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 .
Status of Claims
This Office Action is in response to the communication filed on 19 February 2026.
Claims 1-18 are being considered on the merits.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1 and 10 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claims 1 and 10 each recites the limitation "the side of the two-sided diversity constraint that is violated" in the 10th limitation of each of the claims. There is insufficient antecedent basis for this limitation in the claim.
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 1-18 are rejected under 35 USC § 101
Claim 1:
Step 1: Independent claim 1 recites a computer-implemented method and therefore falls under one of the four statutory categories of patent-eligible subject matter.
Step 2A Prong 1: See the rejection of claim 1 above. The same rationale applies to this dependent claim.
ranking pieces of content in the first set based on the scores; (Mental process: Ranking content is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can take a list of content and judge whether one content is better than another).
building a set of solutions without the diversity constraint; (Mental process: Building a set of solutions is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can take a list of solutions without constraints entirely in their mind or with pen and a piece of paper).
searching the set of solutions for a solution that satisfies one side of the two-sided diversity constraint; and (Mental process: searching a set of solutions is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can take a list of solutions and review them mentally for any solutions that satisfies one side of a diversity constraint or check or cross them out on a paper with a pen).
defining a dual objective function using the one-sided diversity constraint; (Mental process: Defining a dual objection function is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can conceptualize a dual objective function entirely in their mines or with the help of a pen and paper)
producing a solution to the dual objective function; (Mental process: Producing a solution a dual objection function is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can evaluate a dual objective function to find a solution entirely in their mines or with the help of a pen and paper).
recovering a primal solution from the solution; (Mental process: Evaluating a dual objection function to recover a primal solution is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can evaluate a dual objective function to find a primal solution entirely in their mines or with the help of a pen and paper).
re-ranking the pieces of content in the first set based on the primal solution; and (Mental process: Ranking content is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can take a list of content and judge whether one content is better than another).
Step 2A Prong 2: This judicial exception is not integrated into practical application
A system comprising: a processor; and a computer-readable medium having instructions stored thereon, which, when executed by the processor, cause the system to perform operations comprising: (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
accessing a first set comprising pieces of content; (Insignificant extra-solution activity to the judicial exception: storing and retrieving data from memory – See MPEP § 2106.05(g))
executing a machine learning relevance model to provide a score for each piece of content in the first set; (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
accessing a two-sided diversity constraint, the two-sided diversity constraint having an upper bound and a lower bound for a first attribute of the pieces of content; (Insignificant extra-solution activity to the judicial exception: storing and retrieving data from memory – See MPEP § 2106.05(g))
reducing the two-sided diversity constraint to a one-sided diversity constraint by eliminating either the upper bound or the lower bound, the two side-sided diversity constraint reduced by: (Insignificant extra-solution activity to the judicial exception: storing and retrieving data from memory – See MPEP § 2106.05(g))
retaining the side of the two-sided diversity constraint that is violated; (Insignificant extra-solution activity to the judicial exception: storing and retrieving data from memory – See MPEP § 2106.05(g))
adding a dual variable to the one-sided diversity constraint; (Insignificant extra-solution activity to the judicial exception: storing and retrieving data from memory – See MPEP § 2106.05(g))
causing display of one or more pieces of content in the first set based on the re-ranking. (insignificant extra-solution activity to the judicial exception: Presenting offers and gathering statistics – See MPEP § 2106.05(g))
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception
A system comprising: a processor; and a computer-readable medium having instructions stored thereon, which, when executed by the processor, cause the system to perform operations comprising: (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
accessing a first set comprising pieces of content; (Insignificant Extra Solution Activity: storing and retrieving data from memory is well-understood, routine, conventional activity – see Berkheimer evidence MPEP § 2106.05(d))
executing a machine learning relevance model to provide a score for each piece of content in the first set; (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
accessing a two-sided diversity constraint, the two-sided diversity constraint having an upper bound and a lower bound for a first attribute of the pieces of content; (Insignificant Extra Solution Activity: storing and retrieving data from memory is well-understood, routine, conventional activity – see Berkheimer evidence MPEP § 2106.05(d))
reducing the two-sided diversity constraint to a one-sided diversity constraint by eliminating either the upper bound or the lower bound, the two side-sided diversity constraint reduced by: (Insignificant Extra Solution Activity: storing and retrieving data from memory is well-understood, routine, conventional activity – see Berkheimer evidence MPEP § 2106.05(d))
retaining the side of the two-sided diversity constraint that is violated; (Insignificant Extra Solution Activity: storing and retrieving data from memory is well-understood, routine, conventional activity – see Berkheimer evidence MPEP § 2106.05(d))
adding a dual variable to the one-sided diversity constraint; (Insignificant Extra Solution Activity: storing and retrieving data from memory is well-understood, routine, conventional activity – see Berkheimer evidence MPEP § 2106.05(d))
causing display of one or more pieces of content in the first set based on the re-ranking. (Insignificant Extra Solution Activity: Presenting offers and gathering statistics is well-understood, routine, conventional activity – see Berkheimer evidence from MPEP 2106.05(d))
Claim 10:
Step 1: Independent claim 10 recites a computer-implemented method and therefore falls under one of the four statutory categories of patent-eligible subject matter.
Step 2A Prong 1: See the rejection of claim 10 above. The same rationale applies to this dependent claim.
ranking pieces of content in the first set based on the scores; (Mental process: Ranking content is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can take a list of content and judge whether one content is better than another).
building a set of solutions without the diversity constraint; (Mental process: Building a set of solutions is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can take a list of solutions without constraints entirely in their mind or with pen and a piece of paper).
searching the set of solutions for a solution that satisfies one side of the two-sided diversity constraint; and (Mental process: searching a set of solutions is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can take a list of solutions and review them mentally for any solutions that satisfies one side of a diversity constraint or check or cross them out on a paper with a pen).
defining a dual objective function using the one-sided diversity constraint; (Mental process: Defining a dual objection function is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can conceptualize a dual objective function entirely in their mines or with the help of a pen and paper)
producing a solution to the dual objective function; (Mental process: Producing a solution a dual objection function is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can evaluate a dual objective function to find a solution entirely in their mines or with the help of a pen and paper).
recovering a primal solution from the solution; (Mental process: Evaluating a dual objection function to recover a primal solution is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can evaluate a dual objective function to find a primal solution entirely in their mines or with the help of a pen and paper).
re-ranking the pieces of content in the first set based on the primal solution; and (Mental process: Ranking content is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can take a list of content and judge whether one content is better than another).
Step 2A Prong 2: This judicial exception is not integrated into practical application
A computerized method comprising: (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
accessing a first set comprising pieces of content; (Insignificant extra-solution activity to the judicial exception: storing and retrieving data from memory – See MPEP § 2106.05(g))
executing a machine learning relevance model to provide a score for each piece of content in the first set; (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
accessing a two-sided diversity constraint, the two-sided diversity constraint having an upper bound and a lower bound for a first attribute of the pieces of content; (Insignificant extra-solution activity to the judicial exception: storing and retrieving data from memory – See MPEP § 2106.05(g))
reducing the two-sided diversity constraint to a one-sided diversity constraint by eliminating either the upper bound or the lower bound, the two side-sided diversity constraint reduced by: (Insignificant extra-solution activity to the judicial exception: storing and retrieving data from memory – See MPEP § 2106.05(g))
retaining the side of the two-sided diversity constraint that is violated; (Insignificant extra-solution activity to the judicial exception: storing and retrieving data from memory – See MPEP § 2106.05(g))
adding a dual variable to the one-sided diversity constraint; (Insignificant extra-solution activity to the judicial exception: storing and retrieving data from memory – See MPEP § 2106.05(g))
causing display of one or more pieces of content in the first set based on the re-ranking. (insignificant extra-solution activity to the judicial exception: Presenting offers and gathering statistics – See MPEP § 2106.05(g))
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception
A computerized method comprising: (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
accessing a first set comprising pieces of content; (Insignificant Extra Solution Activity: storing and retrieving data from memory is well-understood, routine, conventional activity – see Berkheimer evidence MPEP § 2106.05(d))
executing a machine learning relevance model to provide a score for each piece of content in the first set; (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
accessing a two-sided diversity constraint, the two-sided diversity constraint having an upper bound and a lower bound for a first attribute of the pieces of content; (Insignificant Extra Solution Activity: storing and retrieving data from memory is well-understood, routine, conventional activity – see Berkheimer evidence MPEP § 2106.05(d))
reducing the two-sided diversity constraint to a one-sided diversity constraint by eliminating either the upper bound or the lower bound, the two side-sided diversity constraint reduced by: (Insignificant Extra Solution Activity: storing and retrieving data from memory is well-understood, routine, conventional activity – see Berkheimer evidence MPEP § 2106.05(d))
retaining the side of the two-sided diversity constraint that is violated; (Insignificant Extra Solution Activity: storing and retrieving data from memory is well-understood, routine, conventional activity – see Berkheimer evidence MPEP § 2106.05(d))
adding a dual variable to the one-sided diversity constraint; (Insignificant Extra Solution Activity: storing and retrieving data from memory is well-understood, routine, conventional activity – see Berkheimer evidence MPEP § 2106.05(d))
causing display of one or more pieces of content in the first set based on the re-ranking. (Insignificant Extra Solution Activity: Presenting offers and gathering statistics is well-understood, routine, conventional activity – see Berkheimer evidence from MPEP 2106.05(d))
Claims 2 and 11:
Step 2A Prong 1: See the rejection of claims 1 and 10 above. The same rationale applies to this dependent claim.
maintaining and updating an interval between a minimum value of the dual variable and a maximum value of the dual variable and shrinking the interval by half in each iteration. (Mental process: Maintaining, updating, and shrinking an interval by half is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components.
Step 2A Prong 2: This judicial exception is not integrated into practical application
The system of claim 1 and method of claim 10, wherein the bisection method iterates a plurality of times (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception
The system of claim 1 and method of claim 10, wherein the bisection method iterates a plurality of times (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
Claims 3 and 12:
Step 2A Prong 1: See the rejection of claim 2 and 11 above. The same rationale applies to this dependent claim.
The system of claim two and method of claim 11, wherein the solution is selected from the interval after at least 3 iterations. (Mental process: Selecting a solution is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can look at a choice of 3 solutions and opine on a selection entirely in their mind).
Step 2A Prong 2 and Step 2B: The claim does not include additional elements.
Claims 4 and 13:
Step 2A Prong 1: See the rejection of claim 1 above. The same rationale applies to this dependent claim.
The system of claim 3 and method of claim 12, further comprising screening out one or more pieces of content from the first set of pieces of content that would not be contained in a top n coordinates of a vector containing the solution, based on the scores. (Mental process: Screening out content that would not be contained in coordinates of a vector is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind; nothing in this claim element precludes the step from practically being performed in the mind. For example, a person can look coordinates of contents compared to coordinates of a vector and judge which contents to screen out).
Step 2A Prong 2 and Step 2B: The claim does not include additional elements.
Claims 5 and 14:
Step 2A Prong 1: See the rejection of claims 1 and 10 above. The same rationale applies to this dependent claim.
Step 2A Prong 2: This judicial exception is not integrated into practical application
The system of claim 1 and method of claim 10, wherein the machine learning relevance model is trained by passing training data through a machine learning algorithm. (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception
The system of claim 1 and method of claim 10, wherein the machine learning relevance model is trained by passing training data through a machine learning algorithm. (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
Claims 6 and 15:
Step 2A Prong 1: See the rejection of claims 5 and 14 above. The same rationale applies to this dependent claim.
Step 2A Prong 2: This judicial exception is not integrated into practical application
The system of claim 5 and method of claim 14, wherein the machine learning algorithm is a neural network. (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception
The system of claim 5 and method of claim 14, wherein the machine learning algorithm is a neural network. (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
Claims 7 and 16:
Step 2A Prong 1: See the rejection of claims 1 and 10 above. The same rationale applies to this dependent claim.
Step 2A Prong 2: This judicial exception is not integrated into practical application
The system of claim 1 and method of claim 10, wherein the machine learning relevance model is retrained based on feedback from a viewer. (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception
The system of claim 1 and method of claim 10, wherein the machine learning relevance model is retrained based on feedback from a viewer. (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
Claims 8 and 17:
Step 2A Prong 1: See the rejection of claims 1 and 10 above. The same rationale applies to this dependent claim.
Step 2A Prong 2: This judicial exception is not integrated into practical application
The system of claim 1 and method of claim 10, wherein the pieces of content are user profiles. (Generally linking the use of a judicial exception to a particular technological environment or field of use – See MPEP § 2106.05(h))
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception
The system of claim 1 and method of claim 10, wherein the pieces of content are user profiles. (Generally linking the use of a judicial exception to a particular technological environment or field of use – See MPEP § 2106.05(h))
Claims 9 and 18:
Step 2A Prong 1: See the rejection of claims 1 and 10 above. The same rationale applies to this dependent claim.
Step 2A Prong 2: This judicial exception is not integrated into practical application
The system of claim 1 and method of claim 10, wherein the method is run in a recommender, and the method further comprises: (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
causing display of the re-ranked pieces of content in an electronic display as recommendations to a user operating a device containing or connected to the electronic display. (insignificant extra-solution activity to the judicial exception: Presenting offers and gathering statistics – See MPEP § 2106.05(g))
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception
The system of claim 1 and method of claim 10, wherein the method is run in a recommender, and the method further comprises: (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea - see MPEP 2106.05(f))
causing display of the re-ranked pieces of content in an electronic display as recommendations to a user operating a device containing or connected to the electronic display. (Insignificant Extra Solution Activity: Presenting offers and gathering statistics is well-understood, routine, conventional activity – see Berkheimer evidence from MPEP 2106.05(d))
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 1-18 are rejected under 35 U.S.C. 103 as being unpatentable over Yin, et. al. (“Diversity Preference-Aware Link Recommendation for Online Social Networks”; arXiv:2205.10689v2 [cs.LG] 18 Oct 2022; hereinafter “Yin”) in view of Do, et al. (“Two-sided fairness in rankings via Lorenz dominance.” arXiv:2110.15781v1 [cs.IR] 28 Oct 2021; hereinafter “Do”), in view of Maes (US 2002/0002502 A1; hereinafter “Maes”), and further in view of Aziz, H., Biró, P., & Yokoo, M. (Matching Market Design with Constraints. Proceedings of the AAAI Conference on Artificial Intelligence, 36(11), 12308-12316. 2022. https://doi.org/10.1609/aaai.v36i11.21495; hereinafter, “Aziz”).
Claim 1, Yin teaches:
accessing a first set comprising pieces of content; (Yin, sec. 5.1: “The Google+ data set used in our evaluation contains data about Google+ users’ profiles as well as their linkages in July, August, and September 2011, which we denote as time periods 0, 1, and 2.”)
executing a machine learning relevance model (Yin, sec. 2.1: “Once training data is constructed, machine learning methods can be applied to the data to build a model, which in turn predicts the linkage likelihood for each pair of currently unlinked users. Machine learning methods commonly used for link recommendation include boosted decision tree (Benchettara et al., 2010), supervised random walk (Backstrom and Leskovec, 2011), support vector machine (SVM) (Gong et al., 2014), and Bayesian network (Li et al., 2017a). For example, Gong et al. (2014) apply a SVM to an attribute-augmented social network to predict linkage likelihoods of Google+ users.”) to provide a score for each piece of content in the first set; (Yin, sec. 2.2: “Specifically, an item is selected if it maximizes the weighted sum of its relevance score (i.e., recommendation accuracy) and its average dissimilarity to the items already in the final recommendation list (i.e., recommendation diversity).”)
ranking pieces of content in the first set based on the scores; (Yin, sec. 2.2: “The majority of these methods re-rank the items generated by an accuracy-maximizing recommender system to meet the diversity requirement”)
adding a dual variable to the one-sided diversity constraint; (Yin, appendix B. Proof of Theorem 2: “Let…H be the dual variables associated with the respective constraints in problem”)
defining a dual objective function using the one-sided diversity constraint; (Yin, Appendix B. Proof of Theorem 2: “H be the dual variables associated with the respective constraints in problem (3). Then the Lagrangian function of the problem is given by [equation omitted]” Examiner notes that Yin teaches a Lagrangian function i.e. as part of a primal-dual optimization function)
producing a solution to the dual objective function; (Yin, Appendix B. Proof of Theorem 2: “Note that problem (4) is a convex optimization problem and the KKT conditions are sufficient and necessary to the optimal solution (Boyd et al., 2004). Theorem 2 then follows immediately.” Examiner notes Yin teaches a convex optimization function which is a dual objective function.)
recovering a primal solution from the solution; (Yin, Appendix C. Time Complexity of Analysis of DPA-LR, Table A.1: “Sort entries in ys and set the top-k entries to 1 and the rest to 0.” Examiner notes Yin teaches selecting top k-entries where the top entry is a primal solution)
re-ranking the pieces of content in the first set based on the primal solution; and (Yin, sec. 2.2: “The majority of these methods re-rank the items generated by an accuracy-maximizing recommender system to meet the diversity requirement”)
Yin does not explicitly disclose, but Do teaches:
accessing a two-sided diversity constraint, the two-sided diversity constraint having an upper bound and a lower bound for a first attribute of the pieces of content; (Do, sec. 2.2: “We consider exposure as the item-side utility to follow prior work and for simplicity. Our framework and algorithm readily applies in a more general case of two-sided preferences, where items also have preferences over users (for instance, in hiring, job seekers have preferences over which recruiters they are recommended to”)
reducing the two-sided diversity constraint to a one-sided diversity constraint by eliminating either the upper bound or the lower bound, the two-side diversity constraint reduced by: (Do, sec. 2.2: “Our notion of fairness aims at improving the utility of the worse-off users and items. Since this does not prescribe exactly which fraction of the worse-off users/items should be prioritized, the assessment of trade-offs requires looking at all fractions of the population. This is captured by the generalized Lorenz curve used in cardinal welfare economics [53]… Inference is carried out by maximizing W” Examiner notes Do teaches eliminating the upper and lower bound by converting the two-sided constraint into a single constraint.)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Do into Yin. Yin teaches a new recommendation system utilizing the concept of diversity preference; Do teaches generating recommendation rankings by maximizing concave welfare functions. One of ordinary skill would have been motivated to combine the teachings of Do into Yin in order to guarantee recommendation rankings are Pareto efficient (Do, abstract).
Neither Yin or Do explicitly disclose, but Maes teaches:
A system comprising: a processor; and a computer-readable medium having instructions stored thereon, which, when executed by the processor, cause the system to perform operations comprising: (Maes, para. 0065: “The main sequence of instructions effectuating the functions of the invention and facilitating interaction among the user and the system reside on a mass storage device (such as a hard disk, floppy disk, or optical storage unit) 202 as well as in a main system memory 204 during operation. Execution of these instructions and performance of the functions of the invention is accomplished by a central-processing unit (“CPU”) 206, which may be a processor such as with Intel Pentium, Motorola PowerPC, or Sun SPARC. A communications interface 208 is a telephone modem or network controller that connects, via a gateway or an Internet access provider, to a data network 222.”)
causing display of one or more pieces of content in the first set based on the re-ranking. (Maes, para. 0067: “The code and graphical data are received by the user's web browser, and the web page that is displayed facilitates user selection of products. The user's selection of products causes the user's web browser running on the client computer 220, 221 to communicate the selection information back to the system 200. In a direct access embodiment, the user interface 231 generates data for the screen display 214 that facilitates user selection of products via the keyboard 210 and the position sensing device 212.”)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Maes into Yin, as modified. Maes teaches a computer-based tool helps a person choose a product from among many possible products. One of ordinary skill would have been motivated to combine the teachings of Maes into Yin, as modified, in order to guarantee recommendation rankings are Pareto efficient (Maes, abstract).
building a set of solutions without the diversity constraint; (Aziz, sec. 8: “Yokoi (2019) discussed that if the feasibility constraints are madroidal and if there is a total order on the individuals, then the greedy algorithm of selecting agents while not violating feasibility constraints gives rise to a choice function that satisfies the SUB condition.” Examiner notes Aziz teaches a total order of individuals as a set of solutions)
searching the set of solutions for a solution that satisfies one side of the two-sided diversity constraint; and (Aziz, sec. 8: “Yokoi (2019) discussed that if the feasibility constraints are madroidal and if there is a total order on the individuals, then the greedy algorithm of selecting agents while not violating feasibility constraints gives rise to a choice function that satisfies the SUB condition.” Examiner notes Aziz teaches a total order of individuals to be searched using a greedy algorithm on just one side of the two-sided diversity constraint as a set of solutions)
retaining the side of the two-sided diversity constraint that is violated; (Aziz, sec. 8: “Yokoi (2019) discussed that if the feasibility constraints are madroidal and if there is a total order on the individuals, then the greedy algorithm of selecting agents while not violating feasibility constraints gives rise to a choice function that satisfies the SUB condition.” Examiner notes Aziz teaches a total order of individuals to be searched using a greedy algorithm on just one side of the two-sided diversity constraint as a set of solutions where the pool of agents (opposed to the institutions) violating the constraints of the institution are nevertheless retained as part of the pool to be searched and evaluated for constraint violation)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Aziz into Yin, as modified. Aziz teaches developments in the field of two-sided matching with various constraints, including those based on regions, diversity, multi-dimensional capacities, and matroids. One of ordinary skill would have been motivated to combine the teachings of Aziz into Yin, as modified, in order to implement operations research and in particular combinatorial optimization (Aziz, introduction).
Claim 2, Yin as modified teaches claim 1 above. Yin further teaches:
The system of claim 1, wherein the bisection method iterates a plurality of times maintaining and updating an interval between a minimum value of the dual variable and a maximum value of the dual variable and shrinking the interval by half in each iteration. (Yin, Appendix B. Proof of Theorem 2: “Note that problem (4) is a convex optimization problem and the KKT conditions are sufficient and necessary to the optimal solution (Boyd et al., 2004). Theorem 2 then follows immediately.” Examiner notes that the algorithm of Appendix B teaches an iterative update of
β
h
, which converges monotonically to the optimal value.)
Claim 3, Yin as modified teaches claim 2 above. Yin further teaches:
The system of claim 2, wherein the solution is selected from the interval after at least 3 iterations. (Yin, Appendix B. Proof of Theorem 2: “Note that problem (4) is a convex optimization problem and the KKT conditions are sufficient and necessary to the optimal solution (Boyd et al., 2004). Theorem 2 then follows immediately.” Examiner notes that the algorithm of Appendix B teaches an iterative update of
β
h
, which converges monotonically to the optimal value which may be at least 3 iterations.)
Claim 4, Yin as modified teaches claim 3 above. Yin further teaches:
The system of claim 3, wherein the operations further comprise screening out one or more pieces of content from the first set of pieces of content that would not be contained in a top n coordinates of a vector containing the solution, based on the scores. (Yin, Appendix C. Time Complexity of Analysis of DPA-LR, Table A.1: “Sort entries in ys and set the top-k entries to 1 and the rest to 0.” Examiner notes Yin teaches selecting top k-entries where such entries not in the top-k are sorted into 0 and effectively screened out).
Claim 5, Yin as modified teaches claim 1 above. Yin further teaches:
The system of claim 1, wherein the machine learning relevance model is trained by passing training data through a machine learning algorithm. (Yin, sec. 2.1: “Once training data is constructed, machine learning methods can be applied to the data to build a model, which in turn predicts the linkage likelihood for each pair of currently unlinked users. Machine learning methods commonly used for link recommendation include boosted decision tree (Benchettara et al., 2010), supervised random walk (Backstrom and Leskovec, 2011), support vector machine (SVM) (Gong et al., 2014), and Bayesian network (Li et al., 2017a). For example, Gong et al. (2014) apply a SVM to an attribute-augmented social network to predict linkage likelihoods of Google+ users.”)
Claim 6, Yin as modified teaches claim 5 above. Yin further teaches:
The system of claim 5, wherein the machine learning algorithm is a neural network. (Yin, sec. 2.2: “State-of-the-art graph neural networks address these limitations by generating node representations through neural networks, e.g., convolutional neural networks.”)
Claim 7, Yin as modified teaches claim 1 above. Yin further teaches:
The system of claim 1, wherein the machine learning relevance model is retrained based on feedback from a viewer. (Yin, sec. 6: “. In addition, in an experiment, we can observe how users react to recommended friends and which recommended friends they actually connect with in real time. Finally, conventional accuracy-oriented measures for link recommendation, such as precision and recall, are defined based on whether a recommended friend is accepted by a user.”)
Claim 8, Yin as modified teaches claim 1 above. Yin further teaches:
The system of claim 1, wherein the pieces of content are user profiles. (Yin, sec. 3: “Let G =< UE > be an online social network of n users…Each user is described by a H-dimensional user profile”)
Claim 9, Yin as modified teaches claim 1 above. Yin further teaches:
The system of claim 1, wherein the system is contained in a recommender, and (Yin sec. 1: Two streams of research are closely related to our study: link recommendation for social networks and diversification methods for recommender system.”)
Yin does not explicitly disclose, but Maes teaches:
the operations further comprise: causing display of the re-ranked pieces of content in an electronic display as recommendations to a user operating a device containing or connected to the electronic display. (Maes, para. 0067: “The code and graphical data are received by the user's web browser, and the web page that is displayed facilitates user selection of products. The user's selection of products causes the user's web browser running on the client computer 220, 221 to communicate the selection information back to the system 200. In a direct access embodiment, the user interface 231 generates data for the screen display 214 that facilitates user selection of products via the keyboard 210 and the position sensing device 212.”)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Maes into Yin, as modified, as set forth above with respect to claim 1.
Claim 10, Yin teaches:
A computerized method comprising: accessing a first set comprising pieces of content; (Yin, sec. 5.1: “The Google+ data set used in our evaluation contains data about Google+ users’ profiles as well as their linkages in July, August, and September 2011, which we denote as time periods 0, 1, and 2.”)
ranking pieces of content in the first set based on the scores; (Yin, sec. 2.2: “The majority of these methods re-rank the items generated by an accuracy-maximizing recommender system to meet the diversity requirement”)
adding a dual variable to the one-sided diversity constraint; (Yin, appendix B. Proof of Theorem 2: “Let…H be the dual variables associated with the respective constraints in problem”)
defining a dual objective function using the one-sided diversity constraint; (Yin, Appendix B. Proof of Theorem 2: “H be the dual variables associated with the respective constraints in problem (3). Then the Lagrangian function of the problem is given by [equation omitted]” Examiner notes that Yin teaches a Lagrangian function i.e. as part of a primal-dual optimization function)
producing a solution to the dual objective function; (Yin, Appendix B. Proof of Theorem 2: “Note that problem (4) is a convex optimization problem and the KKT conditions are sufficient and necessary to the optimal solution (Boyd et al., 2004). Theorem 2 then follows immediately.” Examiner notes Yin teaches a convex optimization function which is a dual objective function.)
recovering a primal solution from the solution; (Yin, Appendix C. Time Complexity of Analysis of DPA-LR, Table A.1: “Sort entries in ys and set the top-k entries to 1 and the rest to 0.” Examiner notes Yin teaches selecting top k-entries where the top entry is a primal solution)
re-ranking the pieces of content in the first set based on the primal solution; and (Yin, sec. 2.2: “The majority of these methods re-rank the items generated by an accuracy-maximizing recommender system to meet the diversity requirement”)
Yin does not explicitly disclose, but Do teaches:
accessing a two-sided diversity constraint, the two-sided diversity constraint having an upper bound and a lower bound for a first attribute of the pieces of content; (Do, sec. 2.2: “We consider exposure as the item-side utility to follow prior work and for simplicity. Our framework and algorithm readily applies in a more general case of two-sided preferences, where items also have preferences over users (for instance, in hiring, job seekers have preferences over which recruiters they are recommended to”)
reducing the two-sided diversity constraint to a one-sided diversity constraint by eliminating either the upper bound or the lower bound, the two-side diversity constraint reduced by: (Do, sec. 2.2: “Our notion of fairness aims at improving the utility of the worse-off users and items. Since this does not prescribe exactly which fraction of the worse-off users/items should be prioritized, the assessment of trade-offs requires looking at all fractions of the population. This is captured by the generalized Lorenz curve used in cardinal welfare economics [53]… Inference is carried out by maximizing W” Exminer notes Do teaches eliminating the upper and lower bound by converting the two-sided constraint into a single constraint.)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Do into Yin as set forth above with respect to claim 1.
Neither Yin or Do explicitly disclose, but Maes teaches:
executing a machine learning relevance model to provide a score for each piece of content in the first set; (Maes, para. 0065: “The main sequence of instructions effectuating the functions of the invention and facilitating interaction among the user and the system reside on a mass storage device (such as a hard disk, floppy disk, or optical storage unit) 202 as well as in a main system memory 204 during operation. Execution of these instructions and performance of the functions of the invention is accomplished by a central-processing unit (“CPU”) 206, which may be a processor such as with Intel Pentium, Motorola PowerPC, or Sun SPARC. A communications interface 208 is a telephone modem or network controller that connects, via a gateway or an Internet access provider, to a data network 222.”)
causing display of one or more pieces of content in the first set based on the re-ranking. (Maes, para. 0067: “The code and graphical data are received by the user's web browser, and the web page that is displayed facilitates user selection of products. The user's selection of products causes the user's web browser running on the client computer 220, 221 to communicate the selection information back to the system 200. In a direct access embodiment, the user interface 231 generates data for the screen display 214 that facilitates user selection of products via the keyboard 210 and the position sensing device 212.”)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Maes into Yin, as modified, as set forth above with respect to claim 1.
building a set of solutions without the diversity constraint; (Aziz, sec. 8: “Yokoi (2019) discussed that if the feasibility constraints are madroidal and if there is a total order on the individuals, then the greedy algorithm of selecting agents while not violating feasibility constraints gives rise to a choice function that satisfies the SUB condition.” Examiner notes Aziz teaches a total order of individuals as a set of solutions)
searching the set of solutions for a solution that satisfies one side of the two-sided diversity constraint; and (Aziz, sec. 8: “Yokoi (2019) discussed that if the feasibility constraints are madroidal and if there is a total order on the individuals, then the greedy algorithm of selecting agents while not violating feasibility constraints gives rise to a choice function that satisfies the SUB condition.” Examiner notes Aziz teaches a total order of individuals to be searched using a greedy algorithm on just one side of the two-sided diversity constraint as a set of solutions)
retaining the side of the two-sided diversity constraint that is violated; (Aziz, sec. 8: “Yokoi (2019) discussed that if the feasibility constraints are madroidal and if there is a total order on the individuals, then the greedy algorithm of selecting agents while not violating feasibility constraints gives rise to a choice function that satisfies the SUB condition.” Examiner notes Aziz teaches a total order of individuals to be searched using a greedy algorithm on just one side of the two-sided diversity constraint as a set of solutions where the pool of agents (opposed to the institutions) violating the constraints of the institution are nevertheless retained as part of the pool to be searched and evaluated for constraint violation)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Aziz into Yin, as modified, as set forth above with respect to claim 1.
Claim 11, Yin as modified teaches claim 10 above. Yin further teaches:
The method of claim 10, wherein the bisection method iterates a plurality of time, maintaining and updating an interval between a minimum value of the dual variable and a maximum value of the dual variable, shrinking the interval by half in each iteration. (Yin, Appendix B. Proof of Theorem 2: “Note that problem (4) is a convex optimization problem and the KKT conditions are sufficient and necessary to the optimal solution (Boyd et al., 2004). Theorem 2 then follows immediately.” Examiner notes that the algorithm of Appendix B teaches an iterative update of
β
h
, which converges monotonically to the optimal value.)
Claim 12, Yin as modified teaches claim 11 above. Yin further teaches:
The method of claim 11, wherein the solution is selected from the interval after at least 3 iterations. (Yin, Appendix B. Proof of Theorem 2: “Note that problem (4) is a convex optimization problem and the KKT conditions are sufficient and necessary to the optimal solution (Boyd et al., 2004). Theorem 2 then follows immediately.” Examiner notes that the algorithm of Appendix B teaches an iterative update of
β
h
, which converges monotonically to the optimal value which may be at least 3 iterations.)
Claim 13, Yin as modified teaches claim 12 above. Yin further teaches:
The method of claim 12, further comprising screening out one or more pieces of content from the first set of pieces of content that would not be contained in a top n coordinates of a vector containing the solution, based on the scores. (Yin, Appendix C. Time Complexity of Analysis of DPA-LR, Table A.1: “Sort entries in ys and set the top-k entries to 1 and the rest to 0.” Examiner notes Yin teaches selecting top k-entries where such entries not in the top-k are sorted into 0 and effectively screened out).
Claim 14, Yin as modified teaches claim 10 above. Yin further teaches:
The method of claim 10, wherein the machine learning relevance model is trained by passing training data through a machine learning algorithm. (Yin, sec. 2.1: “Once training data is constructed, machine learning methods can be applied to the data to build a model, which in turn predicts the linkage likelihood for each pair of currently unlinked users. Machine learning methods commonly used for link recommendation include boosted decision tree (Benchettara et al., 2010), supervised random walk (Backstrom and Leskovec, 2011), support vector machine (SVM) (Gong et al., 2014), and Bayesian network (Li et al., 2017a). For example, Gong et al. (2014) apply a SVM to an attribute-augmented social network to predict linkage likelihoods of Google+ users.”)
Claim 15, Yin as modified teaches claim 14 above. Yin further teaches:
The method of claim 14, wherein the machine learning algorithm is a neural network. (Yin, sec. 2.2: “State-of-the-art graph neural networks address these limitations by generating node representations through neural networks, e.g., convolutional neural networks.”)
Claim 16, Yin as modified teaches claim 10 above. Yin further teaches:
The method of claim 10, wherein the machine learning relevance model is retrained based on feedback from a viewer. (Yin, sec. 6: “. In addition, in an experiment, we can observe how users react to recommended friends and which recommended friends they actually connect with in real time. Finally, conventional accuracy-oriented measures for link recommendation, such as precision and recall, are defined based on whether a recommended friend is accepted by a user.”)
Claim 17, Yin as modified teaches claim 10 above. Yin further teaches:
The method of claim 10, wherein the pieces of content are user profiles. (Yin, sec. 3: “Let G =< UE > be an online social network of n users…Each user is described by a H-dimensional user profile”)
Claim 18, Yin as modified teaches claim 10 above. Yin further teaches:
The method of claim 10, wherein the method is run in a recommender, and (Yin sec. 1: Two streams of research are closely related to our study: link recommendation for social networks and diversification methods for recommender system.”)
Yin does not explicitly disclose, but Maes teaches:
the method further comprises: causing display of the re-ranked pieces of content in an electronic display as recommendations to a user operating a device containing or connected to the electronic display. (Maes, para. 0067: “The code and graphical data are received by the user's web browser, and the web page that is displayed facilitates user selection of products. The user's selection of products causes the user's web browser running on the client computer 220, 221 to communicate the selection information back to the system 200. In a direct access embodiment, the user interface 231 generates data for the screen display 214 that facilitates user selection of products via the keyboard 210 and the position sensing device 212.”)
It would have obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Maes into Yin, as modified, as set forth above with respect to claim 1.
Response to Applicant Arguments/Remarks
Objection
The previously asserted objections have been withdrawn in light of applicants’ amendments
35 USC § 101
On page 8 of applicant’s remarks, applicant argues that the claims do not recite a mental process because the limitations cannot be practically performed in the human mind. First, applicant argues that a person cannot mentally execute a machine learning relevance model. Second, applicant argues that it is not practical for a person to mentally score the content of a machine learning model nor reduce a constraint by building a new set of solutions, searching for space, and finding duel objective functions within the context claimed because the period to functionally respond to the user would be over.
In this case, under section 101, each limitation must be examined. Therefore, the limitation of executing a machine learning model is merely application of a computer. The remaining limitations are a method—there is no particular application that applicant is claiming that requires the method be executed within any particular timeframe. While a machine learning model may apply and solve the method faster, this is generically and generally true when using a computer to solve algorithms and mathematical or search models.
On page 9 of applicant’s remarks, applicant argues that limitations as a whole integrates any recited exceptions into a practical application. Applicant specifically argues that the claims solve specific technical challenges associated with real-time recommender models that are subject to one or more diversity constraints.
However, applicant does not recite any particular improvement to the functioning of a computer, technology or technical field. Instead, applicant’s claims are a search optimization and recommender model which can be implemented using a computer and computer model. For example, in claim 1, applicant claims “executing a machine learning relevance model to provide a score for each piece of content in the first set”. Assigning a score to each piece—without anything more—can be accomplished entirely within the human mind. Moreover, depending on the number in the set, scores can be assigned at random in seconds. There are no claimed improvements to a computer or computer model itself.
At the top of page 10, applicant argues that the claimed techniques are a “structured linear program…able to obtain a result set to the primal problem without the inefficient calculations required to directly solve the primal problem”
In other words, an improvement to a linear program to solve a problem is claimed. A linear program is a method rather than any particular computer or technology.
Starting at the bottom of page 10, applicant argues that the claims amount to significantly more. Applicant argues that the claims limitations amount to significantly more and improve the functioning of a technology or other technical field.
However, as set forth above, the claims are not directed to the improvement of a technology or technological field.
At the bottom of page 11, applicant argues that the claims add specific limitations other than those that are well-understood, routine, or conventional. Applicant argues that the claims offer a new, more efficient architecture for a real-time recommender.
However, applicant does not offer any architecture but rather a specific methodology.
35 USC § 103
In light of applicant’s amendments and arguments, the § 103 rejection has been updated where claims 1-18 now stand rejected pursuant to Yin, in view of Do, in view of Maes, and further in view of Aziz, H., Biró, P., & Yokoo, M.
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 Sally T. Ley whose telephone number is (571)272-3406. The examiner can normally be reached Monday - Thursday, 10:00am - 6:00pm ET.
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, Viker Lamardo can be reached at (571) 270-5871. 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.
/STL/Examiner, Art Unit 2147
/VIKER A LAMARDO/Supervisory Patent Examiner, Art Unit 2147