Detailed Action
Status of Claims
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
This Action is in reply to the Amendment filed on 3/9/2026. Claims 1-16 & 21-24 are currently pending and have been examined. Claims 1, 10 & 21 have been amended. The claim objections have been overcome by amendment.
Claim Rejection - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-16 & 21-24 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
First, it is determined whether the claims are directed to a statutory category of invention. In the instant case, claims 1-9 are directed to a machine, claims 10-16 are directed to a process, and claims 21-24 are directed to an article of manufacture. Therefore, claims 1-16 and 21-24 are directed to statutory subject matter under Step 1 as described in MPEP 2106 (Step 1: YES).
The claims are then analyzed to determine whether the claims are directed to a judicial exception. In determining whether the claims are directed to a judicial exception, the claims are analyzed to evaluate whether the claims recite a judicial exception (Prong One of Step 2A), as well as analyzed to evaluate whether the claims recite additional elements that integrate the judicial exception into a practical application of the judicial exception (Prong Two of Step 2A).
Claims 1, 10, and 21 recite at least the following limitations that are believed to recite an abstract idea:
receive a request for information including a first context for the information, wherein the first context comprises one or more of an account page, a home page, a cart view page, a product page, a program enrollment page, an order specific page, a checkout page, or a search page;
receive a user identifier associated with a set of user features based on the request;
receive the set of user features based on the user identifier;
receive a set of assets each including a set of asset features based on the first context;
select a context-specific asset prediction process from a plurality of context-specific asset prediction processes based on the first context, wherein each of the plurality of context-specific asset prediction processes is generated using data to adjust corresponding parameters to minimize a cost value characterizing a likelihood of interaction between users and assets, and wherein the data comprises user-preference features characterizing user interactions of a plurality of users with a plurality of pages, asset features characterizing a plurality of assets of the plurality of pages, and user-asset features characterizing interactions between the plurality of users represented by the user-preference features and the plurality of assets represented by the asset features;
determine, using the context-specific asset prediction process, an amount of engagement associated with each of the assets in the set of assets;
generate a set of predicted assets using the context-specific asset prediction process, wherein the context-specific asset prediction process comprises a procedure that receives the set of user features and the set of asset features for each asset in the set of assets and outputs the set of predicted assets, and wherein the context-specific asset prediction process maximizes a likelihood of engagement for the set of predicted assets for the first context;
generate a page of the first context including a predetermined number of assets selected from the set of predicted assets generated using the context-specific asset prediction process in descending ranked order based on the determined amount of engagement;
transmit the page in response to the request to display the page;
receive user input directed to a displayed first asset from the predetermined number of assets;
based on the user input, update the page to include a second asset; and
transmit the updated page to display the page with the second asset.
The above limitations recite the concept of user-specific asset recommendations. These limitations, under their broadest reasonable interpretation, fall within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas, enumerated in MPEP 2106, in that they recite commercial interactions, e.g. sales activities/behaviors, and managing personal behavior or relationships or interactions between people, e.g., following rules or instructions. Accordingly, under Prong One of Step 2A, claims 1-16 and 21-24 an abstract idea (Step 2A, Prong One: YES).
Prong Two of Step 2A is the next step in the eligibility analyses and looks at whether the abstract idea is integrated into a practical application. This requires an additional element or combination of additional elements in the claims to apply, rely on, or user the judicial exception in a manner that imposes a meaningful limit on the judicial exception, such that the claim is more than a drafting effort designed to monopolize the exception.
In this instance, the claims recite the additional elements of:
A system comprising a processor and a non-transitory memory storing processor-executed instructions
Interfaces
A database
Asset prediction models trained using training data
A machine learning model
Automatic steps
A computer-implemented method
A non-transitory computer-readable medium having instructions stored thereon that, when executed by at least one process, cause at least one device to perform operations
However, these elements do not amount to an improvement in the functioning of a computer or any other technology or technical field; apply the judicial exception with, or by use of, a particular machine; or apply or use the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is more than a drafting effort to monopolize the exception.
In addition, the recitations are recited at a high level of generality and also do not amount to an improvement in the functioning of a computer or any other technology or technical field; apply the judicial exception with, or by use of, a particular machine; or apply or use the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is more than a drafting effort to monopolize the exception.
The dependent claims also fail to recite elements which amount to an improvement in the functioning of a computer or any other technology or technical field; apply the judicial exception with, or by use of, a particular machine; or apply or use the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is more than a drafting effort to monopolize the exception. For example, claims 3, 6-7, 9, 12-15, and 23-24 are directed to the abstract idea itself and do not amount to an integration according to any one of the considerations above. As for claims 2, 4-5, 8, 11, 16, and 22, these claims are similar to the independent claims except that they recite the further additional elements of a random forest model, clicks, and a click rate. These additional elements are recited at a high level of generality and also do not amount to an improvement in the functioning of a computer or any other technology or technical field; apply the judicial exception with, or by use of, a particular machine; or apply or use the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is more than a drafting effort to monopolize the exception. Therefore the dependent claims do not create an integration for the same reasons.
Step 2B is the next step in the eligibility analyses and evaluates whether the claims recite additional elements that amount to an inventive concept (i.e., “significantly more”) than the recited judicial exception. According to Office procedure, revised Step 2A overlaps with Step 2B, and thus, many of the considerations need not be re-evaluated in Step 2B because the answer will be the same.
In Step 2A, several additional elements were identified as additional limitations:
A system comprising a processor and a non-transitory memory storing processor-executed instructions
Interfaces
A database
Asset prediction models trained using training data
A machine learning model
Automatic steps
A computer-implemented method
A non-transitory computer-readable medium having instructions stored thereon that, when executed by at least one process, cause at least one device to perform operations
These additional limitations, including the limitations in the dependent claims, do not amount to an inventive concept because they were already analyzed under Step 2A and did not amount to a practical application of the abstract idea. Therefore, the claims lack one or more limitations which amount to an inventive concept in the claims.
For these reasons, the claims are rejected under 35 U.S.C. 101.
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
Claim Rejection – 35 USC § 103
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or non-
obviousness.
Claims 1-3, 6, 8-12, 14, 16, 21-22, and 24 are rejected under 35 U.S.C. 103 as being unpatentable over Wang et al (US 20160140622 A1), hereinafter Wang, in view of Nath et al (US 20230342831 A1), hereinafter Nath, and further in view of Agarwal et al (US 20100217648 A1), hereinafter Agarwal
Regarding Claim 1, Wang discloses a system, comprising: a processor; and a non-transitory memory storing instructions (Wang: [0124-0126]) that, when executed, cause the processor to:
receive a request for an interface [page] including a first context for the interface, wherein the first context comprises one or more of an account page, a home page, a cart view page, a product page, a program enrollment page, an order specific page, a checkout page, or a search page (Wang: “in response to a request from the user 406 … the social networking system 202 may present a storefront page” [0087] – “detecting a user request to access the merchant page” [Claim 2] – See Figures 4, 6.);
receive a user identifier associated with a set of user features [profile information] based on the request (Wang: “The affinity score engine 206 can utilize user profile information associated with a user to determine product affinity scores for products” [0056] – “User profile information can include an age, a gender, … preferences (e.g., product preferences), likes, dislikes, and/or attitudes of a user” [0055]);
receive, from a database, the set of user features based on the user identifier (Wang: “ a social networking system 1302 can comprise one or more data stores. …the social networking system 1302 can store a social graph comprising user nodes, … Each user node can comprise one or more data objects corresponding to information associated with or describing a user” [0143] – “create and store in the social networking system a user profile associated with the user.” [0131]);
receive from the database, a set of assets [products] each including a set of asset features [product information] based on the first context (Wang: “automatically identify and obtain product information for each of products 602” [0089] – “product information can include text associated with a product (e.g., a title of the product, a description for the product, specifications for the product, a name of a manufacturer of the product” [0032] – “associating, based on a plurality of social networking communications by a merchant, a plurality of products with a merchant page.” [0115] – “store product information in another locating, such as in a remote database.” [0050]);
determine, using a context-specific asset prediction model [affinity score engine], an amount of engagement [e.g. sales volume, number of likes] associated with each of the assets in the set of assets (Wang: “the affinity score engine 206 can determine a product affinity score or a set of product affinity scores based on social trends. Social trends can include overall popularity, rating, and/or ranking of a product. …social trends can include sales volume, sales rate, and sales history of a product. …products associated with a large number of “likes” by friends of a user can be given a more favorable product affinity score ” [0058-0059] );
generate a set of predicted assets using the context-specific asset prediction model (Wang: “determine a product affinity score for each product and then rank the products according to their product affinity scores” [0028] – “the affinity score engine 206 can calculate a product affinity score based on a correlation of one or more product characteristics, product features, and/or pieces of product information to user profile information associated with a user.” [0056] – “a user's product affinity score for a product can represent a predicted level of the user's interest in the product” [0033]),
wherein the context-specific asset prediction model receives the set of user features and the set of asset features for each asset in the set of assets and outputs the set of predicted assets (Wang: “the affinity score engine 206 can determine a product affinity score based on one or more of the following: user profile information, user activity, social trends, merchant input, any other suitable factors, or a combination thereof. ” [0054] – “the affinity score engine 206 can calculate a product affinity score based on a correlation of one or more product characteristics, product features, and/or pieces of product information to user profile information associated with a user. ” [0056]), and
wherein the context-specific asset prediction model maximizes a likelihood of engagement for the set of predicted assets for the first context (Wang: “the affinity score engine 206 can use a user's product preferences to identify products that the user is likely to prefer or not prefer based on the user product preferences,” [0056] – “rank products having a favorable product affinity score above products having an unfavorable product affinity score. By ranking products for a particular user, the customization manager 208 can prioritize products according to the user's potential interest in each product when customizing the merchant page.” [0064] – “products 602 that are most likely to be of interest to user 406 may be presented to user 406 first or at the top” [0090]);
automatically generate an interface of the first context including a predetermined number of assets selected from the set of predicted assets generated using the context-specific asset prediction model in descending ranked order based on the determined amount of engagement (Wang: “customize a display of products on a merchant page based on a ranking of the products based on their product affinity scores … if the shopping window can only include of four (4) products. The customization manager 208 can use the rank to identify the top four (4) ranked products” [0075-0076] – “Like the shopping portion 416, the social networking system 202 can customize the storefront page 600 for the user 406. For example, the social networking system 202 can customize the product page 602 to display the products 602 in an order that corresponds to product affinity scores for the products 602 (e.g., as calculated by customization manager 208). … products 602 that are most likely to be of interest to user 406 may be presented to user 406 first or at the top” [0090] – See Figures 4, 6.);
transmit the interface in response to the request to display the interface (Wang: “customize a display of products on a merchant page based on a ranking of the products based on their product affinity scores … if the shopping window can only include of four (4) products. The customization manager 208 can use the rank to identify the top four (4) ranked products” [0075-0076] – See Figures 4-6);
receive user input directed to a displayed first asset from the predetermined number of assets (Wang: “A user 406 can interact with a product within the shopping portion 416 to obtain additional information regarding the product and/or to purchase the product. For example, upon selecting a product within the shopping portion 416, the social networking system 202 can present additional product information associated with the product within a product page” [0084] – “the storefront page 600 includes … multiple selectable options 606 for the user to “See Details” for the product or “Buy” the product.” [0088] – See Figures 4, 6.);
based on the user input, update the interface to include a second asset (Wang: “based on user interactions, updated social information, and/or merchant changes, the affinity score engine 206 can send updated product affinity scores to the customization manager 208, and the customization manager 208 can change the display of products within the shopping portion based on the updated product affinity scores.” [0069]); and
transmit the updated interface to display the interface with the second asset (Wang: “based on user interactions, updated social information, and/or merchant changes, the affinity score engine 206 can send updated product affinity scores to the customization manager 208, and the customization manager 208 can change the display of products within the shopping portion based on the updated product affinity scores.” [0069] – “The social networking system 202 can present the product page 502 as a new page or as an overlay to the merchant page 402” [0084]).
While Wang teaches iterative updating the output of the engine based on new information/feedback [0069], it does not specifically teach selecting a context-specific asset prediction model from a plurality of context-specific asset prediction models based on the first context, wherein each of the plurality of context-specific asset prediction models is trained using training data to adjust corresponding parameters to minimize a cost value characterizing a likelihood of interaction between users and assets, and wherein the training data comprises user-preference features characterizing user interactions of a plurality of users with a plurality of interfaces, asset features characterizing a plurality of assets of the plurality of interfaces, and user-asset features characterizing interactions between the plurality of users represented by the user-preference features and the plurality of assets represented by the asset features; and that the context-specific asset prediction model comprises a machine learning model.
However, Nath teaches a system for providing recommendations [Abstract] including:
selecting a context-specific asset prediction model from a plurality of context-specific asset prediction models (Nath: “builds and trains one or more prediction models 125 a-125 n (‘n’ represents any natural number) to be used by the other stages (which may be referred to herein individually as a prediction model 125 or collectively as the prediction models 125). For example, the prediction models 125 can include a model for recommending to a first party (e.g., a customer) a most relevant product offered by a second party (e.g., a merchant), a model for recommending to a first party (e.g., a customer) a highest value product offered by a second party (e.g., a merchant), and a model for recommending to a first party (e.g., a customer) a most relevant and highest value product offered by a second party (e.g., a merchant).” [0031]),
wherein each of the plurality of context-specific asset prediction models is trained using training data to adjust corresponding parameters to minimize a cost value characterizing a likelihood of interaction between users and assets (Nath: “determining hyperparameters for the model 125 and performing iterative operations of inputting examples from the training data 145 a into the model 125 to find a set of model parameters (e.g., weights and/or biases) that minimizes a cost function(s) such as loss or error function for the model” [0035] – “minimize a worst-case potential loss that is evaluated with a loss function comprising a first component that represents error in a prediction of a user and product combination and a second component that represents error in a prediction of a value of a product.” [0028]), and
wherein the training data comprises user-preference features characterizing user interactions of a plurality of users with a plurality of interfaces, asset features characterizing a plurality of assets of the plurality of interfaces, and user-asset features characterizing interactions between the plurality of users represented by the user-preference features and the plurality of assets represented by the asset features (Nath: “the model 210 is trained to find out the products that are most coherent to the preference of certain first parties according to historical data.” [0041] –“ The training data 145 a may include at least a subset of historical data about first parties (e.g., customers) and products offered by a second party (e.g., enterprises banks or other merchants). … The historical data can be transactional data, customer profiles, products, and product characteristics. ” [0034]); and
that the context-specific asset prediction model comprises a machine learning model (Nath: “A prediction model 125 can be a machine-learning (“ML”) model” [0032]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because the results would be predictable. Specifically, Wang would continue to teach generate a set of predicted assets using the context-specific asset prediction model, except that now it would also teach selecting a context-specific asset prediction model from a plurality of context-specific asset prediction models based on the first context, wherein each of the plurality of context-specific asset prediction models is trained using training data to adjust corresponding parameters to minimize a cost value characterizing a likelihood of interaction between users and assets, and wherein the training data comprises user-preference features characterizing user interactions of a plurality of users with a plurality of interfaces, asset features characterizing a plurality of assets of the plurality of interfaces, and user-asset features characterizing interactions between the plurality of users represented by the user-preference features and the plurality of assets represented by the asset features; and that the context-specific asset prediction model comprises a machine learning model, according to the teachings of Nath. This is a predictable result of the combination.
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because it would result in an improved ability to increase the accuracy of recommended products (Nath: [0029]).
While Wang/Nath do not specifically teach that selecting a model from a plurality of models is based on the first context, Agarwal teaches methods and systems for determining a probability that a user with interact with digital assets (Agarwal: Abstract), including:
selecting a model from a plurality of models based on the first context (Agarwal: “after predetermined features in a web page have been detected, one or more expert statistical models to which the web page belongs are selected and weightings are assigned” [0036] – “A determination as to which expert statistical models a web page belongs is based on a comparison of predetermined features determined at operation 200 to predetermined features associated with one or more expert statistical models. … By matching a web page with most closely related expert statistical models, a probability of a user clicking on a web advertisement may be determined with higher accuracy than would be possible if only a single statistical model were used to represent user behavior on the entire corpus of available web pages.” [0037]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because the results would be predictable. Specifically, Wang/Nath would continue to teach selecting a context-specific asset prediction model from a plurality of context-specific asset prediction models, except that now it would also teach that selecting a model from a plurality of models is based on the first context, according to the teachings of Agarwal. This is a predictable result of the combination.
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because it would result in an improved ability to predict user interactions with higher accuracy (Agarwal: [0037]).
Regarding Claim 2, Wang/Nath/Agarwal teach the system of claim 1, wherein the context-specific asset prediction model comprises a random forest model (Nath: “A prediction model 125 can also be any other suitable ML model trained for providing a recommendation, such as a … Random Forest Model, ” [0032]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Nath with Wang/Agarwal for the reasons identified above with respect to claim 1.
Regarding Claim 3, Wang/Nath/Agarwal teach the system of claim 1, wherein the instructions cause the processor to receive a set of user-asset interaction features, and wherein the context-specific asset prediction model receives the set of user-asset interaction features, and wherein the set of predicted assets is selected, in part, based on the user-asset interaction features (Wang: “the affinity score engine 206 can determine a product affinity score based on user activities. User activity can include a user's browsing history (e.g. products browsing history, non-product browsing history, user searches, website history, etc.), shopping activity (e.g., product purchases, additions of products to shopping carts, product views, visits to brick-and-mortar merchant locations, etc.)” [0057]).
Regarding Claim 6, Wang/Nath/Agarwal teach the system of claim 1, wherein the context-specific asset prediction model comprises a context-specific asset prediction model that generates a context-specific set of predicted assets for a predetermined interface context (Wang: “the affinity score engine 206 can use a user's product preferences to identify products that the user is likely to prefer or not prefer based on the user product preferences,” [0056] – “rank products having a favorable product affinity score above products having an unfavorable product affinity score. By ranking products for a particular user, the customization manager 208 can prioritize products according to the user's potential interest in each product when customizing the merchant page.” [0064] – “customize a display of products on a merchant page based on a ranking of the products based on their product affinity scores … if the shopping window can only include of four (4) products. The customization manager 208 can use the rank to identify the top four (4) ranked products” [0075-0076]).
Regarding Claim 8, Wang/Nath/Agarwal teach the system of claim 1, wherein the context-specific asset prediction model maximizes the likelihood of engagement for the set of predicted assets by maximizing a likely click rate for the set of assets (Agarwal: “a click-through-rate probability is estimated for a web advertisement matched with a web page. Such a click-through-rate probability may be utilized to determine which web advertisements should be placed on a particular web page in order to maximize expected advertising revenue.” [0039]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Agarwal with Wang/Nath for the reasons identified above with respect to claim 1.
Regarding Claim 9, Wang/Nath/Agarwal teach the system of claim 1, wherein the set of user features includes a number of transactions feature, a context affinity feature, an inter-purchase interval feature, an items viewed feature, an add-to-cart feature, and a fulfillment intent feature (Wang: “the affinity score engine 206 can determine a product affinity score based on user activities. User activity can include a user's browsing history (e.g. products browsing history, non-product browsing history, user searches, website history, etc.), shopping activity (e.g., product purchases, additions of products to shopping carts, product views, visits to brick-and-mortar merchant locations, etc.) … determine, based on a user's shopping activity, that a user frequently purchases or shops for products within a particular product category.” [0057])
Regarding Claim 10, Wang discloses a computer-implemented method, comprising:
receiving a request for an interface [page] including a first context for the interface, wherein the first context comprises one or more of an account page, a home page, a cart view page, a product page, a program enrollment page, an order specific page, a checkout page, or a search page (Wang: “in response to a request from the user 406 … the social networking system 202 may present a storefront page” [0087] – “detecting a user request to access the merchant page” [Claim 2] – See Figures 4, 6.);
receiving, from a database, a set of user features [profile information] based on the first context (Wang: “The affinity score engine 206 can utilize user profile information associated with a user to determine product affinity scores for products” [0056] – “User profile information can include an age, a gender, … preferences (e.g., product preferences), likes, dislikes, and/or attitudes of a user” [0055] - “ a social networking system 1302 can comprise one or more data stores. …the social networking system 1302 can store a social graph comprising user nodes, … Each user node can comprise one or more data objects corresponding to information associated with or describing a user” [0143] – “create and store in the social networking system a user profile associated with the user.” [0131]);
receiving, from the database, a set of asset features [product information] for each of a plurality of assets [products] based on the first context (Wang: “automatically identify and obtain product information for each of products 602” [0089] – “product information can include text associated with a product (e.g., a title of the product, a description for the product, specifications for the product, a name of a manufacturer of the product” [0032] – “associating, based on a plurality of social networking communications by a merchant, a plurality of products with a merchant page.” [0115] – “store product information in another locating, such as in a remote database.” [0050]);
determining, using a context-specific asset prediction model [affinity score engine], an amount of engagement [e.g. sales volume, number of likes] associated with each of the assets in the set of assets (Wang: “the affinity score engine 206 can determine a product affinity score or a set of product affinity scores based on social trends. Social trends can include overall popularity, rating, and/or ranking of a product. …social trends can include sales volume, sales rate, and sales history of a product. …products associated with a large number of “likes” by friends of a user can be given a more favorable product affinity score ” [0058-0059])
executing the context-specific asset prediction model to generate a set of ranked assets (“determine a product affinity score for each product and then rank the products according to their product affinity scores” [0028] – “the affinity score engine 206 can calculate a product affinity score based on a correlation of one or more product characteristics, product features, and/or pieces of product information to user profile information associated with a user.” [0056] – “a user's product affinity score for a product can represent a predicted level of the user's interest in the product” [0033]),
wherein the context-specific asset prediction model receives the set of user features and the set of asset features for each asset in the plurality of assets and outputs the set of ranked assets (Wang: “the affinity score engine 206 can determine a product affinity score based on one or more of the following: user profile information, user activity, social trends, merchant input, any other suitable factors, or a combination thereof. ” [0054] – “the affinity score engine 206 can calculate a product affinity score based on a correlation of one or more product characteristics, product features, and/or pieces of product information to user profile information associated with a user. ” [0056] - “determine a product affinity score for each product and then rank the products according to their product affinity scores” [0028]), and
wherein the context-specific asset prediction model maximizes a likelihood of engagement for the set of predicted assets for the first context (Wang: “the affinity score engine 206 can use a user's product preferences to identify products that the user is likely to prefer or not prefer based on the user product preferences,” [0056] – “rank products having a favorable product affinity score above products having an unfavorable product affinity score. By ranking products for a particular user, the customization manager 208 can prioritize products according to the user's potential interest in each product when customizing the merchant page.” [0064] – “products 602 that are most likely to be of interest to user 406 may be presented to user 406 first or at the top” [0090]);
automatically generating an interface of the first context including a predetermined number of assets selected from the set of ranked assets generated using the context-specific asset prediction model in descending ranked order based on the determined amount of engagement (Wang: “customize a display of products on a merchant page based on a ranking of the products based on their product affinity scores … if the shopping window can only include of four (4) products. The customization manager 208 can use the rank to identify the top four (4) ranked products” [0075-0076] – “Like the shopping portion 416, the social networking system 202 can customize the storefront page 600 for the user 406. For example, the social networking system 202 can customize the product page 602 to display the products 602 in an order that corresponds to product affinity scores for the products 602 (e.g., as calculated by customization manager 208). … products 602 that are most likely to be of interest to user 406 may be presented to user 406 first or at the top” [0090] – See Figures 4, 6.);
transmit the interface in response to the request to display the interface (Wang: “customize a display of products on a merchant page based on a ranking of the products based on their product affinity scores … if the shopping window can only include of four (4) products. The customization manager 208 can use the rank to identify the top four (4) ranked products” [0075-0076] – See Figures 4-6);
receiving user input directed to a displayed first asset from the predetermined number of assets (Wang: “A user 406 can interact with a product within the shopping portion 416 to obtain additional information regarding the product and/or to purchase the product. For example, upon selecting a product within the shopping portion 416, the social networking system 202 can present additional product information associated with the product within a product page” [0084] – “the storefront page 600 includes … multiple selectable options 606 for the user to “See Details” for the product or “Buy” the product.” [0088] – See Figures 4, 6);
based on the user input, updating the interface to include a second asset (Wang: “based on user interactions, updated social information, and/or merchant changes, the affinity score engine 206 can send updated product affinity scores to the customization manager 208, and the customization manager 208 can change the display of products within the shopping portion based on the updated product affinity scores.” [0069]); and
transmit the updated interface to display the interface with the second asset (Wang: “based on user interactions, updated social information, and/or merchant changes, the affinity score engine 206 can send updated product affinity scores to the customization manager 208, and the customization manager 208 can change the display of products within the shopping portion based on the updated product affinity scores.” [0069] – “The social networking system 202 can present the product page 502 as a new page or as an overlay to the merchant page 402” [0084]).
While Wang teaches iterative updating the output of the engine based on new information/feedback [0069], it does not specifically teach selecting a context-specific asset prediction model from a plurality of context- specific asset prediction models based on the first context, wherein each of the plurality of context-specific asset prediction models is trained using training data to adjust corresponding parameters to minimize a cost value characterizing a likelihood of interaction between users and assets, and wherein the training data comprises user- preference features characterizing user interactions of a plurality of users with a plurality of interfaces, asset features characterizing a plurality of assets of the plurality of interfaces, and user-asset features characterizing interactions between the plurality of users represented by the user-preference features and the plurality of assets represented by the asset features.
However, Nath teaches a system for providing recommendations [Abstract] including:
selecting a context-specific asset prediction model from a plurality of context-specific asset prediction models (Nath: “builds and trains one or more prediction models 125 a-125 n (‘n’ represents any natural number) to be used by the other stages (which may be referred to herein individually as a prediction model 125 or collectively as the prediction models 125). For example, the prediction models 125 can include a model for recommending to a first party (e.g., a customer) a most relevant product offered by a second party (e.g., a merchant), a model for recommending to a first party (e.g., a customer) a highest value product offered by a second party (e.g., a merchant), and a model for recommending to a first party (e.g., a customer) a most relevant and highest value product offered by a second party (e.g., a merchant).” [0031]),
wherein each of the plurality of context-specific asset prediction models is trained using training data to adjust corresponding parameters to minimize a cost value characterizing a likelihood of interaction between users and assets (Nath: “determining hyperparameters for the model 125 and performing iterative operations of inputting examples from the training data 145 a into the model 125 to find a set of model parameters (e.g., weights and/or biases) that minimizes a cost function(s) such as loss or error function for the model” [0035] – “minimize a worst-case potential loss that is evaluated with a loss function comprising a first component that represents error in a prediction of a user and product combination and a second component that represents error in a prediction of a value of a product.” [0028]), and
wherein the training data comprises user-preference features characterizing user interactions of a plurality of users with a plurality of interfaces, asset features characterizing a plurality of assets of the plurality of interfaces, and user-asset features characterizing interactions between the plurality of users represented by the user-preference features and the plurality of assets represented by the asset features (Nath: “the model 210 is trained to find out the products that are most coherent to the preference of certain first parties according to historical data.” [0041] –“ The training data 145 a may include at least a subset of historical data about first parties (e.g., customers) and products offered by a second party (e.g., enterprises banks or other merchants). … The historical data can be transactional data, customer profiles, products, and product characteristics. ” [0034]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because the results would be predictable. Specifically, Wang would continue to teach generate a set of predicted assets using the context-specific asset prediction model, except that now it would also teach selecting a context-specific asset prediction model from a plurality of context-specific asset prediction models based on the first context, wherein each of the plurality of context-specific asset prediction models is trained using training data to adjust corresponding parameters to minimize a cost value characterizing a likelihood of interaction between users and assets, and wherein the training data comprises user-preference features characterizing user interactions of a plurality of users with a plurality of interfaces, asset features characterizing a plurality of assets of the plurality of interfaces, and user-asset features characterizing interactions between the plurality of users represented by the user-preference features and the plurality of assets represented by the asset features, according to the teachings of Nath. This is a predictable result of the combination.
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because it would result in an improved ability to increase the accuracy of recommended products (Nath: [0029]).
While Wang/Nath do not specifically teach that selecting a model from a plurality of models is based on the first context, Agarwal teaches methods and systems for determining a probability that a user with interact with digital assets (Agarwal: Abstract), including:
selecting a model from a plurality of models based on the first context (Agarwal: “after predetermined features in a web page have been detected, one or more expert statistical models to which the web page belongs are selected and weightings are assigned” [0036] – “A determination as to which expert statistical models a web page belongs is based on a comparison of predetermined features determined at operation 200 to predetermined features associated with one or more expert statistical models. … By matching a web page with most closely related expert statistical models, a probability of a user clicking on a web advertisement may be determined with higher accuracy than would be possible if only a single statistical model were used to represent user behavior on the entire corpus of available web pages.” [0037]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because the results would be predictable. Specifically, Wang/Nath would continue to teach selecting a context-specific asset prediction model from a plurality of context-specific asset prediction models, except that now it would also teach that selecting a model from a plurality of models is based on the first context, according to the teachings of Agarwal. This is a predictable result of the combination.
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because it would result in an improved ability to predict user interactions with higher accuracy (Agarwal: [0037]).
Regarding Claim 11, Wang/Nath/Agarwal teach the computer-implemented method of claim 10, wherein the context-specific asset prediction model comprises a random forest model (Nath: “A prediction model 125 can also be any other suitable ML model trained for providing a recommendation, such as a … Random Forest Model, ” [0032]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Nath with Wang/Agarwal for the reasons identified above with respect to claim 1.
Regarding Claim 12, Wang/Nath/Agarwal teach the computer-implemented method of claim 10, comprising receiving a set of user-asset interaction features, and wherein the context-specific asset prediction model receives the set of user-asset interaction features, and wherein the set of ranked assets is selected, in part, based on the user-asset interaction features (Wang: “the affinity score engine 206 can determine a product affinity score based on user activities. User activity can include a user's browsing history (e.g. products browsing history, non-product browsing history, user searches, website history, etc.), shopping activity (e.g., product purchases, additions of products to shopping carts, product views, visits to brick-and-mortar merchant locations, etc.)” [0057]).
Regarding Claim 14, Wang/Nath/Agarwal teach the computer-implemented method of claim 10, wherein the context-specific asset prediction model comprises a context-specific asset prediction model that generates a context-specific set of ranked assets for a predetermined interface context (Wang: “the affinity score engine 206 can use a user's product preferences to identify products that the user is likely to prefer or not prefer based on the user product preferences,” [0056] – “rank products having a favorable product affinity score above products having an unfavorable product affinity score. By ranking products for a particular user, the customization manager 208 can prioritize products according to the user's potential interest in each product when customizing the merchant page.” [0064] – “customize a display of products on a merchant page based on a ranking of the products based on their product affinity scores … if the shopping window can only include of four (4) products. The customization manager 208 can use the rank to identify the top four (4) ranked products” [0075-0076]).
Regarding Claim 16, Wang/Nath/Agarwal teach the computer-implemented method of claim 10, wherein the context-specific asset prediction model maximizes the likelihood of engagement for the set of ranked assets by maximizing a likely click rate for the plurality of assets (Agarwal: “a click-through-rate probability is estimated for a web advertisement matched with a web page. Such a click-through-rate probability may be utilized to determine which web advertisements should be placed on a particular web page in order to maximize expected advertising revenue.” [0039]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Agarwal with Wang for the reasons identified above with respect to claim 10.
Regarding Claim 21, Wang discloses a non-transitory computer-readable medium having instructions stored thereon that, when executed by at least one processor, cause at least one device to perform operations (Wang: [0124-0126]) comprising:
receiving a request for an interface [page] including a first context for the interface, wherein the first context comprises one or more of an account page, a home page, a cart view page, a product page, a program enrollment page, an order specific page, a checkout page, or a search page (Wang: “in response to a request from the user 406 … the social networking system 202 may present a storefront page” [0087] – “detecting a user request to access the merchant page” [Claim 2] – See Figures 4, 6);
receiving, from a database, a set of user features [profile information] based on the first context (Wang: “The affinity score engine 206 can utilize user profile information associated with a user to determine product affinity scores for products” [0056] – “User profile information can include an age, a gender, … preferences (e.g., product preferences), likes, dislikes, and/or attitudes of a user” [0055] - “ a social networking system 1302 can comprise one or more data stores. …the social networking system 1302 can store a social graph comprising user nodes, … Each user node can comprise one or more data objects corresponding to information associated with or describing a user” [0143] – “create and store in the social networking system a user profile associated with the user.” [0131);
receiving, from the database, a set of asset features [product information] for each of a plurality of assets [products] based on the first context (Wang: “automatically identify and obtain product information for each of products 602” [0089] – “product information can include text associated with a product (e.g., a title of the product, a description for the product, specifications for the product, a name of a manufacturer of the product” [0032] – “associating, based on a plurality of social networking communications by a merchant, a plurality of products with a merchant page.” [0115] – “store product information in another locating, such as in a remote database.” [0050]);
executing a context-specific asset prediction model [affinity score engine] to generate a set of ranked assets (Wang: “the affinity score engine 206 can determine a product affinity score or a set of product affinity scores based on social trends. Social trends can include overall popularity, rating, and/or ranking of a product. …social trends can include sales volume, sales rate, and sales history of a product. …products associated with a large number of “likes” by friends of a user can be given a more favorable product affinity score ” [0058-0059] - determine a product affinity score for each product and then rank the products according to their product affinity scores” [0028] – “the affinity score engine 206 can calculate a product affinity score based on a correlation of one or more product characteristics, product features, and/or pieces of product information to user profile information associated with a user.” [0056] – “a user's product affinity score for a product can represent a predicted level of the user's interest in the product” [0033]),
wherein the context-specific asset prediction model receives the set of user features and the set of asset features for each asset in the plurality of assets and outputs the set of ranked assets (Wang: “the affinity score engine 206 can determine a product affinity score based on one or more of the following: user profile information, user activity, social trends, merchant input, any other suitable factors, or a combination thereof. ” [0054] – “the affinity score engine 206 can calculate a product affinity score based on a correlation of one or more product characteristics, product features, and/or pieces of product information to user profile information associated with a user. ” [0056]), and
wherein the context-specific asset prediction model maximizes a likelihood of engagement for the set of ranked assets (Wang: “the affinity score engine 206 can use a user's product preferences to identify products that the user is likely to prefer or not prefer based on the user product preferences,” [0056] – “rank products having a favorable product affinity score above products having an unfavorable product affinity score. By ranking products for a particular user, the customization manager 208 can prioritize products according to the user's potential interest in each product when customizing the merchant page.” [0064] – “products 602 that are most likely to be of interest to user 406 may be presented to user 406 first or at the top” [0090]); and
automatically generating an interface of the first context including a predetermined number of assets selected from the set of ranked assets generated using the context-specific asset prediction model in descending ranked order based on the determined amount of engagement (Wang: “customize a display of products on a merchant page based on a ranking of the products based on their product affinity scores … if the shopping window can only include of four (4) products. The customization manager 208 can use the rank to identify the top four (4) ranked products” [0075-0076] – “Like the shopping portion 416, the social networking system 202 can customize the storefront page 600 for the user 406. For example, the social networking system 202 can customize the product page 602 to display the products 602 in an order that corresponds to product affinity scores for the products 602 (e.g., as calculated by customization manager 208). … products 602 that are most likely to be of interest to user 406 may be presented to user 406 first or at the top” [0090] – See Figures 4, 6.);
transmit the interface in response to the request to display the interface (Wang: “customize a display of products on a merchant page based on a ranking of the products based on their product affinity scores … if the shopping window can only include of four (4) products. The customization manager 208 can use the rank to identify the top four (4) ranked products” [0075-0076] – See Figures 4-6);
receiving user input directed to a displayed first asset from the predetermined number of assets (Wang: “A user 406 can interact with a product within the shopping portion 416 to obtain additional information regarding the product and/or to purchase the product. For example, upon selecting a product within the shopping portion 416, the social networking system 202 can present additional product information associated with the product within a product page” [0084] – “the storefront page 600 includes … multiple selectable options 606 for the user to “See Details” for the product or “Buy” the product.” [0088] – See Figures 4, 6);
based on the user input, updating the interface to include a second asset (Wang: “based on user interactions, updated social information, and/or merchant changes, the affinity score engine 206 can send updated product affinity scores to the customization manager 208, and the customization manager 208 can change the display of products within the shopping portion based on the updated product affinity scores.” [0069]).
While Wang teaches iterative updating the output of the engine based on new information/feedback [0069], it does not specifically teach selecting a context-specific asset prediction model from a plurality of context- specific asset prediction models based on the first context, wherein each of the plurality of context-specific asset prediction models is trained using training data to adjust corresponding parameters to minimize a cost value characterizing a likelihood of interaction between users and assets, and wherein the training data comprises user- preference features characterizing user interactions of a plurality of users with a plurality of interfaces, asset features characterizing a plurality of assets of the plurality of interfaces, and user-asset features characterizing interactions between the plurality of users represented by the user-preference features and the plurality of assets represented by the asset features.
However, Nath teaches a system for providing recommendations [Abstract] including:
selecting a context-specific asset prediction model from a plurality of context-specific asset prediction models (Nath: “builds and trains one or more prediction models 125 a-125 n (‘n’ represents any natural number) to be used by the other stages (which may be referred to herein individually as a prediction model 125 or collectively as the prediction models 125). For example, the prediction models 125 can include a model for recommending to a first party (e.g., a customer) a most relevant product offered by a second party (e.g., a merchant), a model for recommending to a first party (e.g., a customer) a highest value product offered by a second party (e.g., a merchant), and a model for recommending to a first party (e.g., a customer) a most relevant and highest value product offered by a second party (e.g., a merchant).” [0031]),
wherein each of the plurality of context-specific asset prediction models is trained using training data to adjust corresponding parameters to minimize a cost value characterizing a likelihood of interaction between users and assets (Nath: “determining hyperparameters for the model 125 and performing iterative operations of inputting examples from the training data 145 a into the model 125 to find a set of model parameters (e.g., weights and/or biases) that minimizes a cost function(s) such as loss or error function for the model” [0035] – “minimize a worst-case potential loss that is evaluated with a loss function comprising a first component that represents error in a prediction of a user and product combination and a second component that represents error in a prediction of a value of a product.” [0028]), and
wherein the training data comprises user-preference features characterizing user interactions of a plurality of users with a plurality of interfaces, asset features characterizing a plurality of assets of the plurality of interfaces, and user-asset features characterizing interactions between the plurality of users represented by the user-preference features and the plurality of assets represented by the asset features (Nath: “the model 210 is trained to find out the products that are most coherent to the preference of certain first parties according to historical data.” [0041] –“ The training data 145 a may include at least a subset of historical data about first parties (e.g., customers) and products offered by a second party (e.g., enterprises banks or other merchants). … The historical data can be transactional data, customer profiles, products, and product characteristics. ” [0034]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because the results would be predictable. Specifically, Wang would continue to teach generate a set of predicted assets using the context-specific asset prediction model, except that now it would also teach selecting a context-specific asset prediction model from a plurality of context-specific asset prediction models based on the first context, wherein each of the plurality of context-specific asset prediction models is trained using training data to adjust corresponding parameters to minimize a cost value characterizing a likelihood of interaction between users and assets, and wherein the training data comprises user-preference features characterizing user interactions of a plurality of users with a plurality of interfaces, asset features characterizing a plurality of assets of the plurality of interfaces, and user-asset features characterizing interactions between the plurality of users represented by the user-preference features and the plurality of assets represented by the asset features, according to the teachings of Nath. This is a predictable result of the combination.
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because it would result in an improved ability to increase the accuracy of recommended products (Nath: [0029]).
While Wang/Nath do not specifically teach that selecting a model from a plurality of models is based on the first context, Agarwal teaches methods and systems for determining a probability that a user with interact with digital assets (Agarwal: Abstract), including:
selecting a model from a plurality of models based on the first context (Agarwal: “after predetermined features in a web page have been detected, one or more expert statistical models to which the web page belongs are selected and weightings are assigned” [0036] – “A determination as to which expert statistical models a web page belongs is based on a comparison of predetermined features determined at operation 200 to predetermined features associated with one or more expert statistical models. … By matching a web page with most closely related expert statistical models, a probability of a user clicking on a web advertisement may be determined with higher accuracy than would be possible if only a single statistical model were used to represent user behavior on the entire corpus of available web pages.” [0037]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because the results would be predictable. Specifically, Wang/Nath would continue to teach selecting a context-specific asset prediction model from a plurality of context-specific asset prediction models, except that now it would also teach that selecting a model from a plurality of models is based on the first context, according to the teachings of Agarwal. This is a predictable result of the combination.
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because it would result in an improved ability to predict user interactions with higher accuracy (Agarwal: [0037]).
Regarding Claim 22, Wang/Agarwal teach the non-transitory computer-readable medium of claim 21, wherein the context-specific asset prediction model comprises a random forest model (Nath: “A prediction model 125 can also be any other suitable ML model trained for providing a recommendation, such as a … Random Forest Model, ” [0032]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Nath with Wang/Agarwal for the reasons identified above with respect to claim 21.
Regarding Claim 24, Wang/Nath/Agarwal teach the non-transitory computer-readable medium of claim 21, wherein the context-specific asset prediction model comprises a context-specific asset prediction model that generates a context-specific set of ranked assets for a predetermined interface context (Wang: “the affinity score engine 206 can use a user's product preferences to identify products that the user is likely to prefer or not prefer based on the user product preferences,” [0056] – “rank products having a favorable product affinity score above products having an unfavorable product affinity score. By ranking products for a particular user, the customization manager 208 can prioritize products according to the user's potential interest in each product when customizing the merchant page.” [0064] – “customize a display of products on a merchant page based on a ranking of the products based on their product affinity scores … if the shopping window can only include of four (4) products. The customization manager 208 can use the rank to identify the top four (4) ranked products” [0075-0076]).
Claims 4, 7, and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Wang/Nath/Agarwal, and further in view of Okerlund et al (US20140279420A1), hereinafter Okerlund.
Regarding Claim 4, Wang/Nath/Agarwal teach the system of claim 3, wherein the set of user-asset interaction features includes a number of views feature, a number of clicks feature (Wang: “User activity can include a user's browsing history (e.g. products browsing history, non-product browsing history, user searches, website history, etc.), shopping activity (e.g., product purchases, additions of products to shopping carts, product views, visits to brick-and-mortar merchant locations, etc.), social networking activity (e.g., “likes,” shares, comments,… product ratings, … tracked” [0057]).
While Wang/Nath/Agarwal further teaches that user activity can include “social networking group membership… and/or any other activity associated with a user,” (Wang: [0057]) it does not specifically teach that the set of user-asset interaction features further includes a number of cancels feature, a loyalty program feature, and a tip action feature representative of prior tipping or gratuity actions for a user when interacting with delivery assets.
However, Okerlund teaches a system and method for making customized offers to a user based on consumer data (Okerlund: [0352]), including that the set of user-asset interaction features further includes
a number of cancels feature (Okerlund: “registrant's past acceptance or rejection of other incentive offers” [0281] – “An incentive offer may be declined either actively, such as by deleting the incentive offer from a web or smartphone accessible list of available incentive offers, ” [0268]),
a loyalty program feature (Okerlund: “download the transaction data from the program registrant's account.” [0094] – “transaction data may be transmitted from … a loyalty card issued by the merchant itself.” [0099] – See also [0149], wherein the system operates a loyalty program based on customer loyalty activity.), and
a tip action feature representative of prior tipping or gratuity actions for a user when interacting with delivery assets (Okerlund: “an analysis of the program registrant's transaction history, including in some embodiments their spending and tipping history. …identifies program registrants that have patronized … a history of tipping … at the particular restaurant” [0147]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because the results would be predictable. Specifically, Wang/Nath/Agarwal would continue to teach that the set of user-asset interaction features includes a number of views feature, a number of clicks feature, except that now it would also teach that the set of user-asset interaction features further includes a number of cancels feature, a loyalty program feature, and a tip action feature representative of prior tipping or gratuity actions for a user when interacting with delivery assets, according to the teachings of Okerlund. This is a predictable result of the combination.
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because it would result in an improved accuracy of incentives delivered to a user (Okerlund: [0073]).
Regarding Claim 7, Wang/Nath/Agarwal teach the system of claim 1, but do not specifically teach that the instructions cause the processor to: receive an asset dismissal for the displayed first asset in the predetermined number of assets, wherein the asset dismissal removes the displayed first asset from the interface; select a second asset, wherein the second asset includes an asset in the set of predicted assets but not in the predetermined number of assets; and update the interface to include the second asset.
However, Okerlund teaches a system and method for making customized offers to a user based on consumer data (Okerlund: [0352]), including the ability to:
receive an asset dismissal for the displayed first asset in the predetermined number of assets, wherein the asset dismissal removes the displayed first asset from the interface (Okerlund: “An incentive offer may be declined either actively, such as by deleting the incentive offer from a web or smartphone accessible list of available incentive offers” [0268]);
select a second asset, wherein the second asset includes an asset in the set of predicted assets but not in the predetermined number of assets; and update the interface to include the second asset (Okerlund: “Data on whether or not an incentive offer is redeemed …may be used …to … tailor incentive offers delivered to program registrants to maximize redemptive behavior” [0268] – “the data represented in 16B6 may include a list of any incentive offers currently pending for that program registrant …may also contain data regarding that program registrant's past acceptance or rejection of other incentive offers” [0281] – “an update is made to a predetermined category of data, that update may be configured to trigger an automated incentive-related query in order to determine if one or more current incentive campaigns that previously were not relevant to a program registrant now become relevant.” [0123] – See also [0324]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because the results would be predictable. Specifically, Wang/Nath/Agarwal would continue to teach receive user input directed to a displayed first asset from the predetermined number of assets, except that now it would also teach receive an asset dismissal for the displayed first asset in the predetermined number of assets, wherein the asset dismissal removes the displayed first asset from the interface; select a second asset, wherein the second asset includes an asset in the set of predicted assets but not in the predetermined number of assets; and update the interface to include the second asset, according to the teachings of Okerlund. This is a predictable result of the combination.
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because it would result in an improved accuracy of incentives delivered to a user (Okerlund: [0073]).
Regarding Claim 15, Wang/Nath/Agarwal teach the computer-implemented method of claim 10, but do not specifically teach receiving a dismissal notification for the first asset included in the interface; selecting a second asset, wherein the second asset includes an asset in the set of ranked assets but not in the predetermined number of assets; and updating the interface to include the second asset.
However, Okerlund teaches a system and method for making customized offers to a user based on consumer data (Okerlund: [0352]), including the ability to:
receiving a dismissal notification for the first asset included in the interface (Okerlund: “An incentive offer may be declined either actively, such as by deleting the incentive offer from a web or smartphone accessible list of available incentive offers” [0268]);
selecting a second asset, wherein the second asset includes an asset in the set of ranked assets but not in the predetermined number of assets; and updating the interface to include the second asset (Okerlund: “Data on whether or not an incentive offer is redeemed …may be used …to … tailor incentive offers delivered to program registrants to maximize redemptive behavior” [0268] – “the data represented in 16B6 may include a list of any incentive offers currently pending for that program registrant …may also contain data regarding that program registrant's past acceptance or rejection of other incentive offers” [0281] – “an update is made to a predetermined category of data, that update may be configured to trigger an automated incentive-related query in order to determine if one or more current incentive campaigns that previously were not relevant to a program registrant now become relevant.” [0123] – See also [0324]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because the results would be predictable. Specifically, Wang/Nath/Agarwal would continue to teach receive user input directed to a displayed first asset from the predetermined number of assets, except that now it would also teach receiving a dismissal notification for the first asset included in the interface; selecting a second asset, wherein the second asset includes an asset in the set of ranked assets but not in the predetermined number of assets; and updating the interface to include the second asset, according to the teachings of Okerlund. This is a predictable result of the combination.
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because it would result in an improved accuracy of incentives delivered to a user (Okerlund: [0073]).
Claim 5 is rejected under 35 U.S.C. 103 as being unpatentable over Wang/Nath/Agarwal/Okerlund, and further in view of Goad et al (US 20110137776 A1), hereinafter Goad.
Regarding Claim 5, Wang/Nath/Agarwal/Okerlund teach the system of claim 4, but do not specifically teach the loyalty program feature includes an initial enrollment feature and a renewal status feature.
However, Goad teaches a recommendation system (Goad: [0009]), including that the loyalty program feature includes an initial enrollment feature and a renewal status feature (Goad: “an alert if … the user's cellular phone plan is about to expire” [0157] – “the notification module 1014 retrieves current service usage information for all services for which notifications are required. … contract expiry dates or bill due dates may be retrieved from the customer information database 116.” [0160] – “alerts 1018 include … plan expiry dates” [0164] – “the advisor engine 134 provides users with up-to-date and personalized information and/or advice on their home” [0082] – “if a given user has previously rejected, on multiple occasions, offers … then the engine 134 may determine that the user is uninterested” [0103]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because the results would be predictable. Specifically, Wang/Nath/Agarwal/Okerlund would continue to teach the set of user-asset interaction features includes a loyalty program feature, except now it would also teach that the loyalty program feature includes an initial enrollment feature and a renewal status feature, according to the teachings of Goad. This is a predictable result of the combination.
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because it would result in an improved ability to provide the most relevant results to the user (Goad: [0141]).
Claims 13, and 23 are rejected under 35 U.S.C. 103 as being unpatentable over Wang/Nath/Agarwal, and further in view of Goad.
Regarding Claim 13, Wang/Nath/Agarwal teach the computer-implemented method of claim 12, but do not specifically teach that the set of user-asset interaction features includes an initial loyalty program enrollment feature and a loyalty program renewal status feature.
However, Goad teaches a recommendation system (Goad: [0009]), including the set of user-asset interaction features includes an initial loyalty program enrollment feature and a loyalty program renewal status feature (Goad: “an alert if … the user's cellular phone plan is about to expire” [0157] – “the notification module 1014 retrieves current service usage information for all services for which notifications are required. … contract expiry dates or bill due dates may be retrieved from the customer information database 116.” [0160] – “alerts 1018 include … plan expiry dates” [0164] – “the advisor engine 134 provides users with up-to-date and personalized information and/or advice on their home” [0082] – “if a given user has previously rejected, on multiple occasions, offers … then the engine 134 may determine that the user is uninterested” [0103]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because the results would be predictable. Specifically, Wang/Nath/Agarwal would continue to teach the set of user-asset interaction features, except now it would also teach that the set of user-asset interaction features includes an initial loyalty program enrollment feature and a loyalty program renewal status feature, according to the teachings of Goad. This is a predictable result of the combination.
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because it would result in an improved ability to provide the most relevant results to the user (Goad: [0141]).
Regarding Claim 23, Wang/Nath/Agarwal teach the non-transitory computer-readable medium of claim 21, wherein the instructions cause the at least one device to perform operations comprising receiving a set of user-asset interaction features, and wherein the context-specific asset prediction model receives the set of user-asset interaction features, and wherein the set of ranked assets is selected, in part, based on the user-asset interaction features (Wang: “the affinity score engine 206 can determine a product affinity score based on user activities. User activity can include a user's browsing history (e.g. products browsing history, non-product browsing history, user searches, website history, etc.), shopping activity (e.g., product purchases, additions of products to shopping carts, product views, visits to brick-and-mortar merchant locations, etc.)” [0057]),
but do not specifically teach that the set of user-asset interaction features includes an initial loyalty program enrollment feature and a loyalty program renewal status feature.
However, Goad teaches a recommendation system (Goad: [0009]), including that the set of user-asset interaction features includes an initial loyalty program enrollment feature and a loyalty program renewal status feature (Goad: “an alert if … the user's cellular phone plan is about to expire” [0157] – “the notification module 1014 retrieves current service usage information for all services for which notifications are required. … contract expiry dates or bill due dates may be retrieved from the customer information database 116.” [0160] – “alerts 1018 include … plan expiry dates” [0164] – “the advisor engine 134 provides users with up-to-date and personalized information and/or advice on their home” [0082] – “if a given user has previously rejected, on multiple occasions, offers … then the engine 134 may determine that the user is uninterested” [0103]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because the results would be predictable. Specifically, Wang/Nath/Agarwal would continue to teach the set of user-asset interaction features, except now it would also teach that that the set of user-asset interaction features includes an initial loyalty program enrollment feature and a loyalty program renewal status feature, according to the teachings of Goad. This is a predictable result of the combination.
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of invention to combine these references because it would result in an improved ability to provide the most relevant results to the user (Goad: [0141]).
Response to Arguments
Applicant's arguments filed 3/9/2026 have been fully considered but they are not persuasive.
Claim Rejection – 35 §USC 101
Applicant argues that “claim 1 recites subject matter that improves computer functionality through parameter optimization of context-specific predictive models, and thus is not directed to an abstract idea,” with reference to Desjardins. Applicant refers to limitations from the claim that the models are trained “to adjust corresponding parameters to minimize a cost value characterizing a likelihood of interaction.”
Examiner respectfully disagrees. The argued limitation of adjusting parameters is part of the abstract idea itself, which cannot form the sole basis for a technological improvement to itself. Rather than reciting a specific method of parameter optimization, the claim merely recites a general abstract step of adjusting parameters to determine a likelihood value, and recites additional elements, such as the models being trained ML models, as mere instructions to apply this abstract idea to a technological environment [MPEP 2106.05(f)]. Whereas Desjardins recites a specific technological problem in the Specification and provides a specific technological solution to it in the claims, the pending claims merely recite an abstract idea for determining personalized recommendation, including adjusting parameters of a model for predicting interest, along with additional elements which create only a general linking to computer technology.
Applicant further argues that the claimed subject matter “amounts to an improvement in computer functionality at least by enabling structured, feature-based parameter optimization within a predictive modeling framework,” and that “claim 1 provides an improvement to user interfaces that allows for the display of user specific assets,” which Applicant argues “can reduce redundant interface reconstruction operations and improve computational efficiencies in generating and updating context-specific interfaces.”
Examiner disagrees. The claims do not recite a structured, feature-based methodology for optimizing parameters, but merely the ability “to adjust corresponding parameters to minimize a cost value characterizing a likelihood of interaction,” which is part of the abstract idea itself. Similarly, the ability to display and receive interactions with user-specific asset recommendations is part of the abstract idea, such that any alleged reduced operations and efficiencies stem only from the abstract idea instead of being rooted in computer technology. The additional elements are recited at a high level of generality and amount to mere instructions to apply this abstract idea to a technological environment [MPEP 2106.05(f)]. At best, these additional elements offer only the improved speed or efficiency inherent to a general purpose computer [MPEP 2106.05(a)], which does not integrate the abstract idea into a practical application.
Applicant further argues that “the claimed combination applies a particular machine learning architecture and interface generation mechanism to achieve context-dependent prediction and structured interface updating, thereby providing significantly more than the abstract idea.” Applicant argues that “the claimed subject matter is confined to a particular useful application that provides several technical benefits.”
Examiner disagrees. Similar to the discussion above, the ability to update displayed recommendations and to perform the specific analysis/prediction claimed are part of the abstract idea itself. The additional elements are recited at a high level of generality and amount to mere instructions to apply this abstract idea to a technological environment [MPEP 2106.05(f)]. Rather than reciting a “particular machine learning architecture,” the claims merely invoke a generic “machine learning model” or trained models as instructions to apply the abstract steps to computer technology.
Claim Rejection – 35 §USC 103
Applicant’s arguments with respect to the prior art rejection have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument. Specifically, the amended selecting step is taught be newly-relied-upon reference Nath, except for the recitation that the selecting is based on the context, which is taught by Agarwal in the rejection above.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Liang et al (US 20170097741 A1) teaches content presentation systems that use machine learning to predict the likelihood of user engagement with content items.
Carbune et al (US 20180131655 A1) teaches machine learning techniques for predicting the likelihood of user engagement with notifications, and present notifications in ranked importance based on a predetermined number of notifications for presentation.
Westphal et al (US 12106357 B2) teaches systems for dynamic reorganization of a user interface based on user historical engagement with icons on the page, such as a website home screen.
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to THOMAS JOSEPH SULLIVAN whose telephone number is (571)272-9736. The examiner can normally be reached on Mon - Fri 8-5 PT.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Marissa Thein can be reached on 571-272-6764. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov.
Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/T.J.S./Examiner, Art Unit 3689
/MARISSA THEIN/Supervisory Patent Examiner, Art Unit 3689
Prosecution Insights