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 .
Response to Amendment
The Amendment filed on 15 October 2025 has been entered. The following is in reply to the Amendments and Arguments.
Claims amended: 1, 7, 8, 15
Claims cancelled: 2, 3, 5, 6, 9, 10, 12, 13, 16, 17, 19, 20
Claims added: 21-23
Claims currently pending: 1, 4, 7, 8, 11, 14, 15, 18, 21-23
Response to Arguments
Applicant, in the “Disposition of the Claims”, “Claim Amendments”, and “Examiner Interview” sections, presents opening remarks regarding the disposition of the claims, the amendments to the claims, and the previously conducted Examiner Interview. As no specific argument is raised in this/these section(s) with respect to the instant application, no rebuttal is required.
Applicant, in the “Rejections under 35 U.S.C. § 101” section, presents Applicant’s overview of analysis under 35 U.S.C. § 101. Substantial argument begins in sub-section 2.a.
Applicant, in sub-section 2.a of the “Rejections under 35 U.S.C. § 101” section, refers to the “teeter-totter” verbiage from the MPEP and argues that the Examiner has not considered the newly added claims 21-23. Examiner notes that this argument is moot in view of the fact that these claims were added in the instant amendment to the claims. These new claims have now been considered and, unfortunately, the claims still fail to represent eligible subject matter under § 101, as described herein. Applicant then argues, “the human mind is not equipped to perform this method”. This argument is unpersuasive as the grounds of rejection are not based on a finding of a “mental process”. Applicant then argues that the claims “are directed towards modifying a machine learning model using feedback from previous recommendations for a specific task and not directed to the abstract idea of selecting addresses”. This argument ignores the fact that a “machine learning model” is trained using data, often from previous uses of the model, which is where the “learning” comes from. Simply employing a machine learning model or artificial intelligence does not rise beyond the level of “apply it”. As noted in the grounds of rejection herein, Wikipedia entries for machine learning and the specific machine learning technique of a “multi-armed bandit” were well-known, routine, and conventional techniques to apply to a recommendation system. This is little more than having a goal of determining postal address recommendations and stating the solution involves algebra or invoking functions. Therefore, Examiner reasserts that the claims are directed towards the abstract idea of improving address recommendations based on a user's feedback to address recommendations.
Applicant, in sub-section 2.b of the “Rejections under 35 U.S.C. § 101” section, refers to Enfish and argues that the claims represent “an improvement to computer functionality”. The claims here are unlike the claims in Enfish. In Enfish, the court relied on the distinction made in Alice between computer-functionality improvements and uses of existing computers as tools in aid of processes focused on “abstract ideas”. Enfish, 822 F.3d at 1335–36; see Alice, 134 S. Ct. at 2358–59. In Enfish, this distinction was applied to reject the claims under § 101 because the claims at issue focused not on asserted advances in uses to which existing computer capabilities could be put, but on a specific improvement—a particular database technique—in how computers could carry out one of their basic functions of storage and retrieval of data. Enfish, 822 F.3d at 1335–36; see Bascom, 2016 WL 3514158, at *5; cf. Alice, 134 S. Ct. at 2360 (noting basic storage function of generic computer). The present case is different: the focus of the claims is not on such an improvement in computers as tools, but on certain independently abstract ideas that use computers as tools. Examiner notes that this line of argument is paraphrased from the Federal Circuit's precedential ruling in Electric Power Group, LLC v. Alstom S.A., Alstom Grid Inc., Psymetrix, LTD,. Alstom Limited, 2015-1778; directly quoting from the decision was not used to avoid too many layers of quotation. Here, the claims address a business problem of correcting inaccurately entered/received addresses and providing recommended address suggestions to an expert and using the expert’s input to improve future recommendations. Improving the suggestions does not improve the operation of the computer itself. Furthermore, Applicant has not identified which additional elements Applicant feels render the claim into a practical application. As such, Applicant’s argument is merely conclusory and unpersuasive.
Applicant, in sub-section 3 of the “Rejections under 35 U.S.C. § 101” section, argues that the claims "provide an unconventional technological solution (a machine learning model) to a technological problem (using the machine learning model to make address recommendations and updating the machine learning model based on user feedback)". Examiner disagrees to the notion that using "a machine learning model" represents an unconventional technological solution as machine learning is not a new topic but a class of well-understood, routine, and conventional algorithms used to solve a variety of problems as can be seen in the Wikipedia articles noted in the grounds of rejection under 35 U.S.C. § 101 presented herein. Furthermore, the problem of recommending addresses for the purposes of accurately shipping goods to customers is not a "technological problem", but rather a business problem. The claimed invention merely applies machine learning to solve a business problem. Therefore, the grounds of rejection under 35 U.S.C. § 101 is herein maintained albeit updated to reflect Applicant’s amendments to the claims.
Applicant, in the “Rejections under 35 U.S.C. § 103” section, argues that the newly amended claim language overcomes the previously cited prior art rejections. Examiner notes that Applicant has “rolled-up” claims 5 and 6 into claim 1 and similar dependent claims into the other independent claims representing alternate embodiments. Claim 6 was previously indicated by the Office as overcoming the prior art rejections. A further search confirms that the newly amended claim language overcomes the previously cited prior art. The grounds of rejection under 35 U.S.C. § 103 have been withdrawn.
Applicant, in the “New Claims” section, notes the addition of new claims 21-23, which have been taken into account in the grounds of rejection under 35 U.S.C. § 101 presented herein.
Applicant does not present substantial argument in the “Conclusion” section. As no specific argument is raised, no rebuttal is required.
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, 4, 7, 8, 11, 14, 15, 18, 21-23 are rejected under 35 U.S.C. § 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more.
Step 1: Claims 1, 4, 7, 21 are directed towards a method. Claims 8, 11, 14, 22 are directed towards a manufacture (computer-readable medium). Claims 15, 18, 23 are directed towards a system. Thus, these claims, on their face, are directed to one of the statutory categories of 35 U.S.C. § 101.
Step 2A - Prong One: As per MPEP 2106.04, Prong One asks does the claim recite an abstract idea, law of nature, or natural phenomenon. In Prong One examiners evaluate whether the claim(s) recites a judicial exception; that is, whether the claim(s) set forth or describe a law of nature, natural phenomenon, or abstract idea.
Claim 1 is presented here as a representative claim for specific analysis (The underlined claim terms here are interpreted as additional elements beyond the abstract idea and are further analyzed under Step 2A - Prong Two):
A method for generating address recommendation, comprising:
obtaining from a client device, by a recommendation system, an address recommendation request associated with an address, wherein the address is associated with a user and a purchase order;
identify, by a selection feature, a plurality of potential addresses from a user information repository based on a customer identifier of the user and a postal code associated with the address;
in response to obtaining the address recommendation request and the potential addresses:
generating a context vector associated with the address and the potential addresses,
wherein generating the context vector associated with the address comprises:
generating similarity scores between the address and each of the potential addresses,
generating dissimilarity scores between each of the potential addresses,
obtaining historical information associated with the user from the user information repository,
obtaining behavior information associated with the user from the user information repository and wherein the context vector comprises the similarity scores, the dissimilarity scores, the historical information, and the behavior information;
generating an address recommendation based on the context vector, the potential addresses, and a payoff function using a recommendation model, wherein:
the recommendation model is a machine learning model that implements an upper confidence bound multi-armed bandit algorithm,
the payoff function is designed to specify an expected reward and an upper confidence level with the address recommendation, and
the recommendation model balances an exploitation specified by the expected reward and an exploration specified by an upper confidence level;
send the address recommendation to an expert user via a graphical user interface;
obtaining user feedback associated with the address recommendation, wherein the feedback specifies the expert user used the address recommendation to process the purchase order;
generating a reward based on the user feedback with a value of one indicating the address recommendation was used by the expert user;
and updating neural network weights of the recommendation model to generate a new payoff function based on the context vector, the address recommendation and the reward wherein the new payoff function is optimized to decrease the upper confidence level for future address recommendations causing the recommendation model to exploit new expected rewards in generating the future address recommendations.
The claims here are based on the recitation of an abstract idea (i.e. recitation other than the additional elements delineated here with underlining and further addressed per Step 2B - Prong Two). The claims recite the abstract idea of improving address recommendations based on a user's feedback to address recommendations which falls within certain methods of organizing human activity.
The phrase "certain methods of organizing human activity" applies to fundamental economic principles or practices including hedging insurance, mitigating risk; commercial or legal interactions including agreements in the form of contracts, legal obligations, advertising, marketing or sales activities or behaviors business relations; managing personal behavior or relationships or interactions between people including social activities teaching, and following rules or instructions. Refer to MPEP 2106.04(a)(2) II. A-C.
The Remaining Claims: The analysis of claim 1 is applicable to the additional independent claims 8 and 15 as these additional claims comprise alternate embodiments that implement the same steps of the method of the claim analyzed above. The additional independent claims recite the same additional elements as the claim analyzed above except for the following additional elements: "non-transitory computer readable medium", "computer processor", "a client", "recommendation manager", and "a processor and memory".
The dependent claims recite the same additional elements as the parent claim(s) and fail to recite any additional elements. Dependent claims 4-7, 11-14, and 18-20 reiterate the same abstract idea with further embellishments. Claims 4, 11, 18 further describe the reward process. Claims 5, 6, 12, 13, 19, and 20 further describe the generation of the context vector. Claim 7 describes the recommended address.
Therefore, the identified claims fall within the subject matter groupings of abstract ideas enumerated in MPEP 2106.04(a)(2). Thus, the analysis proceeds to Prong Two to evaluate whether the claim integrates the abstract idea into a practical application.
Step 2A - Prong Two: As per MPEP 2106.04.II.A.2, Prong Two asks does the claim recite additional elements that integrate the judicial exception into a practical application.
The claims offer the additional elements of: "client device", "a recommendation system", “user information repository”, "non-transitory computer readable medium", "computer processor", "a client", "recommendation manager", “graphical user interface”, and "a processor and memory". It would have been readily apparent to one having ordinary skill in the art (PHOSITA) at the time the invention was filed that the additional elements represent generic computing devices. The additional element(s) are simply utilized as generic computing tools to implement the abstract idea, functioning as mere instructions to apply the exception as noted in MPEP 2106.05(f). The specification, in at least 0026 and 0027 describes the recommendation manager as a "hardware processor" or "computer instructions". The description of the elements as generic computing hardware continues in at least 0062. Furthermore, the claims appear to be a solution to a commercial/business problem providing address recommendations for addresses found in purchase orders.
The ordered combination of these additional elements amounts to generally linking the use of the abstract idea to a particular technological environment or field of use (MPEP 2106.05(h)). The ordered combination offers nothing more than employing a generic configuration of computer devices and computer functions. The claims do not amount to a practical application, similar to how limiting the abstract idea in Flook to petrochemical and oil-refining industries was insufficient.
Step 2B: As per MPEP 2106.05, the additional elements are analyzed, both individually and in combination, to determine whether an "inventive concept" is furnished by an element or combination of elements that is recited in the claim in addition to (beyond) the judicial exception, and is sufficient to ensure that the claim, as a whole, amounts to significantly more than the judicial exception itself. The analysis under Step 2B does not consider the elements describing the abstract ideas that are set forth above in Step 2A. Instead, the analysis only assesses the claim limitations other than the invention's use of the ineligible concepts to which the claims are directed. The court's precedent has consistently employed this same approach, and as a matter of law, narrowing or reformulating an abstract idea does not add "significantly more" to it. BSG Tech LLC v. Buyseasons, Inc., 899 F.3d 1281 (Fed. Cir. 2018).
Utilizing a "context vector" within a "recommendation model" represents using a "recommender system" which was well-understood, routine, and conventional use of machine learning techniques as described in the Wikipedia article titled, "Recommender system". Examiner also notes that presence of a "vector space representation" in the "Content-based filtering" section. Similarly, a "machine learning model that implements an upper confidence bound multi-armed bandit algorithm" represents a type of machine learning model that is well-understood, routine, and conventional as illustrated by the Wikipedia article titled, "Multi-armed bandit".
Potentially Allowable Subject Matter
The claims, as currently written, have overcome the prior art. However, the grounds of rejection under 35 U.S.C. § 101 are currently pending and represent a barrier to allowability.
Examiner notes the following references as disclosing the closest prior art or being pertinent to understanding the state of the art at the time the invention was filed.
Cetintas et al. (Pub. #: US 2021/0398193 A1) discloses a recommendation model that utilizes a feature vector (i.e., context vector) for both the user input and the candidates for matching and a payoff function to make a recommendation.
Gartner et al. (Pub. #: US 2022/0050824 A1) discloses receiving an address recommendation request from a user device and retrieving a "candidate address set" based upon "zip codes".
Nandury et al. (Pub. #: US 2019/0005439 A1) discloses a technique for providing an address recommendation to an "admin" whereby the admin provides input as to whether "corrective actions" with respect to the address are utilized.
Subramanian et al. (Pub. #: US 2023/0162830 A1) discloses a system using machine learning models to assist in order processing by predicting missing values for fields in the order.
Zheng et al. (Pub. #: US 2023/0334308 A1) discloses two disparate rewards that are used for "rewarding desired behaviors or punishing undesired ones".
The Wikipedia article, "Multi-armed bandit", discloses a class of algorithms used to solve problems under the "stochastic scheduling" category.
Ye (Pub. #: CN 115205085 A) discloses a system for determining address similarity that uses a user's history as an input.
Li, in "Approximate Address Matching", discloses an address matching algorithm for use in increasing postal delivery accuracy.
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 SCOTT SNIDER whose telephone number is (571)272-9604. The examiner can normally be reached M-W: 9:00-4:30 Mountain (11:00-6:30 Eastern).
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, Waseem Ashraf can be reached at (571)270-3948. 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.
/S.S/Examiner, Art Unit 3621
/WASEEM ASHRAF/Supervisory Patent Examiner, Art Unit 3621