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 .
This communication is in response to applicant’s response and amendments filed 03/23/2026. Claims 1, 12-13, and 18-19 are amended and hereby entered. Claim 11 has been canceled. Claims 1-2, 4-5, 9-10, and 12-19 are pending. No claims are allowed.
Response to Arguments
Applicant's arguments filed 3/23/2026 have been fully considered but they are not persuasive.
Regarding 35 USC 101:
The applicant submits the claims are not directed towards a mental process because the machine learning operations require computer implemented processing that cannot practically be performed in the human mind. Specifically, the applicant cites, (i) determining weights for a plurality of machine-learned labels based on the learning of a presence/absence estimation model, and (ii) calculating a score indicating the presence or absence of a relation type using the generation machine learning model and those learned weights.
First, determining weights to be used in the calculation performed by the generation model based on the output of the presence/absence model, can be performed by the human mind. Using the output of the first model (presence/absence model) as input the calculations of the second model (generation model) describe a step that can be performed by the human mind. The applicant argues determining weights requires repeated gradient based optimization across large data sets involving matrix multiplication, activation function evaluations, and back propagation. However, these features are not recited in the claims. The claims simply state the weights are based on the learning of the presence/absence estimation machine learning model, with no mention of mathematical functions. The machine model is discussed at a high level of generality such that use of the machine learning model to determine weights amounts to an “apply it” recitation, see MPEP 2106.05(f)
Second, the calculation step is described as being performed by the machine model (generation model). The claims recite calculating a score that indicates presence or absence of a relationship type with a machine model. The applicant submits the step involves high precision numerical parameters, iterative optimization and multilayer nonlinear computations that are intrinsic to machine-learning and are beyond human cognitive capability. However, the claims do not recite these mathematical concepts and instead recite a broad recitation of calculating a score using a model. Therefore, the claim recites an abstract idea (mental process) being performed by an additional element (generation machine learning model) at a high level of generality, amounting to an “apply it” recitation, see MPEP 2106.05(f).
Further the applicant submits there is a technological improvement to machine learning estimation systems because existing systems cannot “fully grasp” context. However, there is no improvement to the underlying technology itself (machine learning). The claims recite the use of machine learning models to determine weights and estimate relationships. There is no improvement to the function of the machine learning technology itself; rather there is use of machine learning to perform the mental process. Describing training to adjust weights is not an improvement in machine learning model technology itself; rather, it adjusts what the machine model outputs and thereby the intended result of the outcome.
Regarding 35 USC 103:
The applicant submits that amended features of claim 1 (previously in dependent claim 11) overcome the referenced prior art. Specifically, the applicant submits that Bruich does not show labels being based on the input data. However, the claims recite a step of obtaining input data and outputting labels based on input data, and Bruich clearly shows input data in paragraph 23 with gathered information from user profile objects, action logs, edge objects, external household data, and timeline data. Further in paragraph 44 of Bruich, received information of users is used to predict household memberships and determines probabilities of household membership. This shows a clear link between input data and labels, as received data is used to determine if a user is a member of a household. Therefore, the examiner respectfully disagrees and the rejection is maintained.
Further, the applicant submits that Bruich does not teach the recited score because the score is not determined in accordance with probabilities of each label associated with weights. However, the calculation step in the claim does not recite probabilities of each label. The claims simply recite weighting labels and determining weights with a model. Therefore, the claims recite and are interpreted as weighted labels determined by a model. Paragraph 47-48 of Bruich shows that weights can be trained by machine learning model and that weights of labels are used when determining scores. Therefore, the examiner respectfully disagrees and the rejection is maintained.
Additionally, the applicant argues Bruich does not include intermediate label outputs or learned weights to generate a final score. However, as previously mentioned, paragraphs 47-48 of Bruich show trained weights for labels as described in the claims. Therefore, the examiner respectfully disagrees and the rejection is maintained.
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 1-2, 4-5, 9-10, and 12-19 are rejected under 35 U.S.C. 101 because the claimed invention is directed to judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) with no practical application and without significantly more.
Claims 1-2, 4-5, 9-10 and 12-19 are systems. Thus, each claim on its face is directed to one of the statutory categories of 35 USC 101. However, the claims are rejected under 35 USC 101 because the claimed invention is directed to an abstract idea without significantly more.
The claimed invention is directed to an abstract idea in that the instant application is directed to a mental process (See MPEP 2106.04(a)(2)(III)). The independent claims (1, 18, and 19) recite a system to evaluate data related to people in separate households to estimate relationships. These claim elements are being interpreted as concepts performed in the human mind (including observation, evaluation, judgement, and opinion). Using attribute data to determine relationships can equivalently be achieved by human observation and evaluation of data. For example, a human can estimate two people have a parent-child relationship based on attributes such as same last name, age gap, shared finances, etc. The claims recite an abstract idea consistent with the “mental process” grouping set forth in the MPEP 2106.04(a)(2)(III).
Further, the claimed invention is also directed to an abstract idea in that the instant application is directed to certain methods of organizing human activity (See MPEP 2106.04(a)(2)(II)). The independent claims recite a system that recommends a product or service based on an estimated relationship. These claim elements are being interpreted as commercial or legal interactions (advertising, marketing or sales activities or behaviors). Recommending a product or service is a form of advertising, marketing and sales. The claims recite an abstract idea consistent with the “certain methods of organizing human activity” grouping set forth in the MPEP 2106.04(a)(2)(II).
The instant application fails to integrate the judicial exception into a practical application because the instant application merely recites an “apply it” (or an equivalent) with the judicial exception, or merely includes instructions to implement an abstract idea. The instant application is directed towards systems to implement the identified abstract ideas of mental processes and commercial interactions (i.e. receiving and processing person and household data to estimate a relationship and recommend a product or service and the like) in a general computer environment. The claims do not include additional elements that integrate the abstract ideas into practical application or amount to significantly more than the judicial exception. The independent claims recite the additional elements “a processor”, “a memory device”, “a presence/absence estimation machine learning model” and “generation machine learning model”. These claim elements are recited at a high level of generality such that it amounts to no more than mere instructions to apply the exception using a general computer environment. The machines merely act as a modality to implement the abstract idea and are not indicative of integration into a practical application (i.e., the additional elements are simply used as a tool to perform the abstract idea), see MPEP 2106.05(f).
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed in Step 2A Prong Two analysis, the additional elements in the claims amount to no more than mere instructions to apply the exception using generic computer components. The same analysis applies here in 2B and does not provide an inventive concept.
In regards to the dependent claims
Claims 2, 4-5, and 9-10, and 12-17 further limit the abstract ideas and introduce no new additional elements.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
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 1-2, 4-5, and 9-10, and 12-19 are rejected under 35 U.S.C. 103 as being unpatentable over Bruich (US 20130151527 A1) in view of Tang (US 11810001 B1) in further view of Bullock (US 20160277526 A1).
Regarding Claims 1 and 19, (substantially similar in scope and language) Bruich teaches:
An information processing system, comprising: at least one processor; and at least one memory device that stores a plurality of instructions which, when executed by the at least one processor, causes the at least one processor to: [see at least Bruich: (Para 0060-0062) “Such a computer program may be stored in a non-transitory, tangible computer readable storage medium, or any type of media suitable for storing electronic instructions, which may be coupled to a computer system bus.”]
first, acquire household information, including a street address, [see at least Bruich: (Para 0005, 0014, 0016, 0020) “such as names of persons living in a household… an address associated with a household”]
second, select a first person living at the first household and a second person living at the second household [see at least Bruich: (Para 0033) “Household objects 116 representing households that have been determined to include selected users are associated with the selected users' user profile objects”]
third, estimate a type of a relation between the first household and the second household based on an attribute of the first person and an attribute of the second person and based on a log of messages sent and received [see at least Bruich: (Para 0036, 0024, 0027, 0040) “household prediction module 114 may classify users into various types of households, such as a single household, a married household, a non-married coupled household”, “generate edges with other users that parallel the users' real-life relationships”, (Para 0034) “The action logger 210 populates the action log 104 with information about user actions to track them. Such actions may include, for example, adding a connection to the other user, sending a message to the other user, uploading an image, reading a message from the other user, viewing content associated with the other user, attending an event posted by another user, among others”, (Para 0031) “The web server 208 may provide the functionality of receiving and routing messages between the social networking system 100 and the user devices 202, for example, instant messages, queued messages (e.g., email), text and SMS (short message service) messages, or messages sent using any other suitable messaging technique.”]
estimate a presence or an absence of a relation type for the person belonging to the first household using a presence/absence estimation machine learning model; [see at least Bruich: (Para 0040, 0042, Fig 3) “As another example, a household object 116 may be generated by the household prediction module 114 for a user based on a social graph correlation with a household… In another embodiment, a social graph correlation may be determined based on relationship status indicated in user profile objects 102 for users, such as a parent-child relationship, marriage, or domestic partnership. As a result of the social graph correlation, the users are associated with the household such that the users' user profile objects 102 are associated with the household object 116 for the household.”]
wherein the presence/absence estimation machine learning model is configured to: obtain a plurality of input data comprising a usage history including a transaction history relating to various computer systems and at least a part of the attribute of the person belonging to the first household; [see at least Bruich: (Para 0023) “This information is gathered from user profile objects 102, the action log 104, edge objects 106, content objects 108, external household data 110, and timeline data 112, as described above.”]
output a plurality of labels indicating whether the person belonging to the first household has the relation type based on each of the plurality of input data; and [see at least Bruich: (Para 0044) “The statistical analysis module 302 may determine a probability that the user is a member of a household based on these factors. In one embodiment, social graph information is used to further determine a probability that a user is a member of a household based on other members of the household and whether they are close connections to the user or whether they have a familial relationship with the user.”]
calculate a score indicating the presence or absence of the relation type in accordance with a weight of each of the plurality of labels using a generation machine learning model; [see at least Bruich: (Para 0038) “Confidence scores are generated by the household prediction module 114 upon associating users with households. Separate confidence scores may be associated with different household objects 116 when a user is associated with more than one household.”]
wherein the weights of each of the plurality of labels are determined based on the learning of the presence/absence estimation machine learning model. [see at least Bruich: (Para 0047) “Different types of households may have different scoring models for determining confidence scores. For example, a married household type may heavily weight social graph relationship status information about married users in the calculation of the confidence score for the household prediction”, (Para 0048) “Weights may also be trained by machine learning algorithms based on received information linking users of the social networking system 100 to households”]
While Bruich teaches estimating relationships between users of the same household, it does not explicitly teach the households being two separate households. However, Bruich does not each but Tang does teach:
for a first household and a second household, wherein each household of which includes one or more persons a plurality of users living together; [wherein clause describes intended results; see at least Tang: (Column 3, Lines 21-27) “knowledge graph data structures with node data and edge data, the node data corresponding to a plurality of homes, and the edge data corresponding to identified relationships between the plurality of homes; building a home knowledge graph having nodes and edges based on the node data and the edge data, the home knowledge graph”]
between the user belonging to the first household and the user belonging to the second household, wherein the street address of the first household is different than the street address of the second household [see at least Tang: (Column 3, Lines 21-27) “knowledge graph data structures with node data and edge data, the node data corresponding to a plurality of homes, and the edge data corresponding to identified relationships between the plurality of homes; building a home knowledge graph having nodes and edges based on the node data and the edge data, the home knowledge graph”, (Column 11, Lines 23-27) “For example, server 140 may employ one or more rules for identifying address information in parsed data, by searching the parsed data for a predetermined format of [house number] [street name] [road label (st./rd./In./pl./ave., etc.)]”]
Further, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the method of determining user relationships (Bruich) with two separate households (Tang). One of ordinary skill would have recognized determining user relationships between different households would be beneficial for creating a more robust database and better representation of social graph data. (see at least Bruich: “the present technology relates to techniques for determining household membership”; see at least Tang: (Column 2, Lines 1-11) “As another example, relational databases typically store individual information about the relationships between any two given entities. When new entities and/or relationships are added, database entries grow exponentially to store all new relationships between individual entity pairs. At the scale required in current systems, the storage and computation requirements for maintaining and updating relational databases are unsustainable. Thus, traditional relational database architectures are unsuitable for use in a dynamic system having multiple complex relationships between entities.”
While the combination of Bruich and Tang teach determining relationships between different households, it does not explicitly teach recommending a product or service. However, Bruich and Tang do not teach but Bullock does teach:
and fourth, recommend a product or a service based on the estimated type of relation. [see at least Bullock: (Para 0025) “This information can be leveraged by the social networking system to optimize the presentation of relevant content, advertising, and other services to the user”]
Further, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the method of estimating relationships between users of different households (Bruich and Tang) with the method of recommending a product or service based on an estimated relationship (Bullock). One of ordinary skill would have recognized the benefits of using relationships to better recommend advertisements (see at least Bruich: (Para 0002) “Advertisers, in an effort to provide relevant advertisements, may use this market research to target their advertisements based on the metrics obtained from these panels of sample households”).
Regarding Claim 2, the combination of Bruich, Tang and Bullock teach the limitations above, Bruich further teaches:
wherein, the acquisition of the household information is further based on a surname of each of a plurality of persons stored in a user database, [see at least Bruich: (Para 0005, 0014, 0016, 0020) “such as names of persons living in a household… an address associated with a household”] the first household and the second household each including one or a plurality of persons living together from the plurality of persons are acquired. [see at least Bruich: (Para 0020, 0016) “such as names of persons living in a household”]
Regarding Claim 4, the combination of Bruich, Tang, and Bullock teach the limitations above, Bruich further teaches:
wherein the plurality of parameters include at least part of whether a surname is the same, a frequency of telephone contact, presence or absence of sending a gift relating to a specific day, a frequency of sending a gift to each other, an age difference, a friend in common, whether gender is the same, and similarity in street addresses. [see it at least Bruich: (Para 0024, 0011, 0036) “age, gender, geographical location, education history, employment history and the like”]
Regarding Claim 5, the combination of Bruich, Tang, and Bullock teach the limitations above, Bruich further teaches:
wherein, in the estimation, the type of relation between the first household and the second household is estimated depending on whether a type of a relation between a first person included in the first household and a second person included in the second household is any of at least part of parent-child, sibling, and neighbor. [see at least Bruich: (Para 0036, 0024, 0025, 0046) “the user profile objects 102 associated with the parent and child may be categorized as a family household. This household classification may, in one embodiment, be used to assign users to households gathered from external household data 110.”]
Regarding Claim 9, the combination of Bruich, Tang, and Bullock teach the limitations above, Bruich further teaches:
wherein, the at least one memory device that stores a plurality of instructions which, when executed by the at least one processor, causes the at least one processor to estimate the type of the relation based on results of clustering for each of a plurality of household pairs. [see at least Bruich: (Para 0005) “Users of a social networking system may be assigned to households using prediction models that rely, in part, on user profile information and social graph data. Information about users may be received by a social networking system through various channels… User attributes, such as previous names, date of birth, social graph data, locations, life events, and check-ins, may be used as factors in generating confidence scores of predicted household memberships.” (Para 0027) “Finally, social graph correlation may indicate whether the social relationship indicated in a household matches the social graph relationship of users in the social networking system 100. For example, a household that includes two members, a married couple, may be associated with two users that have indicated they are married to each other. The social graph correlation would be 1.0”]
Regarding Claim 10, the combination of Bruich, Tang, and Bullock teach the limitations above, Bruich further teaches:
wherein, the at least one memory device that stores a plurality of instructions which, when executed by the at least one processor, causes the at least one processor to estimate the type of the relation based on a record of an exchange of information including a message transmission destination. [see at least Bruich: (Para 0050) “Another use of the heuristics analysis module 308 includes gathering and analyzing different types of information about a user's geographic location such as check-ins at places in a specific geographic location, attending events in the same geographic region, receiving requests for connecting with users from the same geographic area, and geo-location codes embedded in photos and other communications, such as text messages, uploaded to the social networking system by the user”]
Regarding Claim 12, the combination of Bruich, Tang and Bullock teach the limitations above, Bruich further teaches:
wherein the plurality of input data is only information of the person belonging to the first household and not information associated with any other persons. [see at least Bruich: (Para 0005) “In one embodiment, the social networking system uses a machine learning algorithm to analyze user information to determine confidence scores for matching potential households.”]
Regarding Claim 13, the combination of Bruich, Tang, and Bullock teach the limitations above, Bruich further teaches:
wherein the generation machine learning model is trained to minimize loss based on a plurality of label probabilities calculated for the plurality of labels. [see at least Bruich: (Para 0044) “Statistical analysis may also be performed by the statistical analysis module 302 to determine a probability, based on the received information about the user linking the user to the household, that the user is a member of the household based on past predictions and historical household data”, (Para 0045) “Statistical analysis is also performed to improve the weights of the different types of data used in data models and also used to measure performance, such as providing an error rate”]
Regarding Claim 14, the combination of Bruich, Tang, and Bullock teach the limitations above, Bruich further teaches:
wherein the at least one memory device that stores the plurality of instructions which, when executed by the at least one processor, causes the at least one processor to: identify a pair of node data connected by a link data between the first household and the second household; [see at least Bruich: (Para 0040) “In another embodiment, a social graph correlation may be determined based on relationship status indicated in user profile objects 102 for users, such as a parent-child relationship, marriage, or domestic partnership.”]
generate a pair attribute data indicating a common attribute between the first household and the second household; [see at least Bruich: (Para 0018) “Further, the edge object 106 between the connected users included in the check-in event may indicate that the connection is stronger than the connection between users that do not interact with each other as frequently. This type of connection information, extracted from edge objects 106, may be used by the social networking system 100 in assigning users of the social networking system 100 to households.”]
classify the combination of the first household and the second household into one of a plurality of clusters; and wherein each of the plurality of clusters is associated with a type of relation. [see at least Bruich: (Para 0024) “The household prediction module 114 may use one or more fields in the user profile objects 102 associated with users that indicate familial relationships, such as being a son or daughter of one or more users, being married to a user, being the parent of one or more children, and so on”, (Para 0040) “As a result of the social graph correlation, the users are associated with the household such that the users' user profile objects 102 are associated with the household object 116 for the household.”]
Regarding Claim 15, the combination of Bruich, Tang, and Bullock teach the limitations above, Bruich further teaches:
wherein the at least one memory device that stores the plurality of instructions which, when executed by the at least one processor, causes the at least one processor to: estimate the type of the relation between the first household and the second household using a machine learning model; [see at least Bruich: (Para 0027) “Finally, social graph correlation may indicate whether the social relationship indicated in a household matches the social graph relationship of users in the social networking system 100. For example, a household that includes two members, a married couple, may be associated with two users that have indicated they are married to each other. The social graph correlation would be 1.0”, (Para 0052) “A machine learning module 310 may be used in the household prediction module 114 to refine the weights used for making household predictions. In one embodiment, a social networking system 100 uses a machine learning algorithm to retrain weights in the household prediction module 114. Using the data gathered by the social networking system 100 that links a user with a household, the machine learning module 310 may be used to train scoring models for determining confidence scores”]
wherein the machine learning leaning model is trained with learning data comprising a plurality of parameter values acquired for a plurality of pairs of households and groundtruth data indicated the type of relation for each of the plurality of pairs of households. [see at least Bruich: (Para 0005) “Users of a social networking system may be assigned to households using prediction models that rely, in part, on user profile information and social graph data. Information about users may be received by a social networking system through various channels (e.g., declared/profile information…”]
Regarding Claim 16, the combination of Bruich, Tang, and Bullock teach the limitations above, Bruich further teaches:
wherein the at least one memory device that stores the plurality of instructions which, when executed by the at least one processor, causes the at least one processor to: obtain a usage history for the person belonging to the first household and the person belonging to the second household from different computer systems including a purchase history comprising a geographic location information of where a plurality of purchases were made; [see at least Bruich: (Para 0020) “External household data 110 may be used by a social networking system 100 to predict locations of users. External household data 110 may include offline household information retrieved by the social networking system 100 from one or more third-party external systems, such as department of motor vehicle records, magazine subscriptions, voter registration records, postal service mailing address changes, catalogs, and other transactional data, such as purchasing data at retailers. This external household data 110 may include information about households, such as names of persons living in a household, an address associated with a household, previous addresses associated with a household”]
obtain a history relating to purchases and browsing performed by the person belonging to the first household in an electronic commerce transaction system; and wherein the usage history and the history are used to estimate the type of relation. [see at least Bruich: (Para 0012) “A social networking system may also be able to capture external website data that is accessed by its users. This external website data may include websites that are frequently visited, links that are selected, and other browsing data. Information about users, such as stronger interests in particular users and applications than others based on their behavior, can be generated from these recorded actions through analysis and machine learning by the social networking system”, (Para 0015) “Market researchers use households to provide statistically valid measurements of the effectiveness of advertising campaigns to advertisers, track purchases of products across different demographics”, (Para 0027) “social graph correlation may indicate whether the social relationship indicated in a household matches the social graph relationship of users in the social networking system 100.”]
Regarding Claim 17, the combination of Bruich, Tang, and Bullock teach the limitations above, Bruich further teaches:
wherein the estimation model outputs a score relating to whether or not the user belonging to the first household has a spouse based on one of a plurality of input parameters relating to the user belonging to the first household. [see at least Bruich: (Para 0005) “The scoring models may rely on statistical analysis of the received user information to predict household membership for users of the social networking system… the social networking system uses a machine learning algorithm to analyze user information to determine confidence scores for matching potential households”, (Para 0024) “The household prediction module 114 may use one or more fields in the user profile objects 102 associated with users that indicate familial relationships, such as being a son or daughter of one or more users, being married to a user, being the parent of one or more children, and so on”]
Regarding Claim 18, Bruich teaches:
An information processing system, comprising: at least one processor; and at least one memory device that stores a plurality of instructions which, when executed by the at least one processor, causes the at least one processor to: [see at least Bruich: (Para 0060-0062) “Such a computer program may be stored in a non-transitory, tangible computer readable storage medium, or any type of media suitable for storing electronic instructions, which may be coupled to a computer system bus.”]
first, acquire household information, including a street address, [see at least Bruich: (Para 0005, 0014, 0016, 0020) “such as names of persons living in a household… an address associated with a household”]
second, select a first person living at the first household and a second person living at the second household; [see at least Bruich: (Para 0033) “Household objects 116 representing households that have been determined to include selected users are associated with the selected users' user profile objects”]
third, estimate a type of a relation between the first household and the second household based on an attribute of the first person and an attribute of the second [see at least Bruich: (Para 0036, 0024, 0027, 0040) “household prediction module 114 may classify users into various types of households, such as a single household, a married household, a non-married coupled household”, “generate edges with other users that parallel the users' real-life relationships”
and based on results of clustering for each of a plurality of household pairs; wherein the street address of the first household is different than the street address of the second household [see at least Bruich: (Para 0005) “Users of a social networking system may be assigned to households using prediction models that rely, in part, on user profile information and social graph data. Information about users may be received by a social networking system through various channels… User attributes, such as previous names, date of birth, social graph data, locations, life events, and check-ins, may be used as factors in generating confidence scores of predicted household memberships.” (Para 0024) “The household prediction module 114 may use one or more fields in the user profile objects 102 associated with users that indicate familial relationships, such as being a son or daughter of one or more users, being married to a user, being the parent of one or more children, and so on”, (Para 0027) “Finally, social graph correlation may indicate whether the social relationship indicated in a household matches the social graph relationship of users in the social networking system 100”, (Para 0040) “As a result of the social graph correlation, the users are associated with the household such that the users' user profile objects 102 are associated with the household object 116 for the household”, (Para 0020) “ This external household data 110 may include information about households, such as names of persons living in a household, an address associated with a household, previous addresses associated with a household,”]
estimate a presence or an absence of a relation type for the person belonging to the first household using a presence/absence estimation machine learning model; [see at least Bruich: (Para 0040, 0042, Fig 3) “As another example, a household object 116 may be generated by the household prediction module 114 for a user based on a social graph correlation with a household… In another embodiment, a social graph correlation may be determined based on relationship status indicated in user profile objects 102 for users, such as a parent-child relationship, marriage, or domestic partnership. As a result of the social graph correlation, the users are associated with the household such that the users' user profile objects 102 are associated with the household object 116 for the household.”]
wherein the presence/absence estimation machine learning model is configured to: obtain a plurality of input data comprising a usage history including a transaction history relating to various computer systems and at least a part of the attribute of the person belonging to the first household; [see at least Bruich: (Para 0023) “This information is gathered from user profile objects 102, the action log 104, edge objects 106, content objects 108, external household data 110, and timeline data 112, as described above.”]
output a plurality of labels indicating whether the person belonging to the first household has the relation type based on each of the plurality of input data; and [see at least Bruich: (Para 0044) “The statistical analysis module 302 may determine a probability that the user is a member of a household based on these factors. In one embodiment, social graph information is used to further determine a probability that a user is a member of a household based on other members of the household and whether they are close connections to the user or whether they have a familial relationship with the user.”]
calculate a score indicating the presence or absence of the relation type in accordance with a weight of each of the plurality of labels using a generation machine learning model; [see at least Bruich: (Para 0038) “Confidence scores are generated by the household prediction module 114 upon associating users with households. Separate confidence scores may be associated with different household objects 116 when a user is associated with more than one household.”]
wherein the weights of each of the plurality of labels are determined based on the learning of the presence/absence estimation machine learning model. [see at least Bruich: (Para 0047) “Different types of households may have different scoring models for determining confidence scores. For example, a married household type may heavily weight social graph relationship status information about married users in the calculation of the confidence score for the household prediction”, (Para 0048) “Weights may also be trained by machine learning algorithms based on received information linking users of the social networking system 100 to households”]
While Bruich teaches estimating relationships between users of the same household, it does not explicitly teach the households being two separate households. However, Bruich does not each but Tang does teach:
for a first household and a second household, wherein each household includes one or more persons living together; [wherein clause describes intended results; see at least Tang: (Column 3, Lines 21-27) “knowledge graph data structures with node data and edge data, the node data corresponding to a plurality of homes, and the edge data corresponding to identified relationships between the plurality of homes; building a home knowledge graph having nodes and edges based on the node data and the edge data, the home knowledge graph”]
Further, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the method of determining user relationships (Bruich) with two separate households (Tang). One of ordinary skill would have recognized determining user relationships between different households would be beneficial for creating a more robust database and better representation of social graph data. (see at least Bruich: “the present technology relates to techniques for determining household membership”; see at least Tang: (Column 2, Lines 1-11) “As another example, relational databases typically store individual information about the relationships between any two given entities. When new entities and/or relationships are added, database entries grow exponentially to store all new relationships between individual entity pairs. At the scale required in current systems, the storage and computation requirements for maintaining and updating relational databases are unsustainable. Thus, traditional relational database architectures are unsuitable for use in a dynamic system having multiple complex relationships between entities.”]
While the combination of Bruich and Tang teach determining relationships between different households, it does not explicitly teach recommending a product or service. However, Bruich and Tang do not teach but Bullock does teach:
and fourth, recommend a product or a service based on the estimated type of relation. [see at least Bullock: (Para 0025) “This information can be leveraged by the social networking system to optimize the presentation of relevant content, advertising, and other services to the user”]
Further, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the method of estimating relationships between users of different households (Bruich and Tang) with the method of recommending a product or service based on an estimated relationship (Bullock). One of ordinary skill would have recognized the benefits of using relationships to better recommend advertisements (see at least Bruich: (Para 0002) “Advertisers, in an effort to provide relevant advertisements, may use this market research to target their advertisements based on the metrics obtained from these panels of sample households”).
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 Examiner Benjamin Truong, whose telephone number is 703-756-5883. The examiner can normally be reached on Monday-Friday from 9 am to 5 pm (EST).
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, Nathan Uber SPE can be reached on 571-270-3923. 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.
/B.L.T. /Examiner, Art Unit 3626
/NATHAN C UBER/Supervisory Patent Examiner, Art Unit 3626
Prosecution Insights