DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Introductory Remarks
In response to communications filed on 17 July 2026, claims 1, 5, 8, and 14-15 are amended per Applicant's request. No claims were cancelled. No claims were withdrawn. No new claims were added. Therefore, claims 1-20 are presently pending in the application, of which claims 1, 8, and 15 are presented in independent form.
The previously raised 101 rejection of the pending claims is maintained.
The previously raised 103 rejection of the pending claims is withdrawn in view of the amendments to the claims. A new ground(s) of rejection has been issued.
Response to Arguments
Applicant’s arguments filed 17 July 2026 with respect to the rejection of the claims under 35 U.S.C. 101 (see Remarks, p. 11-15) have been fully considered but are not persuasive.
Applicant’s arguments that certain limitations cannot be performed in the human mind including “communicate with a first set of web functions…to receive a first plurality of data structures”, “communicate with a second set of web functions…to receive a second plurality of data structures”, “transmit, to a traffic data source, a request for traffic information”, “pre-process feature sets”, or “train a machine learning model” (see Remarks, p. 12-13) is unpersuasive.
Because the focus of the claimed invention is on analyzing received information including discarding/filtering out certain information and taking into consideration other types of information, the claims still recite an abstract idea. The elements recited by Applicant as examples of not reciting abstract ideas is not persuasive, because these are considered at Step 2A, Prong Two as whether these additional elements amount to significantly more.
The reason these elements were treated as “additional elements” rather than at Step 2A, Prong One is because of the nature of the claims, which focus on the analysis of the received information. Thus, these elements of communicating with web functions for receiving data structures, transmitting a request for traffic information, pre-processing feature sets, or training a machine learning model, do nothing more than attempt to limit the claims to a particular technological environment—namely, implementation via computers. They also amount to well-understood, routine, and conventional extra-solution activities of receiving/transmitting data over a network. The use of web functions also similarly is an insignificant field-of-use limitation, describing the context rather than a particular manner of achieving the result.
However, none of these elements are integrated into how the claimed invention performs its analysis steps. Though the claims invoke the use of a machine learning model, it is simply invoked as a tool in executing the abstract steps. However, the machine learning model is invoked so generically, e.g., taking certain information as input and outputting a corresponding result, that is amounts to nothing more than stating the abstract idea while adding the words “apply it” with a computer—or in this case, something slightly narrower than a computer (machine learning model), which only has the same deficient result (i.e., as being an insignificant field-of-use limitation).
With respect to the claims also reciting certain methods of organizing human activity (see Remarks, p. 13-14), Applicant argues that claim 1 does not recite any enumerated sub-grouping of organizing human activity, e.g., using machine learning to select from a plurality of possible locations (see Remarks, p. 13), and that a practical application of determining a location where an object is acquired using a single machine learning mode that accounts for multiple factors (Remarks, p. 13).
However, these arguments are unpersuasive, as the focus of the claimed invention lies on taking various human- and business-related factors into consideration (e.g., location, traffic, accounts, rewards, price, etc.) for eventually arriving at some sort of optimization solution, without describing the particulars as to how the optimization solution itself is achieved beyond minimally narrowing the machine learning model to being trained on certain datasets that is claimed in a wholly generic manner (e.g., minimal feature sets is not something Applicant has claimed to have invented, and the manner in which it is invoked is not tethered to the rest of the claimed steps), and that the machine learning model is capable of somehow minimizing the cost, without stating how either the training or particular step of minimizing the cost is accomplished.
Additionally, Applicant’s arguments that the claimed invention purportedly improves “conventional machine learning techniques that rely on multiple models to account for multiple factors…which conserves processing resources otherwise associated with the use of multiple models” (see Remarks, p. 14-15) is unpersuasive, as (1) contrary to Applicant’s assertion, many environments also may use one model to account for multiple factors, e.g., a neural network model that may take in multiple inputs to arrive at one output, and (2) as argued above already, the machine learning model is not integrated into a practical application, as it is invoked in a wholly generic manner.
Therefore, for at least the aforementioned reasons and those set forth in the 101 rejection below, the 101 rejection has been maintained.
Applicant’s arguments filed 17 July 2026 with respect to the rejection of the claims under 35 U.S.C. 103 (see Remarks, p. 16) have been fully considered but are not persuasive. Applicant argues solely that the amendments render the prior rejections moot. See Response at page 16. The examiner respectfully disagrees and the rejections have been modified to conform to the current claim language.
Claim Rejections - 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 1-20 are rejected under 35 U.S.C. 101 because the claims are directed to a judicial exception (i.e., an abstract idea) without significantly more.
Independent Claims 1, 8, and 15 recite
receiving an indication of at least one object from a user device (claim 8 utilizes the language “transmitting”);
determining a plurality of possible locations associated with the at least one object that was received (claims 1 and 15);
determining a plurality of possible accounts associated with the user of the user device (claim 8 utilizes the language “transmitting”);
receiving location information associated with the at least one object (claims 1 and 15); determine a plurality of possible accounts associated with a user of the user device (claim 8 utilizes the language “transmitting”);
receiving terms associated with the plurality of possible accounts (claims 1 and 15);
receiving an indication of a location associated with the user device (claim 1 utilizes “current” location, claim 8 uses only “location”, and claim 15 utilizes “starting” location);
transmitting a request for traffic information based on the current location and the plurality of possible locations (claims 1 and 15);
receiving the traffic information in response to the request (claims 1 and 15);
minimizing a cost associated with the at least one object, wherein the cost is based on a price associated with the location information, reward points information, and traffic information (claims 1, 8, and 15);
receiving a particular location and account pairing from the plurality of possible locations and the plurality of possible accounts (claim 15 further includes “terms”);
outputting an indication of the selected location and an indication of the selected account to the user device, receiving an acceptance of the selected location from the user device, and outputting, to the user device, directions to the selected location from the current location (claim 1, and similarly, dependent claim 14).
Dependent Claim 4 recites determining a set of endpoints based on the plurality of possible locations.
Dependent Claim 13 recites receiving an interaction with an indication of the selected location; and outputting data associated with the selected location and the at least one object.
These fall under “Certain methods of organizing human activity” (more particularly including managing personal behavior or relationships or interactions between people). As seen, the claims are primarily about taking a user-specified object (e.g., product), user location and user account information, and using this information to identify relevant locations, which is a form of a human activity. This also represents a form of “filtering content” (which is one of the examples of managing personal behavior). Furthermore, in the case of claim 1, the claim provides driving directions to users to certain shopping/retail locations (a form of a user activity).
Additionally, the independent claims’ step of “pre-process feature sets comprising location and account pairings to generate a minimum feature set, wherein the system is configured to filter the feature sets by rejecting location and account pairings that are incompatible”, “train[ing]…using the minimum feature set”, and “minimize a cost associated with the at least one object”, encompass an evaluation, observation, and/or judgment, as well as some sort of learning process using certain sets of information (which falls under the “Mental Processes” grouping of abstract ideas). Furthermore, the pre-processing step by filtering certain data within the feature sets also represents a form of “filtering content” (which is one of the examples of managing personal behavior).
Because the claims recite limitations that fall within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas, the claims recite an abstract idea.
The claims are not integrated into a practical application of that idea. In particular, the claims recite various computing hardware components, which are recited at a high level of generality and recited so generically that they represent no more than mere instructions to apply the judicial exception on a computer (see MPEP 2106.05(f)). These limitations can also be viewed as nothing more than an attempt to generally link the use of the judicial exception to the technological environment of a computer (see MPEP 2106.05(h)).
Independent Claims 1, 8, and 15 recite that the disclosed steps for selecting from a plurality of possible locations (and accounts) are performed by a machine learning model. This does nothing more than attempt to limit the claims to a particular field-of-use, which does not amount to significantly more, as they only provide a context rather than a particular manner of achieving the result. Additionally or alternatively, the use of machine learning models does nothing more than attempt to limit the claims to a particular technological environment—namely, implementation via machine learning, which is only slightly narrower than a generic computer. However, as a matter of law, narrowing or reformulating an abstract idea does not add “significantly more” to it.
As a result, the claimed limitation “wherein the machine learning model is configured to filter the feature sets by rejecting location and account pairings that are incompatible due to a location not accepting an account type or due to a location not accepting a payment network” (in independent claim 1), do nothing more than provide a context for the machine learning model, i.e., an insignificant field-of-use limitation. Similarly, stating that a machine learning model is “trained” on the feature sets (e.g., instead of a person), also does nothing more than amount to an insignificant field-of-use limitation, i.e., attempting to limit the claims to a particular technological environment—namely, implementation via computers.
Thus, dependent claim 5, which pertains to the machine learning model being configured to minimize a holistic cost associated with the at least one object, is nothing more than an insignificant field-of-use limitation, describing the context rather than a particular manner of achieving the result.
Independent Claim 1 recites that a plurality of possible locations is determined “using an object database”, that a plurality of possible accounts are determined “using a user database”, and that information is retrieved from “a traffic data source”. Again, this does nothing more than attempt to limit the claims to generic storage/retrieval environments. Similarly, the recitation of the use of a “first set of web functions” associated with the plurality of possible locations and “a second set of web functions” associated with the plurality of possible accounts, are also generically recited, as are the recitations to the data being received being “data structures”. However, the databases/data source, web functions, and data structures do nothing more than provide a generic environment in which the claimed method is performed. Such limitations do not amount to “significantly more”.1,2
Thus, dependent claim 3, which recites transmitting, to the database object, a query indicating the at least one object, and receiving, from the object database, a response to the query that indicates the plurality of possible locations, is nothing more than an insignificant field-of-use limitation, describing the context rather than a particular manner of achieving the result, as well as an insignificant extra-solution activity of retrieving data.3 Similarly, dependent claim 18 recites transmitting, to a user database, a query including an identifier associated with the user; and receiving, from the user database, a response to the query including the indication of the plurality of possible accounts.
The claims variously recite additional insignificant field-of-use limitations, describing the context rather than a particular manner of achieving the result, which include:
Independent Claims 1, 8, and 15, which recite “wherein the particular location and account pairing includes an identifier of a selected location and an identifier of a selected account” and “wherein the cost is based on a price associated with the location, reward points information, and traffic information”.
Dependent Claim 2, which recites that the indication of the at least one object comprises a scan of the at least one object. Similarly, dependent Claim 9 recites capturing, by the user device, a scan of the at least one object, wherein the indication of the at least one object is based on the scan.
Dependent Claim 4 recites that the set of endpoints are associated with the first set of web functions, and that the set of endpoints are used to communicate with the first set of web functions.
Dependent Claim 6 recites that at least one of the plurality of possible locations is virtual.
Dependent Claim 7 recites that the processors are configured to communicate with a third-party application programming interface to receive the directions.
Dependent Claim 10 recites transmitting a plurality of string descriptions that correspond to the plurality of possible accounts. Similarly, dependent Claim 11 recites transmitting a plurality of account numbers that correspond to the plurality of possible accounts.
Dependent Claim 12 recites transmitting, for each possible account in the plurality of possible accounts, a set of credentials for the possible account.
Dependent Claim 13 recites that a webpage is outputted, and the outputting of the webpage associated with the selected location and the at least one object is performed by using a web browser executed by the user device.
Dependent Claim 14 recites that the directions to the selected are outputted using a global position system application executed by the user device.
Independent Claim 15 recites that the traffic information includes a time associated with the request.
Dependent Claim 16 recites that the selected location is virtual, and the indication of the selected location comprises a uniform resource locator.
Dependent Claim 17 recites that the indication of the selected account comprises a visual indicator associated with the selected account.
Dependent Claim 19 recites that the indication of the starting location comprises a physical address or a set of coordinates.
Dependent Claim 20 recites that the time associated with the request is included in the request.
Furthermore, independent Claim 15 recites outputting an indication of the selected location and an indication of the selected account to the user device. This is an insignificant post-solution activity.
As such, the additional elements do not integrate the abstract idea into a practical application of that idea.
With respect to the well-understood, routine, and conventional elements, as stated previously above, the claims do not include any additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to the integration of the abstract idea into a practical application, the additional elements reciting the use of various computing software and hardware components amount to no more than mere instructions to apply the judicial exception using generic components. Mere instructions to apply an exception using generic computer components cannot provide an inventive concept.
Additionally, with regards to the claims’ recitation of receiving/transmitting data, accessing (i.e., retrieving) data, and outputting data, are all well-understood, routine, and conventional activities within the computing realm. See MPEP 2106.05(d)(II) (“Receiving or transmitting data over a network, e.g., using the Internet to gather data” with respect to transmitting data, receiving data, and obtaining data; “Storing and retrieving data in memory” with respect to retrieving data; and “Presenting offers and gathering statistics” with regards to the displaying/outputting steps). See also, e.g., Alice4 at p. 15 (finding that using a computer to obtain data and issue automated instructions are well-understood, routine, and conventional activities)5.
Even as an ordered combination, the claims as a whole do not contain any additional elements that amount to significantly more. The focus of the claims on certain methods of organizing human activity, including filtering data and remotely accessing user-specified information (which have previously been found by the courts to be abstract ideas). The most detail with respect to the claimed computer is primarily as a tool for receiving, transmitting, and retrieving data, as well as receiving inputs for storing/maintaining/updating. However, these are wholly conventional computing activities that are well-understood, routine, and conventional activities.
At this level of generality, the claims do no more than primarily describe the overall operations as a desired function or outcome, without providing any limiting detail beyond what is well-understood, routine, and conventional which would confine the claims to a particular solution to the identified problem. As a whole, the claims do not go beyond stating the relevant functions in general terms, with the most specificity being found in well-understood, routine, and conventional computing operations, i.e., are not an advance over conventional computing technologies.
Therefore, for at least the aforementioned reasons, the claims are rejected under 35 U.S.C. 101 for being directed to a judicial exception (i.e., an abstract idea) without significantly more.
Claim Rejections - 35 USC § 103
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.
Claims 1-4, 6, 8-11, 13, and 15-20 are rejected under 35 U.S.C. 103 as being unpatentable over Perks et al. (“Perks”) (US 2016/0104232 A1), in view of Nagaraj et al. (“Nagaraj”) (US 2017/0083930 A1), in further view of Bruss et al. (“Bruss”) (US 2020/0226460 A1).
Regarding claim 1: Perks teaches A system … to select from a plurality of possible locations, the system comprising:
one or more memories; and one or more processors, communicatively coupled to the one or more memories, configured to (Perks, [0063], where the disclosed system may be implemented as a computing device 900 that includes at least one processor 902 executing instructions stored in a memory 904, where the processor 902 may access the memory 904 by way of a system bus (i.e., “communicatively coupled to the one or more memories”)):
receive, from a user device, an indication of at least one object; determine, using an object database, the plurality of possible locations, wherein the plurality of possible locations are associated with the at least one object; communicate with a first set of web functions, associated with the plurality of possible locations, to receive a first plurality of data structures representing location information associated with the at least one object (Perks, [0021-0022], where responsive to receipt of an identification of a product or a list of products from a user, including, e.g., capturing an image of such a product, the identified products can be provided/transmitted to the search component 106, where the search component 106 performs a search over the data 104 in the data store 102 (i.e., “communicate with a first set of web functions”) based upon the identities of the product(s) selected by the user, as well as refine the search by performing the search over retail stores within a particular geographic location of interest to the user. The output of the search performed by the search component 106 can include identities of retail stores, as well as prices pertaining to the products at the retail stores that have the products in stock (see also Perks, [0060] (i.e., “representing location information associated with the at least one object”).
Although Perks does not appear to explicitly state that the information is received as data structures as claimed, one of ordinary skill in the art would have found it obvious to have modified Perks such that the information is provided as a plurality of data structures, as claimed, with the motivation of receiving information in a more organized/structured manner, which may facilitate increased efficiency in data processing);
receive, from the user device, an indication of a current location associated with the user device (Perks, [0023] and [0051], where a current location of a user may be determined from GPS. See also, e.g., Perks, [0026], where driving directions from a current location of the user may be generated (implying that the user’s current location is received));
transmit, to a traffic data source, a request for traffic information based on the current location and the plurality of possible locations; receive, from the traffic data source, the traffic information in response to the request (Perks, [0029], where the map generator component 108 can generate driving directions based upon current or predicted traffic conditions along the routes to be traveled by the user, where one or more sensors may be associated with vehicles traveling along roads near the retail stores (i.e., “traffic data source”), e.g., the retail stores depicted on the map generated by the map generator 108 (i.e., “plurality of possible locations”), which can provide data that indicates the flow of traffic along certain roads and/or at particular intersections. See Perks, [0023] and [0025], where the geographic area of interest comprising those retail stores was based on the determined current location of the user, and that the map generated by the map generator component 108 includes the retail stores in the geographic area of interest which may be displayed on a map (implying that the “request for traffic information” is based on “the current location and the plurality of possible locations”).
Although Perks does not appear to explicitly state that a singular traffic data source is utilized, one of ordinary skill in the art would have found it obvious to have modified Perks such that a single data source is queried with the motivation of greater efficiency in transmission and retrieval));
wherein the [system] is configured to minimize a cost associated with the at least one object, wherein the cost is based on a price associated with the location information … and the traffic information (Perks, [0027-0029], where the map generator component causes one or more optimizations to be performed with respect to a shopping trip of the user, and displays a detailed shopping itinerary to a user to purchase the products at a cheapest price (at a location) and in a least amount of time, including providing the user with directions that will cause the user to complete their shopping list in the least amount of time possible, those directions being based upon current or predicted traffic conditions);
provide the first plurality of data structures, … and the traffic information to a [user] to receive a particular location, wherein the particular location … includes an identifier of a selected location … from the plurality of possible locations…; output an indication of the selected location … to the user device; receive, from the user device, an acceptance of the selected location; and output, to the user device, directions to the selected location from the current location (Perks, [0026], where the map generator component 108 provides selectable hyperlinks corresponding to the graphical icons representative of the retail stores on the map. Data indicative of travel time (i.e., related to the claimed “traffic information”) from the current location of the user to the retail store represented by the graphical icon 204 may also be provided, as well as the price of the at least one product (Perks, [0041-0042] and [0055]). The user can select a selectable link corresponding to the retail store on the map, causing the map generator component 108 to output driving directions to that retail store to the user for display on the display 110. See also Perks, [0029], where the driving directions may be generated based upon current tor predicted traffic conditions along the routes to be traveled by the user).
Although Perks does not appear to explicitly state that the traffic information is provided to the user, one of ordinary skill in the art would have recognized that the travel time is related to traffic information (e.g., more traffic means more time would be required to travel). Therefore, one of ordinary skill in the art would have found it obvious to have supplied the traffic information as well (which is related to Perks’ travel time information) with the motivation of giving the user additional context for deciding which route they want to take (e.g., driving 10 miles across town during rush hour to purchase a $5.00 item might not be financially worthwhile if the same item can be purchased nearby for $6.006).
Perks does not appear to explicitly teach determine, using a user database, a plurality of possible accounts that are associated with a user of the user device; communicate with a second set of web functions, associated with the plurality of possible accounts, to receive a second plurality of data structures representing terms associated with the plurality of possible accounts; pre-process feature sets comprising location and account pairings to generate a minimum feature set, wherein the system is configured to filter the feature sets by rejecting location and account pairings that are incompatible due to a location not accepting an account type or due to a location not accepting a payment network; train a machine learning model using the minimum feature set, [wherein the] machine learning model [is configured to minimize a cost associated with the at least one object, wherein the cost is based on] reward points information; [providing] the second plurality of data structures to a machine learning model to receive [a particular location] and account pairing [from the plurality of possible locations] and the plurality of possible accounts, [wherein the particular location] and account pairing [includes an identifier of a selected location] and an identifier of a selected account; [and] output an indication of the selected account to the user device.
Nagaraj teaches determine, using a user database, a plurality of possible accounts that are associated with a user of the user device; communicate with a second set of web functions, associated with the plurality of possible accounts, to receive a second plurality of data structures representing terms associated with the plurality of possible accounts (Nagaraj, [0067], where stored information may be received from a card database 831, e.g., card-specific information such as a user’s known active cards. See also Nagaraj, [0069], where an optimization manager may collect relevant information including the user’s account or card types, membership information such as membership levels and account standing, rewards program enrollment or terms, rewards balances, user-defined preferences or objectives, and a user’s location information. The optimization manager may run a decision-making operation based at least in part on collected information to determine a number of optimum cards or accounts for use based on the information inputs. Results may be provided to dynamic priority system, which may prioritize the results based on known information.
Although Nagaraj does not appear to explicitly state that the information is received as data structures as claimed, one of ordinary skill in the art would have found it obvious to have modified Nagaraj such that the information is provided as a plurality of data structures, as claimed, with the motivation of receiving information in a more organized/structured manner, which may facilitate increased efficiency in data processing);
[wherein the system is configured to minimize a cost associated with the at least one object, wherein the cost is based on] reward points information (Nagaraj, [0056], where location information may be used during an optimization process such as to identify location-based offers, e.g., physical vendors that may offer promotional rewards rates or programs, and/or specific award programs which may correspond to an establishment at the geographic location. See Perks above with respect to “minimize a cost associated with the at least one object”);
[providing] the second plurality of data structures to [a computer system] to receive [a particular location] and account pairing [from the plurality of possible locations] and the plurality of possible accounts, [wherein the particular location] and account pairing [includes an identifier of a selected location] and an identifier of a selected account (Nagaraj, [0065-0066], where the system selects a rewards account for use by a dynamic priority subsystem 521 of priority ranking system 800. Priority ranking system 800 may use location information of the user device 550 from location identification 525 to optimize priority list 840, e.g., dynamic priority subsystem 521 may suggest a different priority list based on the current location of the user device 550 (i.e., “a particular location and account pairing from the plurality of possible locations and the plurality of possible accounts”, which implies that there is some manner of being able to identify both the location and account, e.g., an “identifier” is implied in order to distinguish the different locations and accounts, i.e., “wherein the particular location and account pairing includes an identifier of a selected location and an identifier of a selected account”). This information may then be used by the dynamic priority subsystem 521 to determine an ordered-list ranking 840 of cards, accounts, programs, rewards types, etc., and may then make this ordered-list available for use in selecting a specific entity for use, e.g., the highest-priority card or account for a particular selection may be automatically selected); [and]
output an indication of the selected account to the user device (Nagaraj, [0069], where a list of prioritized account results may be presented to the user via an interface on their device, including an indication of a highest-priority result to encourage the user to select the ideal card or account for the current transaction).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Perks and Nagaraj (hereinafter “Perks as modified”) with the motivation of intelligently determining ideal accounts to use on a per-transaction basis for the maximum benefit of the user as well as assisting users in reaching specific program goals in an economical fashion (Nagaraj, [0006]), including optimizing use of rewards at certain location places thereby optimizing the economic benefit to the user (see, e.g., Nagaraj, [0056]), as well as presenting the selected account in order to provide information to the user regarding the account being used to optimize the search results while enabling the user to select alternative accounts if possible (Nagaraj, [0070]).
Perks as modified does not appear to explicitly state that the disclosed systems [uses] machine learning, that the inputs were provided to a machine learning model, pre-process feature sets comprising location and account pairings to generate a minimum feature set, wherein the system is configured to filter the feature sets by rejecting location and account pairings that are incompatible due to a location not accepting an account type or due to a location not accepting a payment network; [and] train a machine learning model using the minimum feature set, [wherein the] machine learning model [is configured to minimize a cost associated with the at least one object, wherein the cost is based on various attributes].
Bruss teaches that the disclosed systems [uses] machine learning, that the inputs were provided to a machine learning model, pre-process feature sets comprising location and account pairings to generate a … feature set, wherein the system is configured to filter the feature sets by rejecting location and account pairings that are incompatible due to a location not accepting an account type or due to a location not accepting a payment network; [and] train a machine learning model using the … feature set, [wherein the] machine learning model [is configured to minimize a cost associated with the at least one object, wherein the cost is based on various attributes] (Bruss, [0021], where related entities and/or transactions may be subject to a contextual constraint, where transactions between a customer and merchants A and B may be considered to be within the same context and used as positive samples, e.g., because the customer made purchases at merchants A and B within a time threshold. However, if the customer makes purchases at merchants A and B months apart, the transactions with merchants A and B may not be considered as related, and are therefore not used as positive samples in the training data (i.e., “pre-process feature sets comprising location and account pairings to generate a … feature set, wherein the system is configured to filter the feature sets by rejecting location and account pairings that are incompatible…”).
See Bruss, [0019-0020] and [0040], where the transaction application 103 trains one or more neural networks 105 using the ML algorithm 104 and the training data 107, where the training data 107 includes positive samples from the transaction data 121 and/or the graphs 110, where a positive sample is an actual transaction that has a contextual relationship between a pair of entities in the transaction data 121 and/or the graphs 110. The network graph 110 that is generated from the transaction data, may have been generated using one or more ETL functions, which generally include standardizing the transaction data 121 according to one or more formats (i.e., “pre-process feature sets”).
See Bruss, [0029], where when training of the neural network is complete, the neural network may be used to process new transactions received as runtime data, the runtime data including one or more new transactions received as transaction data 121. The transaction application applies an embedding function to the received transaction in the runtime data that generates a vector representation of the transaction, e.g., generating a vector based on the account ID, merchant name, location, etc. (implying that the neural network was trained on the location and account information, i.e., “pairings”, as claimed, as the neural network is able to provide recommendations based on location, similar accounts, etc. (see, e.g., Bruss, [0046])). The neural network 105 may then process the vector for the transaction in the runtime data 108 (i.e., “using machine learning”, and “inputs are provided to a machine learning model”).
See Perks and Nagaraj above with respect to “minimiz[ing] a cost associated with the least one object, wherein the cost is based on a price associated with the location information, reward points information, and the traffic information”; see Bruss above with respect to the use of a “machine learning model”).
Although Bruss does not appear to explicitly state that the type of information (the location and account pairings that are incompatible) relates to (i.e., due to) a location not accepting an account type or due to a location not accepting a payment network, the claimed invention does not distinguish over the prior art because the differences in the claim limitations and the prior art’s disclosure are only found in the nonfunctional descriptive material and are not functionally involved in the steps recited. The rest of the claimed steps involving the machine learning algorithm, e.g., generating the training set, training the machine learning algorithm on the training set, etc., would have been performed the same regardless of the specific data involved (i.e., incompatibility due to a location not accepting an account type or a location not accepting a payment network as claimed, or other contextual transactional reasons, as disclosed by Bruss). Thus, this descriptive material will not distinguish the claimed invention from the prior art in terms of patentability. See In re Gulack, 703 F.2d 1381, 1385, 217 USPQ2d 401, 404 (Fed. Cir. 1983); In re Lowry, 32 F.3d 1579, 32 USPQ2d 1031 (Fed. Cir. 1994).
Therefore, it would have been obvious to a person of ordinary skill in the art to have referred to Bruss’s teachings in making the claimed invention, because such data does not functionally relate to the steps in the method claimed and because the subjective interpretation of the data does not patentably distinguish the claimed invention over the prior art.
Furthermore, although Bruss does not appear to explicitly state that the pre-processing comprises, e.g., dimensionality reduction, that results in a minimum feature set as claimed, Bruss discloses having the machine learning model learn on a low-dimensional dense representation of each entity in a network graph of transactions (see, e.g., Bruss, [0011]). Therefore, one of ordinary skill in the art would have been suggested by Bruss’ disclosure to perform dimensionality reduction as part of the pre-processing step (thereby resulting in a minimum feature set for Bruss’ machine learning models to train on), with the motivation of reducing storage resource consumption (e.g., by only storing a minimum amount of data needed for training).7,8
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Perks as modified and Bruss (hereinafter “Perks as modified”) with the motivation of (1) utilizing machine learning models in order to provide systems the ability to automatically learn and improve from examples, data sets, direct experience, or instructions without being explicitly programmed, thereby allowing the computers to learn automatically without human intervention or assistance and adjust actions accordingly9, (2) facilitating streamlined process to generate relevant recommendations to users by taking into account certain relevant contextual information (i.e., greater convenience to users, who do not need to double-check that the recommendations are relevant), as well as potentially providing greater accuracy from learning from historical data; and (3) training the machine learning model on reduced dimensionality data, which improves the computational efficiency for training the machine learning model.10,11
Regarding claim 2: Perks as modified teaches The system of claim 1, wherein the indication of the at least one object comprises a scan of the at least one object (Perks, [0022], where the user may be utilizing a mobile computing device capable of capturing images of bar codes or performing scans of bar codes, where the bar codes can identify certain products and can be provided to the search component 106, e.g., the product identities derived from the bar codes can be provided to the search component 106).
Regarding claim 3: Perks as modified teaches The system of claim 1, wherein the one or more processors, to determine the plurality of possible locations, are configured to:
transmit, to the object database, a query indicating the at least one object; and receive, from the object database, a response to the query that indicates the plurality of possible locations (Perks, [0022], where the identified products can be provided to the search component 106, where the search component 106 performs a search over the data 104 in the data store 102 based upon the identities of the product(s) selected by the user, as well as refine the search by performing the search over retail stores within a particular geographic location of interest to the user. The output of the search performed by the search component 106 can include identities of retail stores).
Regarding claim 4: Perks as modified teaches The system of claim 1, wherein the one or more processors are configured to:
determine a set of endpoints associated with the first set of web functions, based on the plurality of possible locations, wherein the set of endpoints are used to communicate with the first set of web functions (Perks, [0020] and [0043-0044], where a data store 102 can include data 104 comprising product identities, price information pertaining to the products, locations of retail stores, and inventory data of the retail stores. Data 104 can include inventory data with respect to one or more retail stores that are at particular geographic locations (i.e., “set of endpoints”), where the aggregator component may obtain inventory data with respect to each of the stores/online providers, or search the inventory of the retail stores directly without aggregating such data at a central source (i.e., “wherein the set of endpoints are used to communicate with the first set of web functions”). See also, e.g., Perks, [0052] and [0058], where inventory of retail stores in a geographic area of interest are searched for the at least one product identified at 504 (i.e., “first set of web functions”), and if the product is in stock at the one or more retail stores (i.e., “set of endpoints…based on the plurality of possible locations”), then the product location and price is displayed on a map (i.e., “determine a set of endpoints associated with the first set of web functions”)).
Regarding claim 6: Perks as modified teaches The system of claim 1, wherein at least one of the plurality of possible locations is virtual (Perks, [0036], where the map generator component 108 can generate a map such that the retail stores are represented in the map by graphical icons to show to the user which retail stores in the geographical area of interest to the user have one or more products in the list of products provided by the user. The user may then select one or more selectable links corresponding to a retail store, which can cause other products in stock at the retail store to be presented to the user that are similar to a product in the list or accessories to the products in the shopping list provided by the user (i.e., “at least one of the plurality of possible locations is virtual”)).
Regarding claim 8: Perks teaches A method … to select from a plurality of possible locations, comprising:
transmitting, from a user device and to a selection system, an indication of at least one object (Perks, [0021-0022], where responsive to receipt of an identification of a product or a list of products from a user, including, e.g., capturing an image of such a product, the identified products can be provided/transmitted to the search component 106, where the search component 106 performs a search over the data 104 in the data store 102);
transmitting, from the user device and to the selection system, an indication of a location (Perks, [0051], where data that indicates a geographic area of interest to the user is received, where this data can be received from a GPS corresponding to a mobile computing device of the user, thereby indicating a current location of the user); and
receiving, from the selection system and at the user device, a particular location … from the plurality of possible locations …, based on [presenting to the user] using the indication of at least one object [and] the indication of the location …, wherein the particular location … includes an identifier of a selected location … (Perks, [0026], where the map generator component 108 provides selectable hyperlinks corresponding to the graphical icons representative of the retail stores on the map. The user can select a selectable link corresponding to the retail store on the map, causing the map generator component 108 to output driving directions to that retail store to the user for display on the display 110), and wherein the [system] is configured to minimize a cost associated with the at least one object, wherein the cost is based on a price associated with the location … and traffic information (Perks, [0027-0029], where the map generator component causes one or more optimizations to be performed with respect to a shopping trip of the user, and displays a detailed shopping itinerary to a user to purchase the products at a cheapest price (at a location) and in a least amount of time, including providing the user with directions that will cause the user to complete their shopping list in the least amount of time possible, those directions being based upon current or predicted traffic conditions).
Perks as modified does not appear to explicitly teach transmitting, from the user device and to the selection system, an indication of a plurality of possible accounts that are associated with a user of the user device; [receiving a particular location] and account pairing [from the plurality of possible locations] and the plurality of possible accounts based on application of a machine learning model using the indication of the plurality of possible accounts, [wherein the particular location] and account pairing [includes an identifier of a selected location] and an identifier of a selected account, and wherein the machine learning model is trained by a machine learning system using a minimum feature set generated based on pre-processing feature sets of location and account pairings, and wherein the machine learning system is configured to filter the feature sets by rejecting location and account pairings that are incompatible due to a location not accepting an account type or due to a location not accepting a payment network, and [wherein the] machine learning model [is configured to minimize a cost associated with the at least one object, wherein the cost is based on] reward points information.
Nagaraj teaches transmitting, from the user device and to the selection system, an indication of a plurality of possible accounts that are associated with a user of the user device (Nagaraj, [0067], where stored information may be received from a card database 831, e.g., card-specific information such as a user’s known active cards. See also Nagaraj, [0069], where an optimization manager may collect relevant information including the user’s account or card types, membership information such as membership levels and account standing, rewards program enrollment or terms, rewards balances, user-defined preferences or objectives, and a user’s location information. The optimization manager may run a decision-making operation based at least in part on collected information to determine a number of optimum cards or accounts for use based on the information inputs. Results may be provided to dynamic priority system, which may prioritize the results based on known information);
[receiving a particular location] and account pairing [from the plurality of possible locations] and the plurality of possible accounts based on an application of [the computing system] using the indication of the plurality of possible accounts, [wherein the particular location] and account pairing [includes an identifier of a selected location] and an identifier of a selected account (Nagaraj, [0065-0066], where the system selects a rewards account for use by a dynamic priority subsystem 521 of priority ranking system 800. Priority ranking system 800 may use location information of the user device 550 from location identification 525 to optimize priority list 840, e.g., dynamic priority subsystem 521 may suggest a different priority list based on the current location of the user device 550 (i.e., “a particular location and account pairing from the plurality of possible locations and the plurality of possible accounts”, which implies that there is some manner of being able to identify both the location and account, e.g., an “identifier” is implied in order to distinguish the different locations and accounts, i.e., “wherein the particular location and account pairing includes an identifier of a selected location and an identifier of a selected account”). This information may then be used by the dynamic priority subsystem 521 to determine an ordered-list ranking 840 of cards, accounts, programs, rewards types, etc., and may then make this ordered-list available for use in selecting a specific entity for use, e.g., the highest-priority card or account for a particular selection may be automatically selected), [and]
[wherein the system is configured to minimize a cost associated with the at least one object, wherein the cost is based on] reward points information (Nagaraj, [0056], where location information may be used during an optimization process such as to identify location-based offers, e.g., physical vendors that may offer promotional rewards rates or programs, and/or specific award programs which may correspond to an establishment at the geographic location. See Perks above with respect to “minimize a cost associated with the at least one object”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Perks and Nagaraj (hereinafter “Perks as modified”) with the motivation of intelligently determining ideal accounts to use on a per-transaction basis for the maximum benefit of the user as well as assisting users in reaching specific program goals in an economical fashion (Nagaraj, [0006]), including optimizing use of rewards at certain location places thereby optimizing the economic benefit to the user (see, e.g., Nagaraj, [0056]), as well as presenting the selected account in order to provide information to the user regarding the account being used to optimize the search results while enabling the user to select alternative accounts if possible (Nagaraj, [0070]).
Perks as modified does not appear to explicitly teach the method [uses] machine learning and “based on application of a machine learning model”; [and] wherein the particular location and account pairing includes an identifier of a selected location and an identifier of a selected account, and wherein the machine learning model is trained by a machine learning system using a minimum feature set generated based on pre-processing feature sets of location and account pairings, and wherein the machine learning system is configured to filter the feature sets by rejecting location and account pairings that are incompatible due to a location not accepting an account type or due to a location not accepting a payment network, and [wherein the] machine learning model [is configured to minimize a cost associated with the at least one object, wherein the cost is based on various attributes].
Bruss teaches [uses] machine learning and “based on application of a machine learning model”; [and] wherein the particular location and account pairing includes an identifier of a selected location and an identifier of a selected account, and wherein the machine learning model is trained by a machine learning system using a … feature set generated based on pre-processing feature sets of location and account pairings, and wherein the machine learning system is configured to filter the feature sets by rejecting location and account pairings that are incompatible due to a location not accepting an account type or due to a location not accepting a payment network, and [wherein the] machine learning model [is configured to minimize a cost associated with the at least one object, wherein the cost is based on various attributes] (Bruss, [0021], where related entities and/or transactions may be subject to a contextual constraint, where transactions between a customer and merchants A and B may be considered to be within the same context and used as positive samples, e.g., because the customer made purchases at merchants A and B within a time threshold. However, if the customer makes purchases at merchants A and B months apart, the transactions with merchants A and B may not be considered as related, and are therefore not used as positive samples in the training data (i.e., “pre-process feature sets comprising location and account pairings to generate a … feature set, wherein the system is configured to filter the feature sets by rejecting location and account pairings that are incompatible…”).
See Bruss, [0020], where the transaction application 103 trains one or more neural networks 105 using the ML algorithm 104 and the training data 107, where the training data 107 includes positive samples from the transaction data 121 and/or the graphs 110, where a positive sample is an actual transaction that has a contextual relationship between a pair of entities in the transaction data 121 and/or the graphs 110. See Bruss, [0029], where when training of the neural network is complete, the neural network may be used to process new transactions received as runtime data, the runtime data including one or more new transactions received as transaction data 121. The transaction application applies an embedding function to the received transaction in the runtime data that generates a vector representation of the transaction, e.g., generating a vector based on the account ID, merchant name, location, etc. (implying that the neural network was trained on the location and account information, i.e., “pairings”, as claimed, as the neural network is able to provide recommendations based on location, similar accounts, etc. (see, e.g., Bruss, [0046])). The neural network 105 may then process the vector for the transaction in the runtime data 108 (i.e., “using machine learning”, and “inputs are provided to a machine learning model”).
See Perks and Nagaraj above with respect to “minimiz[ing] a cost associated with the least one object, wherein the cost is based on a price associated with the location information, reward points information, and the traffic information”; see Bruss above with respect to the use of a “machine learning model”).
Although Bruss does not appear to explicitly state that the type of information (the location and account pairings that are incompatible) relates to (i.e., due to) a location not accepting an account type or due to a location not accepting a payment network, the claimed invention does not distinguish over the prior art because the differences in the claim limitations and the prior art’s disclosure are only found in the nonfunctional descriptive material and are not functionally involved in the steps recited. The rest of the claimed steps involving the machine learning algorithm, e.g., generating the training set, training the machine learning algorithm on the training set, etc., would have been performed the same regardless of the specific data involved (i.e., incompatibility due to a location not accepting an account type or a location not accepting a payment network as claimed, or other contextual transactional reasons, as disclosed by Bruss). Thus, this descriptive material will not distinguish the claimed invention from the prior art in terms of patentability. See In re Gulack, 703 F.2d 1381, 1385, 217 USPQ2d 401, 404 (Fed. Cir. 1983); In re Lowry, 32 F.3d 1579, 32 USPQ2d 1031 (Fed. Cir. 1994).
Therefore, it would have been obvious to a person of ordinary skill in the art to have referred to Bruss’s teachings in making the claimed invention, because such data does not functionally relate to the steps in the method claimed and because the subjective interpretation of the data does not patentably distinguish the claimed invention over the prior art.
Furthermore, although Bruss does not appear to explicitly state that the pre-processing comprises, e.g., dimensionality reduction, that results in a minimum feature set as claimed, Bruss discloses having the machine learning model learn on a low-dimensional dense representation of each entity in a network graph of transactions (see, e.g., Bruss, [0011]). Therefore, one of ordinary skill in the art would have been suggested by Bruss’ disclosure to perform dimensionality reduction as part of the pre-processing step (thereby resulting in a minimum feature set for Bruss’ machine learning models to train on), with the motivation of reducing storage resource consumption (e.g., by only storing a minimum amount of data needed for training).12,13
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Perks as modified and Bruss (hereinafter “Perks as modified”) with the motivation of (1) utilizing machine learning models in order to provide systems the ability to automatically learn and improve from examples, data sets, direct experience, or instructions without being explicitly programmed, thereby allowing the computers to learn automatically without human intervention or assistance and adjust actions accordingly14, (2) facilitating streamlined process to generate relevant recommendations to users by taking into account certain relevant contextual information (i.e., greater convenience to users, who do not need to double-check that the recommendations are relevant), as well as potentially providing greater accuracy from learning from historical data; and (3) training the machine learning model on reduced dimensionality data, which improves the computational efficiency for training the machine learning model.15,16
Regarding claim 9: Perks as modified teaches The method of claim 8, further comprising:
capturing, by the user device, a scan of the at least one object, wherein the indication of the at least one object is based on the scan (Perks, [0022], where the user may be utilizing a mobile computing device capable of capturing images of bar codes or performing scans of bar codes, where the bar codes can identify certain products and can be provided to the search component 106, e.g., the product identities derived from the bar codes can be provided to the search component 106).
Regarding claim 10: Perks as modified teaches The method of claim 8, wherein transmitting the indication of the plurality of possible accounts comprises:
transmitting a plurality of string descriptions that correspond to the plurality of possible accounts (Nagaraj, [0067], where stored information may be received from a card database 831, e.g., card-specific information such as a user’s known active cards. See Nagaraj, [0073] and [FIG. 12], where the displayed accounts are a plurality of string descriptions, e.g., “American Express Premier Rewards Gold Card”, “Capital One Venture Rewards”, “Chase Freedom Card”, and “Citi Debit”. This implies that such string descriptions are associated with the card, thus implying that the transmitted information corresponding to the plurality of possible accounts includes “a plurality of string descriptions”, as claimed).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Perks and Nagaraj with the motivation of enabling the system to associate string descriptions of cards for outputting to the user, e.g., for selection purposes as seen in Nagaraj, [FIG. 12], which increases understandability (e.g., a person may more easily identify their accounts using the credit card brand and name).
Regarding claim 11: Perks as modified teaches The method of claim 8, wherein transmitting the indication of the plurality of possible accounts comprises:
transmitting a plurality of account numbers that correspond to the plurality of possible accounts (Nagaraj, [0067], where stored information may be received from a card database 831, e.g., card-specific information such as a user’s known active cards. Note that one of ordinary skill in the art would have recognized that credit cards, such as those disclosed by Nagaraj, include numbers for identifying the credit cards17,18 (i.e., the “account numbers” as claimed)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Perks and Nagaraj with the motivation of uniquely identifying the accounts, thereby enabling the system to link any transactions, rewards, offers, etc. specifically with a particular account.
Regarding claim 13: Perks as modified teaches The method of claim 8, wherein the selected location is virtual, and the method further comprises: receiving an interaction with an indication of the selected location; and outputting, using a web browser executed by the user device, a webpage associated with the selected location and the at least one object (Perks, [0036], where the map generator component 108 can generate a map such that the retail stores are represented in the map by graphical icons to show to the user which retail stores in the geographical area of interest to the user have one or more products in the list of products provided by the user. The user may then select one or more selectable links corresponding to a retail store, which can cause other products in stock at the retail store to be presented to the user that are similar to a product in the list or accessories to the products in the shopping list provided by the user (i.e., “at least one of the plurality of possible locations is virtual”).
See Perks, [0021], where the graphical user interface can be provided to the user in an Internet browser when the user directs the browser to a particular website (i.e., “outputting, using a web browser executed by the user device, a webpage associated with the selected location and the at least one object”)).
Regarding claim 15: Perks as modified teaches A non-transitory computer-readable medium storing a set of instructions … to select from a plurality of possible locations, the set of instructions comprising: one or more instructions that, when executed by one or more processors of a device, cause the device to (Perks, [0049], where the disclosed steps may be computer-executable instructions implemented by one or more processors stored on a non-transitory computer-readable medium):
receive, from a user device, a request indicating at least one object; receive an indication of the plurality of possible locations, wherein the plurality of possible locations are associated with the at least one object; receive, for each possible location in the plurality of possible locations, location information associated with the at least one object (Perks, [0021-0022], where responsive to receipt of an identification of a product or a list of products from a user, including, e.g., capturing an image of such a product, the identified products can be provided/transmitted to the search component 106, where the search component 106 performs a search over the data 104 in the data store 102 based upon the identities of the product(s) selected by the user, as well as refine the search by performing the search over retail stores within a particular geographic location of interest to the user. The output of the search performed by the search component 106 can include identities of retail stores, as well as prices pertaining to the products at the retail stores that have the products in stock (see also Perks, [0060] (i.e., “location information associated with the at least one object”));
receive, from the user device, an indication of a starting location (Perks, [0023] and [0051], where a current location of a user may be determined from GPS. See also, e.g., Perks, [0026], where driving directions from a current location of the user may be generated (implying that the user’s current location is received));
receive traffic information based on the starting location, the plurality of possible locations, and a time associated with the request (Perks, [0029], where the map generator component 108 can generate driving directions based upon current or predicted traffic conditions along the routes to be traveled by the user, where one or more sensors may be associated with vehicles traveling along roads near the retail stores (i.e., “traffic data source”), e.g., the retail stores depicted on the map generated by the map generator 108 (i.e., “plurality of possible locations”), which can provide data that indicates the flow of traffic along certain roads and/or at particular intersections. See Perks, [0023] and [0025], where the geographic area of interest comprising those retail stores was based on the determined current location of the user, and that the map generated by the map generator component 108 includes the retail stores in the geographic area of interest which may be displayed on a map (implying that the “request for traffic information” is based on “the current location and the plurality of possible locations”).
Although Perks does not appear to explicitly state that the traffic information includes “a time associated with the request” as claimed, Perks discloses that traffic conditions may be predicted, and that historical traffic patterns can be utilized, suggesting the use of time as a factor. Therefore, one of ordinary skill in the art would have found it obvious to have explicitly included a time associated with the request with the motivation of optimizing driving directions provided to a user, e.g., when the user will actually be traveling to the retail stores depicted on the map (Perks, [0029]), as traffic tends to occur at certain times, making time an important component for making predictions as to traffic conditions);
wherein the [system] is configured to minimize a cost associated with the at least one object, wherein the cost is based on a price associated with the location information … and the traffic information (Perks, [0027-0029], where the map generator component causes one or more optimizations to be performed with respect to a shopping trip of the user, and displays a detailed shopping itinerary to a user to purchase the products at a cheapest price (at a location) and in a least amount of time, including providing the user with directions that will cause the user to complete their shopping list in the least amount of time possible, those directions being based upon current or predicted traffic conditions);
provide the location information … and the traffic information to a machine learning model to receive a particular location … from the plurality of possible locations …, wherein the particular location … includes an identifier of a selected location …; and output an indication of the selected location … to the user device (Perks, [0026], where the map generator component 108 provides selectable hyperlinks corresponding to the graphical icons representative of the retail stores on the map. Data indicative of travel time (i.e., related to the claimed “traffic information”) from the current location of the user to the retail store represented by the graphical icon 204 may also be provided, as well as the price of the at least one product (Perks, [0041-0042] and [0055]). The user can select a selectable link corresponding to the retail store on the map, causing the map generator component 108 to output driving directions to that retail store to the user for display on the display 110. See also Perks, [0029], where the driving directions may be generated based upon current tor predicted traffic conditions along the routes to be traveled by the user).
Perks does not appear to explicitly teach [using] machine learning; receive an indication of a plurality of possible accounts that are associated with a user of the user device; receive, for each possible account in the plurality of possible accounts, terms associated with the plurality of possible accounts; process feature sets comprising location and account pairings to generate a minimum feature set, wherein the device is configured to filter the feature sets by rejecting location and account pairings that are incompatible due to a location not accepting an account type or due to a location not accepting a payment network; train a machine learning model using the minimum feature set, [wherein the] machine learning model [is configured to minimize a cost associated with the at least one object, wherein the cost is based on] reward points information; [provide] the terms to the machine learning model [to receive a particular location] and account pairing from the plurality of possible locations and the plurality of possible accounts, [wherein the particular location] and account pairing includes an identifier of a selected location] and an identifier of a selected account; and [output] an indication of the selected account [to the user device].
Nagaraj teaches receive an indication of a plurality of possible accounts that are associated with a user of the user device; receive, for each possible account in the plurality of possible accounts, terms associated with the plurality of possible accounts (Nagaraj, [0067], where stored information may be received from a card database 831, e.g., card-specific information such as a user’s known active cards. See also Nagaraj, [0069], where an optimization manager may collect relevant information including the user’s account or card types, membership information such as membership levels and account standing, rewards program enrollment or terms, rewards balances, user-defined preferences or objectives, and a user’s location information. The optimization manager may run a decision-making operation based at least in part on collected information to determine a number of optimum cards or accounts for use based on the information inputs. Results may be provided to dynamic priority system, which may prioritize the results based on known information);
[wherein the system is configured to minimize a cost associated with the at least one object, wherein the cost is based on] reward points information (Nagaraj, [0056], where location information may be used during an optimization process such as to identify location-based offers, e.g., physical vendors that may offer promotional rewards rates or programs, and/or specific award programs which may correspond to an establishment at the geographic location. See Perks above with respect to “minimize a cost associated with the at least one object”);
[provide] the terms to [a computing system to receive a particular location] and account pairing from the plurality of possible locations and the plurality of possible accounts, [wherein the particular location] and account pairing includes an identifier of a selected location] and an identifier of a selected account (Nagaraj, [0065-0066], where the system selects a rewards account for use by a dynamic priority subsystem 521 of priority ranking system 800. Priority ranking system 800 may use location information of the user device 550 from location identification 525 to optimize priority list 840, e.g., dynamic priority subsystem 521 may suggest a different priority list based on the current location of the user device 550 (i.e., “a particular location and account pairing from the plurality of possible locations and the plurality of possible accounts”, which implies that there is some manner of being able to identify both the location and account, e.g., an “identifier” is implied in order to distinguish the different locations and accounts, i.e., “wherein the particular location and account pairing includes an identifier of a selected location and an identifier of a selected account”). This information may then be used by the dynamic priority subsystem 521 to determine an ordered-list ranking 840 of cards, accounts, programs, rewards types, etc., and may then make this ordered-list available for use in selecting a specific entity for use, e.g., the highest-priority card or account for a particular selection may be automatically selected); and
[output] an indication of the selected account [to the user device] (Nagaraj, [0069], where a list of prioritized account results may be presented to the user via an interface on their device, including an indication of a highest-priority result to encourage the user to select the ideal card or account for the current transaction).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Perks and Nagaraj (hereinafter “Perks as modified”) with the motivation of intelligently determining ideal accounts to use on a per-transaction basis for the maximum benefit of the user as well as assisting users in reaching specific program goals in an economical fashion (Nagaraj, [0006]), as well as presenting the selected account in order to provide information to the user regarding the account being used to optimize the search results while enabling the user to select alternative accounts if possible (Nagaraj, [0070]).
Perks as modified does not appear to explicitly teach [using] machine learning; [providing inputs to] the machine learning model; process feature sets comprising location and account pairings to generate a minimum feature set, wherein the device is configured to filter the feature sets by rejecting location and account pairings that are incompatible due to a location not accepting an account type or due to a location not accepting a payment network; train a machine learning model using the minimum feature set, [wherein the] machine learning model [is configured to minimize a cost associated with the at least one object, wherein the cost is based on various attributes]; wherein the particular location and account pairing includes an identifier of a selected location and an identifier of a selected account.
Bruss teaches [using] machine learning; [providing inputs to] a machine learning model; process feature sets comprising location and account pairings to generate a … feature set, wherein the device is configured to filter the feature sets by rejecting location and account pairings that are incompatible due to a location not accepting an account type or due to a location not accepting a payment network; train a machine learning model using the … feature set, [wherein the] machine learning model [is configured to minimize a cost associated with the at least one object, wherein the cost is based on various attributes]; wherein the particular location and account pairing includes an identifier of a selected location and an identifier of a selected account (Bruss, [0021], where related entities and/or transactions may be subject to a contextual constraint, where transactions between a customer and merchants A and B may be considered to be within the same context and used as positive samples, e.g., because the customer made purchases at merchants A and B within a time threshold. However, if the customer makes purchases at merchants A and B months apart, the transactions with merchants A and B may not be considered as related, and are therefore not used as positive samples in the training data (i.e., “pre-process feature sets comprising location and account pairings to generate a … feature set, wherein the system is configured to filter the feature sets by rejecting location and account pairings that are incompatible…”).
See Bruss, [0020], where the transaction application 103 trains one or more neural networks 105 using the ML algorithm 104 and the training data 107, where the training data 107 includes positive samples from the transaction data 121 and/or the graphs 110, where a positive sample is an actual transaction that has a contextual relationship between a pair of entities in the transaction data 121 and/or the graphs 110. See Bruss, [0029], where when training of the neural network is complete, the neural network may be used to process new transactions received as runtime data, the runtime data including one or more new transactions received as transaction data 121. The transaction application applies an embedding function to the received transaction in the runtime data that generates a vector representation of the transaction, e.g., generating a vector based on the account ID, merchant name, location, etc. (implying that the neural network was trained on the location and account information, i.e., “pairings”, as claimed, as the neural network is able to provide recommendations based on location, similar accounts, etc. (see, e.g., Bruss, [0046])). The neural network 105 may then process the vector for the transaction in the runtime data 108 (i.e., “using machine learning”, and “inputs are provided to a machine learning model”).
See Perks and Nagaraj above with respect to “minimiz[ing] a cost associated with the least one object, wherein the cost is based on a price associated with the location information, reward points information, and the traffic information”; see Bruss above with respect to the use of a “machine learning model”).
Although Bruss does not appear to explicitly state that the type of information (the location and account pairings that are incompatible) relates to (i.e., due to) a location not accepting an account type or due to a location not accepting a payment network, the claimed invention does not distinguish over the prior art because the differences in the claim limitations and the prior art’s disclosure are only found in the nonfunctional descriptive material and are not functionally involved in the steps recited. The rest of the claimed steps involving the machine learning algorithm, e.g., generating the training set, training the machine learning algorithm on the training set, etc., would have been performed the same regardless of the specific data involved (i.e., incompatibility due to a location not accepting an account type or a location not accepting a payment network as claimed, or other contextual transactional reasons, as disclosed by Bruss). Thus, this descriptive material will not distinguish the claimed invention from the prior art in terms of patentability. See In re Gulack, 703 F.2d 1381, 1385, 217 USPQ2d 401, 404 (Fed. Cir. 1983); In re Lowry, 32 F.3d 1579, 32 USPQ2d 1031 (Fed. Cir. 1994).
Therefore, it would have been obvious to a person of ordinary skill in the art to have referred to Bruss’s teachings in making the claimed invention, because such data does not functionally relate to the steps in the method claimed and because the subjective interpretation of the data does not patentably distinguish the claimed invention over the prior art.
Furthermore, although Bruss does not appear to explicitly state that the pre-processing comprises, e.g., dimensionality reduction, that results in a minimum feature set as claimed, Bruss discloses having the machine learning model learn on a low-dimensional dense representation of each entity in a network graph of transactions (see, e.g., Bruss, [0011]). Therefore, one of ordinary skill in the art would have been suggested by Bruss’ disclosure to perform dimensionality reduction as part of the pre-processing step (thereby resulting in a minimum feature set for Bruss’ machine learning models to train on), with the motivation of reducing storage resource consumption (e.g., by only storing a minimum amount of data needed for training).19,20
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Perks as modified and Bruss (hereinafter “Perks as modified”) with the motivation of (1) utilizing machine learning models in order to provide systems the ability to automatically learn and improve from examples, data sets, direct experience, or instructions without being explicitly programmed, thereby allowing the computers to learn automatically without human intervention or assistance and adjust actions accordingly21, (2) facilitating streamlined process to generate relevant recommendations to users by taking into account certain relevant contextual information (i.e., greater convenience to users, who do not need to double-check that the recommendations are relevant), as well as potentially providing greater accuracy from learning from historical data; and (3) training the machine learning model on reduced dimensionality data, which improves the computational efficiency for training the machine learning model.22,23
Regarding claim 16: Perks as modified teaches The non-transitory computer-readable medium of claim 15, wherein the selected location is virtual, and the indication of the selected location comprises a uniform resource locator (Perks, [0036], where the map generator component 108 can generate a map such that the retail stores are represented in the map by graphical icons to show to the user which retail stores in the geographical area of interest to the user have one or more products in the list of products provided by the user. The user may then select one or more selectable links (i.e., “uniform resource locator”) corresponding to a retail store, which can cause other products in stock at the retail store to be presented to the user that are similar to a product in the list or accessories to the products in the shopping list provided by the user (i.e., “the selected location is virtual”)).
Regarding claim 17: Perks as modified teaches The non-transitory computer-readable medium of claim 15, wherein the indication of the selected account comprises a visual indicator associated with the selected account (Nagaraj, [FIG. 12], where the selected account 1225 is shown under the banner “Selected Card:” in a separate section of the screen compared to the “Other Options” listing the other possible accounts).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Perks and Nagaraj with the motivation of visually distinguishing (and thus making more clear to the user) which of the accounts was selected.
Regarding claim 18: Perks as modified teaches The non-transitory computer-readable medium of claim 15, wherein the one or more instructions, that cause the device to receive the indication of the plurality of possible accounts, cause the device to:
transmit, to a user database, a query …; and receive, from the user database, a response to the query including the indication of the plurality of possible accounts (Nagaraj, [0065], where the system may receive a plurality of information from a variety of sources including the databases from a rewards provider 810, e.g., using a structured query language SQL (see Nagaraj, [0045]) (i.e., “transmit, to a user database, a query…”). See also Nagaraj, [0067], where stored information may be received from a card database 831, e.g., card-specific information such as a user’s known active cards (i.e., “receive, from the user database, a response to the query including an indication of the plurality of possible accounts”), as well as an expiry database 832 that provides a plurality of card or account expiration information, or expiration information for specific rewards program offers such as temporary promotions or rates. See also Nagaraj, [0069], where an optimization manager may collect relevant information including the user’s account or card types, membership information such as membership levels and account standing, rewards program enrollment or terms, rewards balances, user-defined preferences or objectives, and a user’s location information. The optimization manager may run a decision-making operation based at least in part on (received) collected information to determine a number of optimum cards or accounts for use based on the information inputs).
Although Nagaraj does not appear to explicitly state that the query for user-specific card information (i.e., a user’s known active cards) includes “an identifier associated with the user” as claimed, one of ordinary skill in the art would have found it obvious to have modified Nagaraj to explicitly include a query with the user’s identifier with the motivation of specifically (i.e., uniquely) identifying the user whose accounts is being retrieved, i.e., greater precision/efficiency/specificity.
Regarding claim 19: Perks as modified teaches The non-transitory computer-readable medium of claim 15, wherein the indication of the starting location comprises a physical address or a set of coordinates (Perks, [0023], where a determined current location of the user may be through utilization of GPS (implying “a set of coordinates”)24,25).
Regarding claim 20: Perks as modified teaches The non-transitory computer-readable medium of claim 15, wherein the time associated with the request is indicated in the request (Perks, [0029], where the map generator component 108 can generate driving directions based upon current or predicted traffic conditions along the routes to be traveled by the user, where one or more sensors may be associated with vehicles traveling along roads near the retail stores (i.e., “traffic data source”), e.g., the retail stores depicted on the map generated by the map generator 108 (i.e., “plurality of possible locations”), which can provide data that indicates the flow of traffic along certain roads and/or at particular intersections. See Perks, [0023] and [0025], where the geographic area of interest comprising those retail stores was based on the determined current location of the user, and that the map generated by the map generator component 108 includes the retail stores in the geographic area of interest which may be displayed on a map (implying that the “request for traffic information” is based on “the current location and the plurality of possible locations”)).
Although Perks does not appear to explicitly state that the traffic information includes “a time associated with the request is indicated in the request” as claimed, Perks discloses that traffic conditions may be predicted, and that historical traffic patterns can be utilized, suggesting the use of time as a factor. Therefore, one of ordinary skill in the art would have found it obvious to have modified Perks to have explicitly included a time associated with the request indicated within the request with the motivation of being able to leverage historical traffic patterns and predicting future traffic for when a user will be traveling to retail stores, as traffic tends to occur at certain times, making time an important component for making predictions as to traffic conditions.
Claim 5 is rejected under 35 U.S.C. 103 as being unpatentable over Perks et al. (“Perks”) (US 2016/0104232 A1), in view of Nagaraj et al. (“Nagaraj”) (US 2017/0083930 A1), in further view of Bruss et al. (“Bruss”) (US 2020/0226460 A1), in further view of Zivin (“Zivin”) (US 2007/0150369 A1).
Regarding claim 5: Perks as modified teaches The system of claim 1, but does not appear to explicitly teach wherein the machine learning model is configured to minimize a holistic cost associated with the at least one object, wherein the cost comprises the holistic cost.
Zivin teaches wherein the machine learning model is configured to minimize a holistic cost associated with the at least one object, wherein the cost comprises the holistic cost (Zivin, [0038-0058], where the system utilizes numerous variables to determine the optimal means of purchasing goods for the lowest total cost, where the optimization method includes selection of the vendors and location, with preference given to those with inventory remaining, availability, lowest prices, closest proximity to the customer’s current location and to other vendors, etc. See Perks as modified in claim 1 above with respect to the “machine learning model” and the machine learning model “minimizing the cost”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Perks as modified and Zivin. Perks discloses in [0028] that the system may perform optimization, e.g., travel expenses being taken into consideration when generating a shopping itinerary, further optimizations taken with respect to various modes of transformation, etc.
Therefore, one of ordinary skill in the art would have found it obvious to have modified Perks such that the optimization explicitly minimizes the holistic cost with the motivation of finding the ideal balance between money saved and time costs (see, e.g., Zivin, [0022]), as well as balancing with other possible variables disclosed in Zivin, [0040-0058].
Claims 7 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Perks et al. (“Perks”) (US 2016/0104232 A1), in view of Nagaraj et al. (“Nagaraj”) (US 2017/0083930 A1), in further view of Bruss et al. (“Bruss”) (US 2020/0226460 A1), in further view of Huang et al. (“Huang”) (US 2011/0197200 A1).
Regarding claim 7: Perks as modified teaches The system of claim 1, but does not appear to explicitly teach wherein the one or more processors are configured to: communicate with a third-party application programming interface to receive the directions.
Huang teaches [communicating] with a third-party application programming interface to receive the directions (Huang, [0043], where location information is decoded from location information parameters associated with a mapping API call made by a mapping script embedded in the content 118. See Perks, [0026], where when the user selecting hyperlinks corresponding to graphical icons representative of the retail stores on the map, causes the map generator component to output driving directions to a particular retail store from the current location of the user or from a reference location provided by the user).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Perks as modified and Huang. Huang discloses in [0018] that web mapping services support the use of special hyperlinks that may be added to third-party content to link to a mapping website furnished by the web mapping service where when selected, these hyperlinks cause the mapping website to be accessed by the device’s browser, e.g., to display a map of a desired location or furnish driving directions, e.g., the driving directions provided by Perks.
Therefore, one of ordinary skill in the art would have found it obvious to have incorporated Huang’s API for receiving directions into Perks’ disclosure with the motivation of enabling communication and interactivity between various device components, for example by causing functionality of an application to be exposed to configure an application for display by the display device and/or allow the user to interact with an application by providing inputs via the touch screen (Huang, [0037]).
Regarding claim 14: Perks as modified teaches The method of claim 8, wherein the selected location is physical, and the method further comprises: receiving an interaction with an indication of the selected location; and outputting … directions to the selected location (Perks, [0026], where the map generator component 108 provides selectable hyperlinks corresponding to graphical icons representative of (physical) retail stores on the map. The user can select a selectable link corresponding to the retail store on the map, causing the map generator component 108 to output driving directions to that retail store to the user for display on the display 110).
Perks as modified does not appear to explicitly teach [outputting], using a global positioning system application executed by the user device[, directions to the location].
Huang teaches [outputting], using a global positioning system application executed by the user device[, directions to the location] (Huang, [0029] and [0031], where the disclosed GPS system may be utilized such that the position-determining module 120 may use signal data 122 received via the GPS receiver 126 in combination with map data 128 that is stored in memory 106 to generate navigation instructions (e.g., turn-by-turn instructions to an input destination or POI), show a current position on a map, etc.).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Perks as modified and Huang with the motivation of providing a determination of geographic position in order to provide specific driving instructions to a user to enable the user to navigate to a particular location (see, e.g., Huang, [0031]).
Claim 12 is rejected under 35 U.S.C. 103 as being unpatentable over Perks et al. (“Perks”) (US 2016/0104232 A1), in view of Nagaraj et al. (“Nagaraj”) (US 2017/0083930 A1), in further view of Bruss et al. (“Bruss”) (US 2020/0226460 A1), in further view of Rathod (“Rathod”) (US 2021/0042724 A1).
Regarding claim 12: Perks as modified teaches The method of claim 8, but does not appear to explicitly teach further comprising: transmitting, for each possible account in the plurality of possible accounts, a set of credentials for the possible account.
Rathod teaches transmitting, for each possible account in the plurality of possible accounts, a set of credentials for the possible account (Rathod, [0485], where the user and/or user device 130 may have identity attributes stored with service provider server 180, and the user and/or user device 130 may have credentials to authenticate or verify identity with service provider server 180. User attributes may include personal information and banking information, where user attributes may be passed to service provider server 180 as part of a login, selection, purchase, and/or payment request, and the user attributes may be utilized by service provider server 180 to associate the user and/or user device 130 with one or more particular user accounts maintained by service provider server 180).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined the teachings of Perks as modified and Rathod with the motivation of automatically authenticating a user based such that the user does not have to provide another set of credentials to sign into a portal to access account feature information of the user.26
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). 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 IRENE BAKER whose telephone number is (408)918-7601. The examiner can normally be reached M-F 8-5PM 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, Boris Gorney can be reached at (571) 270-5626. 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.
/IRENE BAKER/Primary Examiner, Art Unit 2154
27 August 2026
1 See, e.g., TLI Communications LLC v. AV Automotive, L.L.C., 823 F.3d 607 (Fed. Cir. 2016), 823 F.3d at 612–13 (holding claims were directed to an abstract idea despite the claims’ recitation of telephone units and servers, which only merely provided a generic environment in which to carry out the abstract idea).
2 See also, e.g., Ultramercial, Inc. v. Hulu, LLC, 772 F.3d 709, 715 (Fed. Cir. 2014) (finding that claims directed to implementation on the Internet was an attempt to limit the use of the abstract idea to a particular technological environment).
3 See, e.g., Intellectual Ventures I LLC v. Erie Indemnity Co., 850 F.3d 1315, 121 USPQ2d 1928 (Fed. Cir. 2017) (“Remotely accessing and retrieving user-specified information is an age-old practice that existed well before the advent of computers and the Internet…”).
4 Alice Corp. v. CLS Bank Int'l, 573 U.S. __, 134 S. Ct. 2347 (2014)
5 See Alice at p. 15: “Using a computer to create and maintain ‘shadow’ accounts amounts to electronic recordkeeping—one of the most basic functions of a computer…. The same is true with respect to the use of a computer to obtain data, adjust account balances, and issue automated instructions; all of these computer functions are ‘well-understood, routine, conventional activit[ies]’ previously known to the industry…. In short, each step does not do more than require a generic computer to perform generic computer functions”.
6 See, e.g., Zivin (US 2007/0150369 A1) at [0018].
7 Durvasula et al. US 2025/0139378 A1 at [0108].
8 Shah et al. US 2021/0349865 A1 at [0069].
9 Liu et al. US 2018/0314735 A1 at [Background].
10 Durvasula et al. US 2025/0139378 A1 at [0108].
11 Shah et al. US 2021/0349865 A1 at [0069].
12 Durvasula et al. US 2025/0139378 A1 at [0108].
13 Shah et al. US 2021/0349865 A1 at [0069].
14 Liu et al. US 2018/0314735 A1 at [Background].
15 Durvasula et al. US 2025/0139378 A1 at [0108].
16 Shah et al. US 2021/0349865 A1 at [0069].
17 Bomar et al. US 2011/0154467 A1 at [0003] (“Credit cards usually include a payment account number…that includes the credit cards’ 12-19 digit number. The PAN is often embossed into and printed onto the credit card”).
18 Eng. US 2010/0243741 A1 at [0020] (“FIG. 1 is a conventional credit card 10 containing front card information including a complete credit card (account) number 0000 0000 0000 0000”).
19 Durvasula et al. US 2025/0139378 A1 at [0108].
20 Shah et al. US 2021/0349865 A1 at [0069].
21 Liu et al. US 2018/0314735 A1 at [Background].
22 Durvasula et al. US 2025/0139378 A1 at [0108].
23 Shah et al. US 2021/0349865 A1 at [0069].
24 McNeel. US 2015/0356539 A1 at [0029] (“The present system and method allows a user to stand at a location…and capture the location’s geographic coordinates with a GPS enabled device…”).
25 Choi. US 2007/0040742 A1 at [0005] (“A GPS is a system providing positioning, speed, and time information services by using satellites. The GPS can accurately provide 3-dimensional position, altitude, and time information services…”).
26 Louie et al. US 2014/0172660 A1 at [0215].
Prosecution Insights