality of embedded features include a cuisine type feature.
The system of claim 9, the software component further configured to: receive, by the one or more processors, the new set of payment transaction records of transactions made between the users and the merchants; generate, by the one or more processors, from the new set of payment transaction records new raw merchant embeddings and new raw user embeddings; perform, by the one or more processors, a domain-adversarial approach to map the plurality of embedded features from the source merchant embeddings to the target location together to a common embedding space using the GAN that performs a disentanglement pro-cess on new the raw merchant embeddings to remove the effect of the location feature by generating a new set of modified merchant embeddings that are free of the location feature; use, by the one or more processors, a preference model to determine new past preferences of the target user for the merchants based on the new modified merchant embeddings and the new raw user embeddings; automatically generate, by the one or more processors, the list of merchant rankings based on the new set of modified merchant embeddings, the new past prefer-ences of the target user, and the target location to recommend merchants in the target location; and provide, by the one or more processors, the list of mer-chant rankings to the target user.
The system of claim 9, wherein when generating the list of merchant rankings, the software component is further configured to: generate merchant preferences for the target user based on the modified merchant embeddings and the raw user embeddings; search for local merchants located within proximity of a current location of the target user; and rank the local merchants found within proximity of the target user based on the merchant preferences for the target user.
A non-transitory computer-readable medium contain-ing program instructions for training a recommendation system to provide merchant recommendations to a target user, the program instructions for: receiving, by one or more processors, historical payment transaction records of transactions made between users and merchants; generating, by the one or more processors, from the historical payment transaction records a training set of raw merchant embeddings that represent the merchants and raw user embeddings that represent the users within an embedding space using a word embedding process, wherein the raw merchant embeddings include a plurality of embedded features from the historical payment transaction records including a location fea-ture, a price feature, and a number of transactions feature, wherein the raw merchant embeddings include source merchant embeddings from a home location of the target user and target merchant embeddings from a target location; training, by the one or more processors, a generative adversarial network (GAN) to generate modified mer-chant embeddings from the raw merchant embeddings, where the modified merchant embeddings have an effect of the location feature removed, the GAN com-prising a generator and a discriminator, the GAN operational for: 21 22 receiving, by the generator, the source merchant embeddings as a set X of n embeddings X=x1,..., xn from the home location; receiving, by the discriminator, the target merchant embeddings as a set Y of m embeddings Y=y1,..., ym from the target location; and generating, by the generator, the modified merchant embeddings by applying a mapping function F to the set X of the source merchant embeddings, F (X)=F (x1),..., F (xn), wherein the generator uses F (X) to map the modified merchant embeddings to an embeddings space such that the discriminator cannot tell a difference between F (x1),..., F (xn) and the set Y of the target merchant embeddings; and discriminating, by the discriminator, between elements randomly sampled from F (X)=F (x1),..., F (xn) and Y to identify an origin of the modified merchant embeddings; continuing, by the one or more processors, training the GAN until the discriminator cannot tell a difference between the target merchant embeddings and the modi-fied merchant embeddings, indicating the location fea-ture in the source merchant embeddings is mapped to the target merchant embeddings; and subsequent to the training and responsive to receiving a request for merchant recommendations in the target location for the target user, generating by the GAN a list of merchant rankings based on a new set of modi-fied merchant embeddings generated from a new set of payment transaction records made between users and merchants, and the target location to recommend mer-chants in the target location. ∗ ∗ ∗ ∗ ∗
Performance values and ranges asserted in the specification or claims.
| Property | Value | Material |
|---|---|---|
Temperature | 0–30 k | — |
Patents and literature cited by this patent (applicant and examiner references).
Cited patents · 11
Patent
Atlas literature
Patent
US 12,175,504 B2Patent drawings and their descriptions. Click a drawing to enlarge it.
FIG. 1 is a diagram of one embodiment of a card payment processing system in which the disclosed embodiments may be implemented.
FIG. 2 illustrates a flow diagram of a process for unsu- pervised embeddings disentanglement using a GAN to gen- erate merchant recommendations to according to …
FIG. 3 is a diagram illustrating processing of the raw merchant embeddings and the raw user embeddings.
FIG. 4B is a diagram illustrating a principal component analysis (PCA) view of 200-dimension merchant embed- dings of Subway® and McDonalds® restaurants in San …
FIG. 5C illustrates an implementation of the discriminator according to one example embodiment.
FIG. 6 illustrates a mapping transition into 3D space from newly generated modified merchant embeddings of west states (lighter points) to target merchant …
FIG. 7 shows the results of the classification using the GAN without feature enhancement.
FIG. 8 shows the results of the classification using the GAN with feature enhancement.
FIG. 9 illustrates results of a location classification with the different alpha (a) values.
FIG. 10 illustrates results of a category classification with different a values.
FIG. 11 illustrates a Table showing results on nearest neighbor of the SF and NY dataset.
FIG. 12 illustrates a Table showing results on nearest neighbor of the East and West dataset.
Claims define the patent's legal scope. Independent claims stand alone; dependent claims (nested) narrow them. Click a claim to expand its dependents.
A computer-implemented method for training a recom-mendation system to provide merchant recommendations to a target user, comprising: receiving, by one or more processors, historical payment transaction records of transactions made between users and merchants; generating, by the one or more processors, from the historical payment transaction records a training set of raw merchant embeddings that represent the merchants and raw user embeddings that represent the users within an embedding space using a word embedding process, wherein the raw merchant embeddings include a plurality of embedded features from the historical payment transaction records including a location fea-ture, a price feature, and a number of transactions feature, wherein the raw merchant embeddings include source merchant embeddings from a home location of the target user and target merchant embeddings from a target location; training, by the one or more processors, a generative adversarial network (GAN) to generate modified mer-chant embeddings from the raw merchant embeddings, where the modified merchant embeddings have an effect of the location feature removed, the GAN com-prising a generator and a discriminator, the GAN operational for: receiving, by the generator, the source merchant embeddings as a set X of n embeddings X=x1,..., xn from the home location; receiving, by the discriminator, the target merchant embeddings as a set Y of m embeddings Y=y1,..., ym from the target location; and generating, by the generator, the modified merchant embeddings by applying a mapping function F to the set X of the source merchant embeddings, F (X)=F (x1),..., F (xn), wherein the generator uses F (X) to map the modified merchant embeddings to an embeddings space such that the discriminator cannot tell a difference between F (x1),..., F (xn) and the set Y of the target merchant embeddings; and discriminating, by the discriminator, between elements randomly sampled from F (X)=F (x1),..., F (xn) and Y to identify an origin of the modified merchant embeddings; continue, by the one or more processors, training the GAN until the discriminator cannot tell a difference between the target merchant embeddings and the modi-fied merchant embeddings, indicating the location fea-ture in the source merchant embeddings is mapped to the target merchant embeddings; and subsequent to the training and responsive to receiving a request for merchant recommendations in the target location for the target user, generating by the GAN a list of merchant rankings based on a new set of modi-fied merchant embeddings generated from a new set of payment transaction records made between the users and merchants, past preferences of the target user, and the target location to recommend merchants in the target location.
The method of claim 1, wherein the plurality of embedded features include a cuisine type feature.
The method of claim 1, wherein generating a list of merchant rankings based on the new set of modified mer-chant embeddings further comprises: receiving, by the one or more processors, the new set of payment transaction records of transactions made between the users and the merchants; generating, by the one or more processors, from the new set of payment transaction records new raw merchant embeddings and new raw user embeddings; performing, by the one or more processors, a domain-adversarial approach to map the plurality of embedded features from the source merchant embeddings to the target location together to a common embedding space using the GAN that performs a disentanglement pro-cess on the new raw merchant embeddings to remove the effect of the location feature by generating a new set of modified merchant embeddings that are free of the location feature; using, by the one or more processors, a preference model to determine new past preferences of the target user for the merchants based on the new modified merchant embeddings and the new raw user embeddings; automatically generating, by the one or more processors, the list of merchant rankings based on the new set of modified merchant embeddings, the new past prefer-ences of the target user, and the target location to recommend merchants in the target location; and providing, by the one or more processors, the list of merchant rankings to the target user.
The method of claim 1 wherein automatically gener-ating the list of merchant rankings further comprises: generating merchant preferences for the target user based on the modified merchant embeddings and the raw user embeddings; searching for local merchants located within proximity of a current location of the target user; and ranking the local merchants found within proximity of the target user based on the merchant preferences for the target user.
A merchant recommendation system, comprising: a transaction database storing historical payment transac-tion records; one or more processors having access to the transaction database; and a software component executed by the processor that is configured to: receive the historical payment transaction records made between users and merchants; generate from the historical payment transaction records a training set of raw merchant embeddings that repre-sent the merchants and raw user embeddings that represent the users within an embedding space using a word embedding process, wherein the raw merchant embeddings include a plurality of embedded features from the historical payment transaction records includ-ing a location feature, a price feature, and a number of transactions feature, wherein the raw merchant embed-dings include source merchant embeddings from a home location of a target user and target merchant embeddings from a target location; train a generative adversarial network (GAN) to generate modified merchant embeddings from the raw merchant embeddings, where the modified merchant embeddings have an effect of the location feature removed, the GAN comprising a generator and a discriminator, the GAN operational to: receive, by the generator, the source merchant embed-dings as a set X of n embeddings X=x1,..., xn from the home location; receive, by the discriminator, the target merchant embeddings as a set Y of m embeddings Y=y1,..., ym from the target location; and generate, by the generator, the modified merchant embeddings by applying a mapping function F to the set X of the source merchant embeddings, F (X)=F (x1),..., F (xn), wherein the generator uses F (X) to map the modified merchant embeddings to an embeddings space such that the discriminator cannot tell a difference between F (x1),..., F (xn) and the set Y of the target merchant embeddings; and discriminate, by the discriminator, between elements randomly sampled from F (X)=F (x1),..., F (xn) and Y to identify an origin of the modified merchant embeddings; continue, by the one or more processors, to train the GAN until the discriminator cannot tell a difference between the target merchant embeddings and the modified mer-chant embeddings, indicating the location feature in the source merchant embeddings is mapped to the target merchant embeddings; and subsequent to training and responsive to receiving a request for merchant recommendations in the target location for the target user, generating by, the GAN a list of merchant rankings based on a new set of modi-fied merchant embeddings generated from a new set of payment transaction records made between the users and merchants, and the target location to recommend merchants in the target location.
The system of claim 9, wherein the plurality of embedded features include a cuisine type feature.
The system of claim 9, the software component further configured to: receive, by the one or more processors, the new set of payment transaction records of transactions made between the users and the merchants; generate, by the one or more processors, from the new set of payment transaction records new raw merchant embeddings and new raw user embeddings; perform, by the one or more processors, a domain-adversarial approach to map the plurality of embedded features from the source merchant embeddings to the target location together to a common embedding space using the GAN that performs a disentanglement pro-cess on new the raw merchant embeddings to remove the effect of the location feature by generating a new set of modified merchant embeddings that are free of the location feature; use, by the one or more processors, a preference model to determine new past preferences of the target user for the merchants based on the new modified merchant embeddings and the new raw user embeddings; automatically generate, by the one or more processors, the list of merchant rankings based on the new set of modified merchant embeddings, the new past prefer-ences of the target user, and the target location to recommend merchants in the target location; and provide, by the one or more processors, the list of mer-chant rankings to the target user.
The system of claim 9, wherein when generating the list of merchant rankings, the software component is further configured to: generate merchant preferences for the target user based on the modified merchant embeddings and the raw user embeddings; search for local merchants located within proximity of a current location of the target user; and rank the local merchants found within proximity of the target user based on the merchant preferences for the target user.
A non-transitory computer-readable medium contain-ing program instructions for training a recommendation system to provide merchant recommendations to a target user, the program instructions for: receiving, by one or more processors, historical payment transaction records of transactions made between users and merchants; generating, by the one or more processors, from the historical payment transaction records a training set of raw merchant embeddings that represent the merchants and raw user embeddings that represent the users within an embedding space using a word embedding process, wherein the raw merchant embeddings include a plurality of embedded features from the historical payment transaction records including a location fea-ture, a price feature, and a number of transactions feature, wherein the raw merchant embeddings include source merchant embeddings from a home location of the target user and target merchant embeddings from a target location; training, by the one or more processors, a generative adversarial network (GAN) to generate modified mer-chant embeddings from the raw merchant embeddings, where the modified merchant embeddings have an effect of the location feature removed, the GAN com-prising a generator and a discriminator, the GAN operational for: 21 22 receiving, by the generator, the source merchant embeddings as a set X of n embeddings X=x1,..., xn from the home location; receiving, by the discriminator, the target merchant embeddings as a set Y of m embeddings Y=y1,..., ym from the target location; and generating, by the generator, the modified merchant embeddings by applying a mapping function F to the set X of the source merchant embeddings, F (X)=F (x1),..., F (xn), wherein the generator uses F (X) to map the modified merchant embeddings to an embeddings space such that the discriminator cannot tell a difference between F (x1),..., F (xn) and the set Y of the target merchant embeddings; and discriminating, by the discriminator, between elements randomly sampled from F (X)=F (x1),..., F (xn) and Y to identify an origin of the modified merchant embeddings; continuing, by the one or more processors, training the GAN until the discriminator cannot tell a difference between the target merchant embeddings and the modi-fied merchant embeddings, indicating the location fea-ture in the source merchant embeddings is mapped to the target merchant embeddings; and subsequent to the training and responsive to receiving a request for merchant recommendations in the target location for the target user, generating by the GAN a list of merchant rankings based on a new set of modi-fied merchant embeddings generated from a new set of payment transaction records made between users and merchants, and the target location to recommend mer-chants in the target location. ∗ ∗ ∗ ∗ ∗
Performance values and ranges asserted in the specification or claims.
| Property | Value | Material |
|---|---|---|
Temperature | 0–30 k | — |
Patents and literature cited by this patent (applicant and examiner references).
Cited patents · 11
Patent
Atlas literature
Patent
US 12,175,504 B2Patent drawings and their descriptions. Click a drawing to enlarge it.
FIG. 1 is a diagram of one embodiment of a card payment processing system in which the disclosed embodiments may be implemented.
FIG. 2 illustrates a flow diagram of a process for unsu- pervised embeddings disentanglement using a GAN to gen- erate merchant recommendations to according to …
FIG. 3 is a diagram illustrating processing of the raw merchant embeddings and the raw user embeddings.
FIG. 4B is a diagram illustrating a principal component analysis (PCA) view of 200-dimension merchant embed- dings of Subway® and McDonalds® restaurants in San …
FIG. 5C illustrates an implementation of the discriminator according to one example embodiment.
FIG. 6 illustrates a mapping transition into 3D space from newly generated modified merchant embeddings of west states (lighter points) to target merchant …
FIG. 7 shows the results of the classification using the GAN without feature enhancement.
FIG. 8 shows the results of the classification using the GAN with feature enhancement.
FIG. 9 illustrates results of a location classification with the different alpha (a) values.
FIG. 10 illustrates results of a category classification with different a values.
FIG. 11 illustrates a Table showing results on nearest neighbor of the SF and NY dataset.
FIG. 12 illustrates a Table showing results on nearest neighbor of the East and West dataset.
Claims define the patent's legal scope. Independent claims stand alone; dependent claims (nested) narrow them. Click a claim to expand its dependents.
A computer-implemented method for training a recom-mendation system to provide merchant recommendations to a target user, comprising: receiving, by one or more processors, historical payment transaction records of transactions made between users and merchants; generating, by the one or more processors, from the historical payment transaction records a training set of raw merchant embeddings that represent the merchants and raw user embeddings that represent the users within an embedding space using a word embedding process, wherein the raw merchant embeddings include a plurality of embedded features from the historical payment transaction records including a location fea-ture, a price feature, and a number of transactions feature, wherein the raw merchant embeddings include source merchant embeddings from a home location of the target user and target merchant embeddings from a target location; training, by the one or more processors, a generative adversarial network (GAN) to generate modified mer-chant embeddings from the raw merchant embeddings, where the modified merchant embeddings have an effect of the location feature removed, the GAN com-prising a generator and a discriminator, the GAN operational for: receiving, by the generator, the source merchant embeddings as a set X of n embeddings X=x1,..., xn from the home location; receiving, by the discriminator, the target merchant embeddings as a set Y of m embeddings Y=y1,..., ym from the target location; and generating, by the generator, the modified merchant embeddings by applying a mapping function F to the set X of the source merchant embeddings, F (X)=F (x1),..., F (xn), wherein the generator uses F (X) to map the modified merchant embeddings to an embeddings space such that the discriminator cannot tell a difference between F (x1),..., F (xn) and the set Y of the target merchant embeddings; and discriminating, by the discriminator, between elements randomly sampled from F (X)=F (x1),..., F (xn) and Y to identify an origin of the modified merchant embeddings; continue, by the one or more processors, training the GAN until the discriminator cannot tell a difference between the target merchant embeddings and the modi-fied merchant embeddings, indicating the location fea-ture in the source merchant embeddings is mapped to the target merchant embeddings; and subsequent to the training and responsive to receiving a request for merchant recommendations in the target location for the target user, generating by the GAN a list of merchant rankings based on a new set of modi-fied merchant embeddings generated from a new set of payment transaction records made between the users and merchants, past preferences of the target user, and the target location to recommend merchants in the target location.
The method of claim 1, wherein the plurality of embedded features include a cuisine type feature.
The method of claim 1, wherein generating a list of merchant rankings based on the new set of modified mer-chant embeddings further comprises: receiving, by the one or more processors, the new set of payment transaction records of transactions made between the users and the merchants; generating, by the one or more processors, from the new set of payment transaction records new raw merchant embeddings and new raw user embeddings; performing, by the one or more processors, a domain-adversarial approach to map the plurality of embedded features from the source merchant embeddings to the target location together to a common embedding space using the GAN that performs a disentanglement pro-cess on the new raw merchant embeddings to remove the effect of the location feature by generating a new set of modified merchant embeddings that are free of the location feature; using, by the one or more processors, a preference model to determine new past preferences of the target user for the merchants based on the new modified merchant embeddings and the new raw user embeddings; automatically generating, by the one or more processors, the list of merchant rankings based on the new set of modified merchant embeddings, the new past prefer-ences of the target user, and the target location to recommend merchants in the target location; and providing, by the one or more processors, the list of merchant rankings to the target user.
The method of claim 1 wherein automatically gener-ating the list of merchant rankings further comprises: generating merchant preferences for the target user based on the modified merchant embeddings and the raw user embeddings; searching for local merchants located within proximity of a current location of the target user; and ranking the local merchants found within proximity of the target user based on the merchant preferences for the target user.
A merchant recommendation system, comprising: a transaction database storing historical payment transac-tion records; one or more processors having access to the transaction database; and a software component executed by the processor that is configured to: receive the historical payment transaction records made between users and merchants; generate from the historical payment transaction records a training set of raw merchant embeddings that repre-sent the merchants and raw user embeddings that represent the users within an embedding space using a word embedding process, wherein the raw merchant embeddings include a plurality of embedded features from the historical payment transaction records includ-ing a location feature, a price feature, and a number of transactions feature, wherein the raw merchant embed-dings include source merchant embeddings from a home location of a target user and target merchant embeddings from a target location; train a generative adversarial network (GAN) to generate modified merchant embeddings from the raw merchant embeddings, where the modified merchant embeddings have an effect of the location feature removed, the GAN comprising a generator and a discriminator, the GAN operational to: receive, by the generator, the source merchant embed-dings as a set X of n embeddings X=x1,..., xn from the home location; receive, by the discriminator, the target merchant embeddings as a set Y of m embeddings Y=y1,..., ym from the target location; and generate, by the generator, the modified merchant embeddings by applying a mapping function F to the set X of the source merchant embeddings, F (X)=F (x1),..., F (xn), wherein the generator uses F (X) to map the modified merchant embeddings to an embeddings space such that the discriminator cannot tell a difference between F (x1),..., F (xn) and the set Y of the target merchant embeddings; and discriminate, by the discriminator, between elements randomly sampled from F (X)=F (x1),..., F (xn) and Y to identify an origin of the modified merchant embeddings; continue, by the one or more processors, to train the GAN until the discriminator cannot tell a difference between the target merchant embeddings and the modified mer-chant embeddings, indicating the location feature in the source merchant embeddings is mapped to the target merchant embeddings; and subsequent to training and responsive to receiving a request for merchant recommendations in the target location for the target user, generating by, the GAN a list of merchant rankings based on a new set of modi-fied merchant embeddings generated from a new set of payment transaction records made between the users and merchants, and the target location to recommend merchants in the target location.
The system of claim 9, wherein the plurality of embedded features include a cuisine type feature.
The system of claim 9, the software component further configured to: receive, by the one or more processors, the new set of payment transaction records of transactions made between the users and the merchants; generate, by the one or more processors, from the new set of payment transaction records new raw merchant embeddings and new raw user embeddings; perform, by the one or more processors, a domain-adversarial approach to map the plurality of embedded features from the source merchant embeddings to the target location together to a common embedding space using the GAN that performs a disentanglement pro-cess on new the raw merchant embeddings to remove the effect of the location feature by generating a new set of modified merchant embeddings that are free of the location feature; use, by the one or more processors, a preference model to determine new past preferences of the target user for the merchants based on the new modified merchant embeddings and the new raw user embeddings; automatically generate, by the one or more processors, the list of merchant rankings based on the new set of modified merchant embeddings, the new past prefer-ences of the target user, and the target location to recommend merchants in the target location; and provide, by the one or more processors, the list of mer-chant rankings to the target user.
The system of claim 9, wherein when generating the list of merchant rankings, the software component is further configured to: generate merchant preferences for the target user based on the modified merchant embeddings and the raw user embeddings; search for local merchants located within proximity of a current location of the target user; and rank the local merchants found within proximity of the target user based on the merchant preferences for the target user.
A non-transitory computer-readable medium contain-ing program instructions for training a recommendation system to provide merchant recommendations to a target user, the program instructions for: receiving, by one or more processors, historical payment transaction records of transactions made between users and merchants; generating, by the one or more processors, from the historical payment transaction records a training set of raw merchant embeddings that represent the merchants and raw user embeddings that represent the users within an embedding space using a word embedding process, wherein the raw merchant embeddings include a plurality of embedded features from the historical payment transaction records including a location fea-ture, a price feature, and a number of transactions feature, wherein the raw merchant embeddings include source merchant embeddings from a home location of the target user and target merchant embeddings from a target location; training, by the one or more processors, a generative adversarial network (GAN) to generate modified mer-chant embeddings from the raw merchant embeddings, where the modified merchant embeddings have an effect of the location feature removed, the GAN com-prising a generator and a discriminator, the GAN operational for: 21 22 receiving, by the generator, the source merchant embeddings as a set X of n embeddings X=x1,..., xn from the home location; receiving, by the discriminator, the target merchant embeddings as a set Y of m embeddings Y=y1,..., ym from the target location; and generating, by the generator, the modified merchant embeddings by applying a mapping function F to the set X of the source merchant embeddings, F (X)=F (x1),..., F (xn), wherein the generator uses F (X) to map the modified merchant embeddings to an embeddings space such that the discriminator cannot tell a difference between F (x1),..., F (xn) and the set Y of the target merchant embeddings; and discriminating, by the discriminator, between elements randomly sampled from F (X)=F (x1),..., F (xn) and Y to identify an origin of the modified merchant embeddings; continuing, by the one or more processors, training the GAN until the discriminator cannot tell a difference between the target merchant embeddings and the modi-fied merchant embeddings, indicating the location fea-ture in the source merchant embeddings is mapped to the target merchant embeddings; and subsequent to the training and responsive to receiving a request for merchant recommendations in the target location for the target user, generating by the GAN a list of merchant rankings based on a new set of modi-fied merchant embeddings generated from a new set of payment transaction records made between users and merchants, and the target location to recommend mer-chants in the target location. ∗ ∗ ∗ ∗ ∗
Performance values and ranges asserted in the specification or claims.
| Property | Value | Material |
|---|---|---|
Temperature | 0–30 k | — |
Patents and literature cited by this patent (applicant and examiner references).
Cited patents · 11
Patent
Atlas literature
Patent
US 12,175,504 B2Patent drawings and their descriptions. Click a drawing to enlarge it.
FIG. 1 is a diagram of one embodiment of a card payment processing system in which the disclosed embodiments may be implemented.
FIG. 2 illustrates a flow diagram of a process for unsu- pervised embeddings disentanglement using a GAN to gen- erate merchant recommendations to according to …
FIG. 3 is a diagram illustrating processing of the raw merchant embeddings and the raw user embeddings.
FIG. 4B is a diagram illustrating a principal component analysis (PCA) view of 200-dimension merchant embed- dings of Subway® and McDonalds® restaurants in San …
FIG. 5C illustrates an implementation of the discriminator according to one example embodiment.
FIG. 6 illustrates a mapping transition into 3D space from newly generated modified merchant embeddings of west states (lighter points) to target merchant …
FIG. 7 shows the results of the classification using the GAN without feature enhancement.
FIG. 8 shows the results of the classification using the GAN with feature enhancement.
FIG. 9 illustrates results of a location classification with the different alpha (a) values.
FIG. 10 illustrates results of a category classification with different a values.
FIG. 11 illustrates a Table showing results on nearest neighbor of the SF and NY dataset.
FIG. 12 illustrates a Table showing results on nearest neighbor of the East and West dataset.
Claims define the patent's legal scope. Independent claims stand alone; dependent claims (nested) narrow them. Click a claim to expand its dependents.
A computer-implemented method for training a recom-mendation system to provide merchant recommendations to a target user, comprising: receiving, by one or more processors, historical payment transaction records of transactions made between users and merchants; generating, by the one or more processors, from the historical payment transaction records a training set of raw merchant embeddings that represent the merchants and raw user embeddings that represent the users within an embedding space using a word embedding process, wherein the raw merchant embeddings include a plurality of embedded features from the historical payment transaction records including a location fea-ture, a price feature, and a number of transactions feature, wherein the raw merchant embeddings include source merchant embeddings from a home location of the target user and target merchant embeddings from a target location; training, by the one or more processors, a generative adversarial network (GAN) to generate modified mer-chant embeddings from the raw merchant embeddings, where the modified merchant embeddings have an effect of the location feature removed, the GAN com-prising a generator and a discriminator, the GAN operational for: receiving, by the generator, the source merchant embeddings as a set X of n embeddings X=x1,..., xn from the home location; receiving, by the discriminator, the target merchant embeddings as a set Y of m embeddings Y=y1,..., ym from the target location; and generating, by the generator, the modified merchant embeddings by applying a mapping function F to the set X of the source merchant embeddings, F (X)=F (x1),..., F (xn), wherein the generator uses F (X) to map the modified merchant embeddings to an embeddings space such that the discriminator cannot tell a difference between F (x1),..., F (xn) and the set Y of the target merchant embeddings; and discriminating, by the discriminator, between elements randomly sampled from F (X)=F (x1),..., F (xn) and Y to identify an origin of the modified merchant embeddings; continue, by the one or more processors, training the GAN until the discriminator cannot tell a difference between the target merchant embeddings and the modi-fied merchant embeddings, indicating the location fea-ture in the source merchant embeddings is mapped to the target merchant embeddings; and subsequent to the training and responsive to receiving a request for merchant recommendations in the target location for the target user, generating by the GAN a list of merchant rankings based on a new set of modi-fied merchant embeddings generated from a new set of payment transaction records made between the users and merchants, past preferences of the target user, and the target location to recommend merchants in the target location.
The method of claim 1, wherein the plurality of embedded features include a cuisine type feature.
The method of claim 1, wherein generating a list of merchant rankings based on the new set of modified mer-chant embeddings further comprises: receiving, by the one or more processors, the new set of payment transaction records of transactions made between the users and the merchants; generating, by the one or more processors, from the new set of payment transaction records new raw merchant embeddings and new raw user embeddings; performing, by the one or more processors, a domain-adversarial approach to map the plurality of embedded features from the source merchant embeddings to the target location together to a common embedding space using the GAN that performs a disentanglement pro-cess on the new raw merchant embeddings to remove the effect of the location feature by generating a new set of modified merchant embeddings that are free of the location feature; using, by the one or more processors, a preference model to determine new past preferences of the target user for the merchants based on the new modified merchant embeddings and the new raw user embeddings; automatically generating, by the one or more processors, the list of merchant rankings based on the new set of modified merchant embeddings, the new past prefer-ences of the target user, and the target location to recommend merchants in the target location; and providing, by the one or more processors, the list of merchant rankings to the target user.
The method of claim 1 wherein automatically gener-ating the list of merchant rankings further comprises: generating merchant preferences for the target user based on the modified merchant embeddings and the raw user embeddings; searching for local merchants located within proximity of a current location of the target user; and ranking the local merchants found within proximity of the target user based on the merchant preferences for the target user.
A merchant recommendation system, comprising: a transaction database storing historical payment transac-tion records; one or more processors having access to the transaction database; and a software component executed by the processor that is configured to: receive the historical payment transaction records made between users and merchants; generate from the historical payment transaction records a training set of raw merchant embeddings that repre-sent the merchants and raw user embeddings that represent the users within an embedding space using a word embedding process, wherein the raw merchant embeddings include a plurality of embedded features from the historical payment transaction records includ-ing a location feature, a price feature, and a number of transactions feature, wherein the raw merchant embed-dings include source merchant embeddings from a home location of a target user and target merchant embeddings from a target location; train a generative adversarial network (GAN) to generate modified merchant embeddings from the raw merchant embeddings, where the modified merchant embeddings have an effect of the location feature removed, the GAN comprising a generator and a discriminator, the GAN operational to: receive, by the generator, the source merchant embed-dings as a set X of n embeddings X=x1,..., xn from the home location; receive, by the discriminator, the target merchant embeddings as a set Y of m embeddings Y=y1,..., ym from the target location; and generate, by the generator, the modified merchant embeddings by applying a mapping function F to the set X of the source merchant embeddings, F (X)=F (x1),..., F (xn), wherein the generator uses F (X) to map the modified merchant embeddings to an embeddings space such that the discriminator cannot tell a difference between F (x1),..., F (xn) and the set Y of the target merchant embeddings; and discriminate, by the discriminator, between elements randomly sampled from F (X)=F (x1),..., F (xn) and Y to identify an origin of the modified merchant embeddings; continue, by the one or more processors, to train the GAN until the discriminator cannot tell a difference between the target merchant embeddings and the modified mer-chant embeddings, indicating the location feature in the source merchant embeddings is mapped to the target merchant embeddings; and subsequent to training and responsive to receiving a request for merchant recommendations in the target location for the target user, generating by, the GAN a list of merchant rankings based on a new set of modi-fied merchant embeddings generated from a new set of payment transaction records made between the users and merchants, and the target location to recommend merchants in the target location.
The system of claim 9, wherein the plurality of embedded features include a cuisine type feature.
The system of claim 9, the software component further configured to: receive, by the one or more processors, the new set of payment transaction records of transactions made between the users and the merchants; generate, by the one or more processors, from the new set of payment transaction records new raw merchant embeddings and new raw user embeddings; perform, by the one or more processors, a domain-adversarial approach to map the plurality of embedded features from the source merchant embeddings to the target location together to a common embedding space using the GAN that performs a disentanglement pro-cess on new the raw merchant embeddings to remove the effect of the location feature by generating a new set of modified merchant embeddings that are free of the location feature; use, by the one or more processors, a preference model to determine new past preferences of the target user for the merchants based on the new modified merchant embeddings and the new raw user embeddings; automatically generate, by the one or more processors, the list of merchant rankings based on the new set of modified merchant embeddings, the new past prefer-ences of the target user, and the target location to recommend merchants in the target location; and provide, by the one or more processors, the list of mer-chant rankings to the target user.
The system of claim 9, wherein when generating the list of merchant rankings, the software component is further configured to: generate merchant preferences for the target user based on the modified merchant embeddings and the raw user embeddings; search for local merchants located within proximity of a current location of the target user; and rank the local merchants found within proximity of the target user based on the merchant preferences for the target user.
A non-transitory computer-readable medium contain-ing program instructions for training a recommendation system to provide merchant recommendations to a target user, the program instructions for: receiving, by one or more processors, historical payment transaction records of transactions made between users and merchants; generating, by the one or more processors, from the historical payment transaction records a training set of raw merchant embeddings that represent the merchants and raw user embeddings that represent the users within an embedding space using a word embedding process, wherein the raw merchant embeddings include a plurality of embedded features from the historical payment transaction records including a location fea-ture, a price feature, and a number of transactions feature, wherein the raw merchant embeddings include source merchant embeddings from a home location of the target user and target merchant embeddings from a target location; training, by the one or more processors, a generative adversarial network (GAN) to generate modified mer-chant embeddings from the raw merchant embeddings, where the modified merchant embeddings have an effect of the location feature removed, the GAN com-prising a generator and a discriminator, the GAN operational for: 21 22 receiving, by the generator, the source merchant embeddings as a set X of n embeddings X=x1,..., xn from the home location; receiving, by the discriminator, the target merchant embeddings as a set Y of m embeddings Y=y1,..., ym from the target location; and generating, by the generator, the modified merchant embeddings by applying a mapping function F to the set X of the source merchant embeddings, F (X)=F (x1),..., F (xn), wherein the generator uses F (X) to map the modified merchant embeddings to an embeddings space such that the discriminator cannot tell a difference between F (x1),..., F (xn) and the set Y of the target merchant embeddings; and discriminating, by the discriminator, between elements randomly sampled from F (X)=F (x1),..., F (xn) and Y to identify an origin of the modified merchant embeddings; continuing, by the one or more processors, training the GAN until the discriminator cannot tell a difference between the target merchant embeddings and the modi-fied merchant embeddings, indicating the location fea-ture in the source merchant embeddings is mapped to the target merchant embeddings; and subsequent to the training and responsive to receiving a request for merchant recommendations in the target location for the target user, generating by the GAN a list of merchant rankings based on a new set of modi-fied merchant embeddings generated from a new set of payment transaction records made between users and merchants, and the target location to recommend mer-chants in the target location. ∗ ∗ ∗ ∗ ∗
Performance values and ranges asserted in the specification or claims.
| Property | Value | Material |
|---|---|---|
Temperature | 0–30 k | — |
Patents and literature cited by this patent (applicant and examiner references).
Cited patents · 11
Cited non-patent literature · 7
Cited non-patent literature · 7
Cited non-patent literature · 7
Cited non-patent literature · 7
