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 10 December 2025 has been entered. The following is in reply to the Amendments and Arguments.
Claims amended: 1, 9, 17
Claims cancelled: 4-6, 8, 12-14, 16, 18, 20-22
Claims added: none
Claims currently pending: 1-3, 7, 9-11, 15, 16, 17, 19
Examiner notes that claim 18 was not noted by Applicant as having been amended, but contained a small edit with markings.
Response to Arguments
Applicant, in section A, presents opening remarks regarding the disposition of the claims and the amendments to the claims. As no specific argument is raised in this/these section(s) with respect to the instant application, no rebuttal is required.
Applicant, in section B, refers to Figure 1 and asserts that the disclosure therein supports the claim language of “disconnecting, by the recommendation interface computer program, the plurality of channel recommendation engines from the recommendation interface computer program in response to the centralized recommendation engine being trained, wherein the plurality of channel recommendation engines remain available to their respective communication channels after being disconnected”. Applicant asserts “the only connection that could be disconnected is between the recommendation engines and the recommendation interface”. This argument SUPPORTS rather than opposes the grounds of rejection presented herein under 35 U.S.C. § 112(a). The claim language relies on the “plurality of channel recommendation engines” to remain available to “communication channels” after “being disconnected”. If, as depicted in Figure 1, the channel recommendation engines communication to their respective channels only through the “Recommendation Interface”, they cannot remain available to their respective channels if they are “disconnected” from the “Recommendation Interface”. In other words, looking at Figure 1, the only connection the recommendation engines have with the rest of the system is through the recommendation interface. Disconnecting the recommendation engines from the recommendation interface connects them from the system entirely. A completely disconnected recommendation engine would NOT “remain available” to its respective communication channel. Therefore, support for this claim language is NOT found within the specification and Applicant’s argument further reinforces this finding.
Applicant then asserts that the “specification does not disclose that the channel recommendation engines are disconnected from their respective channels”. This is correct, as the specification, in 0052, simply states, “the channel recommendation engines may be disabled, disconnected, etc.”, but does not specify what the recommendation engines are disconnected from. Therefore, the claim language in question is not fully supported by the specification, necessitating the grounds of rejection under 35 U.S.C. § 112. The specification does not contain much detail as to the overall process and Applicant appears to make assumptions about the functionality that are not fully supported. Therefore, the grounds of rejection are herein maintained.
Applicant, in section C.1, refers to the grounds of rejection referencing Lhuillier, and argues that “Desjardins’ ‘feedback’ is not a disclosure of the receipt of recommendation request contexts” because Applicant believes the “feedback” does not include the “recommendation” context. The recommendation context, as per 0039, includes “any information that may be helpful in providing a recommendation”. Therefore, the “recommendation” context is broadly interpreted to mean the information regarding the user that could be used to generate a recommendation. The “user data” of Dejardins neatly reads on this broadly written claim limitation. Applicant then argues, “there is not reason, other than hindsight, to provide a data aggregation module that generates user recommendations from other recommendation engines”. Examiner disagrees to this notion as the disclosure of Lhuillier discloses utilizing the recommendations of disparate engines to train other recommendation engines. Lhuillier itself provides the motivation do gather such data for said usage. Additionally, a “data aggregation” technique does just that, and aggregates a variety of data. Thus, Dejardins is readily combinable with the additional data types of Lhuillier and one of ordinary skill in the art at the time the invention was filed would have anticipated the results of adding additional data to a recommendation engine. Applicant then lists what appears to be the entirety of claim 1 with a claim limitation bolded, the claim limitation largely comprising a roll-up of now cancelled claim 8. Applicant argues that Desjardins’ “user activities” do not read on “whether the first recommendation was displayed, accepted, or declined”. The specification does not make clear which entity performs the “displayed, accepted, or declined” operations, and Examiner has interpreted this claim language to be the results of the recommendation as seen and responded to by the “user”. This is supported by at least 0049, which states that a recommendation could be “presented and declined on a channel other than the requesting channel”; i.e., the recommendation may have been presented to the user, but declined by said user. The claim language appears broader than Applicant seems to be arguing, but the argument here is sparse. Therefore, this argument is unpersuasive and the grounds of rejection are herein maintained, albeit updated to reflect Applicant’s amendments to the claims.
Applicant does not present any arguments in support of the patentability of claims 7 and 15 except to assert that the claims are patentable based on their dependence from independent claim(s) 1 and 9. Therefore, said dependent claims stand rejected under the grounds of rejection presented herein and no detailed rebuttal is required.
35 USC § 101 Discussion
Examiner states the following reasons for overcoming the grounds of rejection under 35 U.S.C. § 101: The structure of disparate recommendation engines for each of a plurality of communication channels with a central recommendation engine that takes over the message recommendation for each channel upon completion of sufficient training of a machine learning model via at least the steps of disabling the plurality of channel recommendation engines upon completion of training of a central recommendation engine and routing of recommendation requests to the central recommendation engine renders the claims subject matter eligible under 35 U.S.C. § 101.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 1-3, 7, 9-11, 15, 16, 17, 19 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention. Independent claim 1 contains the limitation, “disconnecting, by the recommendation interface computer program, the plurality of channel recommendation engines from the recommendation interface computer program in response to the centralized recommendation engine being trained, wherein the plurality of channel recommendation engines remain available to their respective communication channels after being disconnected”, wherein the underlined portion represents subject matter that lacks written description support. Independent claims 11 and 17 contain similar claim language. Examiner does not find support for the “channel recommendation engines” remaining available after they are “disabled” or “disconnected”. Furthermore, the claims contain, “from the recommendation interface computer program”, but the specification does not indicate what the “channel recommendation engines” are disconnected from. Support for disconnecting or disabling the channel recommendation engines appears to derive from 0052 of the Specification, which reads in its entirety: “In one embodiment, once the centralized recommendation engine is sufficiently trained, the channel recommendation engines may be disabled, disconnected, etc.”. Examiner notes that the disposition of the channel recommendation engines is not disclosed within this paragraph and paragraphs 0053 et. seq. appear to provide a description of the computing environment in which the invention is implemented. The drawings do not contain mention of disabling or disconnecting the channel recommendation engines, let alone a description of the disposition of said channels after the step of disabling/disconnecting. Therefore, this element of the claims lacks written description support. The claims have been examined, in light of the specification, as if the invention does not actively delete/destroy the channel recommendation engines after they have been disabled/disconnected.
References of Record but not Applied in the Current Grounds of Rejection
The prior art listed below is made of record as considered pertinent to applicant's disclosure and is not relied upon in the grounds of rejection presented in this Office action. Those starred with '*' were added to this list in this Office action. Those without "*" were added in a previous Office action and are not repeated on a PTO-892 Notice of References Cited form, but are maintained herein for informational purposes only.
* Haze et al. (Pub. #: US 2020/0410017 A1) discloses reviewing the performance of a plurality of recommendation engines and learning to adjust the weights given thereto in order to optimize the overall performance of the recommendation system.
Kalluri (Pub. #: US 10,680,841 B1) discloses using machine learning techniques to optimize the distribution of messages using a central controller and via a plurality of networked devices.
McGovern (Pub. #: CA 02825159) discloses a multi-channel marketing system with the addition of an attribution analysis that uses "clustering" and "neural networks" to enhance the performance of marketing messages in a campaign.
Liou and Liu, in "Hybrid Multiple Channels-Based Recommendation for Mobile Commerce", discusses a system for optimizing presentation of marketing messages to users over multiple channels.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claim(s) 1-3, 8-11, 16-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over:
Desjardins (Pub. #: US 2021/0319478 A1) in view of
Lhuillier et al. (Pub. #: GB 2444520 A) in view of
Minor et al. (Pub. #: US 2015/0161662 A1) in view of
Sarfarez (Pub. #: US 2021/0241168 A1).
Claim(s) 1, 9:
These claims are analogous with different representative embodiments; claim 1 is a method embodiment and claim 9 is a product embodiment. Claim 9 is a system embodiment but is reworded enough to warrant a separate rejection (see below). Desjardins teaches a computer system with computer-readable media in at least 0041 for performing the steps:
A method, comprising: receiving, by a recommendation interface computer program, a first recommendation request context from a first communication channel of a plurality of communication channels,
(Desjardins teaches a plurality of communications channels for presenting content to users in at least 0048, 0057 and teaches that the content may be presented to a user in response to an advertisement request via "web advertising" in at least 0058.)
wherein each of the plurality of communication channels is associated with a different customer interface selected from the group consisting of email, phone, web, and application,
(Desjardins teaches communications channels that comprise different customer interfaces including at least Email, push notification, SMS, Social Networks, Mail, Website in at least 425 of Figure 4 and Figure 5 and a telephony system, telephone voicemail in at least 0058.)
wherein each communication channel is associated with one of a plurality of channel recommendation engines,
wherein the first recommendation request context comprises an identification of the first communication channel and an identification of a first customer that is interacting with the first communication channel;
(Desjardins teaches identifying a user in at least 0051, 0111, and a channel in at least the communications channels 425 in Figure 4 with feedback data returned to the central data aggregation system.)
routing, by the recommendation interface computer program, the first recommendation request context to the channel recommendation engine for the first communication channel;
receiving, by the recommendation interface computer program, a first recommendation from the channel recommendation engine for the first communication channel, wherein the first recommendation is for a product or discount;
(Desjardins discloses “real-time product recommendations” in at least 0043, product details in 0072-0074, 0090-0091.)
providing, by the recommendation interface computer program, the first recommendation and the first recommendation request context to a centralized recommendation engine,
training, by the centralized recommendation engine, a centralized machine learning engine with the first recommendation and the first recommendation request context; and with recommendations and requests from the channel recommendation engines for the other communication channels;
providing, by the recommendation interface computer program, the first recommendation to the first communication channel, wherein the first communication channel provides the first recommendation to the first customer;
(Desjardins teaches that the "communication channel orchestration interface" uses "business logic" to select content to display to a user in at least 0071.)
receiving, by the recommendation interface computer program, a first result of the first recommendation from the first communication channel, wherein the first result comprises a behavioral event including whether the first recommendation was displayed, accepted, declined;
(Desjardins teaches "feedback data" that includes "user interactions" comprising at least "clicks", "opening personalized communication messages", with "opening personalized communications messages" reading on at least “displayed” and “accepted”, and “not opening personalized communication messages” reading on “declined” in at least 0075.)
providing, by the recommendation interface computer program, the first result to the centralized recommendation engine,
(Desjardins teaches identifying a user in at least 0051, 0111, and a channel in at least the communications channels 425 in Figure 4 with feedback data returned to the central data aggregation system. The data aggregation module receives "user interactions" in at least 0075.)
training, by the centralized recommendation engine, the centralized machine learning engine with the first result;
(Desjardins teaches the that "data aggregation module" employs an "AI technique" in at least 0078, 0091, and "AI technologies" in at least 0113 in addition to the "machine learning" algorithm usage taught in at least 0062, 0072. Also see 0122, 0123.)
disconnecting, by the recommendation interface computer program, the plurality of channel recommendation engines in response to the centralized recommendation engine being trained, wherein the plurality of channel recommendation engines remain available to their respective communication channels after being disconnected;
routing, by the recommendation interface computer program, recommendation requests from the plurality of communication channels to the centralized recommendation engine;
(Desjardins discloses a "communication channel orchestration interface" that integrates with the "data aggregation module" in at least Figure 2 that is described as having "business logic" including "trigger conditions" in at least 0055 that is utilized by the "optimization and personalization engine" to recommend content to send to a user in at least 0056-0059.)
receiving, by the recommendation interface computer program, a second recommendation request context from the first communication channel, wherein the second recommendation request context comprises the identification of the first communication channel and an identification of a second customer that is interacting with the first communication channel;
(Desjardins teaches identifying a user in at least 0051, 0111, and a channel in at least item 425 of Figure 4.)
routing, by the recommendation interface computer program, the second recommendation request context to the centralized recommendation engine;
(Desjardins teaches modifying the marketing actions based upon feedback data analyzed in a central recommendation engine in item 350 of Figure 3 and 0062-0064.)
receiving, by the recommendation interface computer program, a second recommendation from the centralized recommendation engine;
(Desjardin teaches an "optimization and personalization engine" in at least 0072, with communication over multiple communication channels taught in at least 0058 and in 0074.)
wherein the centralized recommendation engine is configured to generate the second recommendation using the centralized machine learning engine, wherein the second recommendation is for a second product or discount;
(Desjardins teaches the that "data aggregation module" employs an "AI technique" in at least 0078, 0091, and "AI technologies" in at least 0113 in addition to the "machine learning" algorithm usage taught in at least 0062, 0072. Also see 0122, 0123. Desjardins discloses “real-time product recommendations” in at least 0043, product details in 0072-0074, 0090-0091.)
and providing, by the recommendation interface computer program, the second recommendation to the first communication channel, wherein the first communication channel provides the second recommendation to the second customer.
(Desjardins: 0072, 0058, 0074.)
As for, "wherein each of the plurality of communication channels is associated with a channel recommendation engine", “receiving, by the recommendation interface computer program, a first recommendation from the channel recommendation engine for the first communication channel, wherein the first recommendation is for a product or discount;”, “providing, by the recommendation interface computer program, the first recommendation and the first recommendation request context to a centralized recommendation engine,”, “training, by the centralized recommendation engine, a centralized machine learning engine with the first recommendation and the first recommendation request context; and with recommendations and requests from the channel recommendation engines for the other communication channels;”:
Desjardins teaches the that "data aggregation module" employs an "AI technique" in at least 0078, 0091, and "AI technologies" in at least 0113 in order to process the "feedback data" from each of the channels in order to formulate a centralized response in at least 0075-0077 in addition to the "machine learning" algorithm usage taught in at least 0062, 0072. Also see 0122, 0123. Desjardins teaches feedback data 435 being used for "prediction, labelling, and recommendations" and subsequently used by the central system for "Data Aggregation" module in at least Figure 4. Desjardins does not appear to specify that the data aggregation model operates on a “first recommendation request context” received from a channel recommendation engine. However, Lhuillier teaches a technique of multiple channels each with its own channel recommendation engine that receive channel recommendation contexts from other channels which are then incorporated, via machine learning, into a user’s profile for making further recommendations in at least Page 12, Line 1 – Page 13, Line 5 and Page 29, Line 16 – Page 30, Line 25.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the centralized recommendation system of Desjardins that includes feedback from responses to recommendations made via a number of communication channels with the technique of utilizing the recommendation request context itself (prior to user feedback to the recommendation itself) as taught by Lhuillier. Motivation to combine Desjardins with Lhuillier derives from both references pertaining to multi-channel recommendation systems and from a desire to create “improved recommendations” (Lhuillier: Page 9, Ll. 5-12).
As for, "centralized", "routing, by the recommendation interface computer program, the first recommendation request context to a channel recommendation engine for the first communication channel;" and "...from the channel recommendation engine for the first communication channel": Desjardins teaches communicating to users via multiple channels in at least item 425 of Figure 4 and further describes the communication channels in at least 0071. Desjardins also teaches an interface to control a "plurality of existing communication channels" with each channel having its own business logic as taught in 0071. Desjardins, in view of Lhuillier, discloses disparate communication channels each with its own recommendation engine and communication between each of the channel engines as discussed above. Desjardins, in view of Lhuillier, does not appear to specify a centralized engine that controls the routing of the recommendation contexts. However, Minor teaches a technique of allocating independent data models via a “master system” for use in each respective channel in at least Figure 1, Figure 2, 0034, 0042-0045, with each of the channels feeding data back into a centralized repository in at least 0026-0033 including using "machine learning" in 0032 and a "mean model" in 0033.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the system of Desjardins, in view of Lhuillier, that has channel management via a single module with individual business logic for each channel with the individual "slave modules" for each communications channel of Minor in order to contain the "complexity of purchasing advertisements in individual channels" within each of the "slave computing modules" (Minor: 0012). Motivation to combine Desjardins with Minor comes from the desire to have an "automated way to deal with the complexity and range of online advertising possibilities" (Minor: 0001-0005).
As for, "disconnecting, by the recommendation interface computer program, the plurality of channel recommendation engines in response to the centralized recommendation engine being trained, wherein the plurality of channel recommendation engines remain available to their respective communication channels after being disconnected;": Desjardins discloses an interface to control a "plurality of existing communication channels" with each channel having its own business logic as taught in 0071. Lhuillier teaches a technique of multiple channels each with its own channel recommendation engine that receive channel recommendation contexts from other channels which are then incorporated, via machine learning, into a user’s profile for making further recommendations in at least Page 12, Line 1 – Page 13, Line 5 and Page 29, Line 16 – Page 30, Line 25. In Lhuillier, each of the channel recommendation engines remain available to the channel even if those individual channel recommendation engines are not utilized for each recommendation query, which reads on the engines remaining available after they are “disconnected”. Desjardins, in view of Lhuillier and Minor, does not appear to specify disabling the channel recommendation engines in response to the central machine learning model being trained. However, Sarfarez teaches a technique of training machine learning models wherein the models may be deployed, activated, deactivated, and un-deployed in at least 0069-0072.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the system of Desjardins, in view of Minor, with the technique of activating and de-activating machine learning models as taught by Sarfarez. Motivation to combine Desjardins, in view of Lhuillier and Minor, with Sarfarez to test and change machine learning models as required to achieve the best performance.
Claim(s) 17:
A system, comprising: a plurality of communication channels comprising a mobile communication channel, a web communication channel, a print communication channel and an in-person communication channel;
(Desjardin discloses a myriad of communications channels in at least item 225 of Figure 2 and as described in 0058.)
one or more computer processors executing a plurality of channel recommendation engines; wherein each communication channel is associated with one of the plurality of channel recommendation engines, wherein each of the plurality of communication channels is associated with a different customer interface;
(Desjardins teaches communications channels that comprise different customer interfaces including at least Email, push notification, SMS, Social Networks, Mail, Website in at least 425 of Figure 4 and Figure 5 and a telephony system, telephone voicemail in at least 0058.)
a recommendation electronic device executing a centralized recommendation engine; and a recommendation interface comprising a recommendation interface computer processor executing a recommendation interface computer program, wherein the recommendation interface is in communication with the plurality of communication channels, the plurality of channel recommendation engines, and the centralized recommendation engine;
(Desjardins teaches, in Figure 4, communication channels as item 425, centralized recommendation engine in 405, and 430 with the recommendation interface in 420. Examiner notes that this claim represents a system embodiment that corresponds closely to the method and computer-readable medium independent claims above. This claim isn't grouped with the other embodiments because while the content appears nearly identical in subject matter, the wording is different enough to present independently here.)
wherein:
the recommendation interface computer program is configured to receive a first recommendation request context from a first communication channel of the plurality of communication channels,
wherein the first recommendation request context comprises an identification of the first communication channel and an identification of a first customer that is interacting with the first communication channel;
the recommendation interface computer program is configured to route first recommendation request context to the plurality of channel recommendation engines that is associated with the first communication channel;
the recommendation interface computer program is configured to receive a first recommendation from the channel recommendation engine for the first communication channel, wherein the first recommendation is for a first product or discount;
(Desjardins discloses “real-time product recommendations” in at least 0043, product details in 0072-0074, 0090-0091.)
the recommendation interface computer program is configured to provide the first recommendation and the first recommendation request context to the centralized recommendation engine,
the centralized recommendation engine is configured to train a centralized machine learning engine with the first recommendation and the first recommendation request context
and with recommendations and requests from the channel recommendation engines for the other communication channels;
the recommendation interface computer program is configured to provide the first recommendation to the first communication channel, wherein the first communication channel provides the first recommendation to the first customer,
(Desjardins teaches that the "communication channel orchestration interface" uses "business logic" to select content to display to a user in at least 0071.)
the recommendation interface computer program is configured to receive a first result of the first recommendation from the first communication channel,
wherein the first result comprises a behavioral event including whether the first recommendation was displayed, accepted, declined;
(Desjardins teaches "feedback data" that includes "user interactions" comprising at least "clicks", "opening personalized communication messages", with "opening personalized communications messages" reading on at least “displayed” and “accepted”, and “not opening personalized communication messages” reading on “declined” in at least 0075.)
the recommendation interface computer program is configured to provide the first result to the centralized recommendation engine,
(Desjardins teaches that the "data aggregation module" receives feedback from user interactions with the content presented to the user over the communication channels in at least 0075, 0077.)
wherein the centralized recommendation engine is configured to train the centralized machine learning engine with the first result;
(Desjardins teaches the that "data aggregation module" employs an "AI technique" in at least 0078, 0091, and "AI technologies" in at least 0113 in addition to the "machine learning" algorithm usage taught in at least 0062, 0072. Also see 0122, 0123.)
the recommendation interface computer program is configured to disconnect the plurality of recommendation engines in response to the centralized recommendation engine being trained, wherein the plurality of channel recommendation engines remain available to their respective communication channels after being disconnected;
the recommendation interface computer program is configured to route recommendation requests from the plurality of communication channels to the centralized recommendation engine;
(Desjardins discloses a "communication channel orchestration interface" that integrates with the "data aggregation module" in at least Figure 2 that is described as having "business logic" including "trigger conditions" in at least 0055 that is utilized by the "optimization and personalization engine" to recommend content to send to a user in at least 0056-0059.)
the recommendation interface computer program is configured to receive a second recommendation request context from the first communication channel, wherein the second recommendation request context comprises the identification of the first communication channel and an identification of a second customer that is interacting with the first communication channel;
(Desjardins teaches identifying a user in at least 0051, 0111, and a channel in at least item 425 of Figure 4.)
the recommendation interface computer program is configured to route the second recommendation request context to the centralized recommendation engine;
(Desjardins teaches modifying the marketing actions based upon feedback data analyzed in a central recommendation engine in item 350 of Figure 3 and 0062-0064.)
the recommendation interface computer program is configured to receive a second recommendation from the centralized recommendation engine,
(Desjardin teaches an "optimization and personalization engine" in at least 0072, with communication over multiple communication channels taught in at least 0058 and in 0074.)
wherein the centralized recommendation engine is configured to generate the second recommendation using the centralized machine learning engine, wherein the second recommendation is for a second product or discount;
and the recommendation interface computer program is configured to provide the second recommendation to the first communication channel, wherein the first communication channel provides the second recommendation to the second customer.
(Desjardins: 0072, 0058, 0074. Desjardins discloses “real-time product recommendations” in at least 0043, product details in 0072-0074, 0090-0091.)
As for, “the recommendation interface computer program is configured to receive a first recommendation request context from a first communication channel of the plurality of communication channels,”, “wherein the first recommendation request context comprises an identification of the first communication channel and an identification of a first customer that is interacting with the first communication channel;”, “the recommendation interface computer program is configured to provide the first recommendation and the first recommendation request context to the centralized recommendation engine,”, “the centralized recommendation engine is configured to train a centralized machine learning engine with the first recommendation and the first recommendation request context and with recommendations and requests from the channel recommendation engines for the other communication channels;”:
Desjardins teaches the that "data aggregation module" employs an "AI technique" in at least 0078, 0091, and "AI technologies" in at least 0113 in order to process the "feedback data" from each of the channels in order to formulate a centralized response in at least 0075-0077 in addition to the "machine learning" algorithm usage taught in at least 0062, 0072. Also see 0122, 0123. Desjardins teaches feedback data 435 being used for "prediction, labelling, and recommendations" and subsequently used by the central system for "Data Aggregation" module in at least Figure 4. Desjardins does not appear to specify that the data aggregation model operates on a “first recommendation request context” received from a channel recommendation engine. However, Lhuillier teaches a technique of multiple channels each with its own channel recommendation engine that receive channel recommendation contexts from other channels which are then incorporated, via machine learning, into a user’s profile for making further recommendations in at least Page 12, Line 1 – Page 13, Line 5 and Page 29, Line 16 – Page 30, Line 25.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the centralized recommendation system of Desjardins that includes feedback from responses to recommendations made via a number of communication channels with the technique of utilizing the recommendation request context itself (prior to user feedback to the recommendation itself) as taught by Lhuillier. Motivation to combine Desjardins with Lhuillier derives from both references pertaining to multi-channel recommendation systems and from a desire to create “improved recommendations” (Lhuillier: Page 9, Ll. 5-12).
As for, "centralized", "routes first recommendation request context to one of the plurality of channel recommendation engines that is associated with the first communication channel;", "receives a first recommendation from the channel recommendation engine for the first communication channel;": Desjardins teaches communicating to users via multiple channels in at least item 425 of Figure 4 and further describes the communication channels in at least 0071. Desjardins also teaches an interface to control a "plurality of existing communication channels" with each channel having its own business logic as taught in 0071. Desjardins, in view of Lhuillier, discloses disparate communication channels each with its own recommendation engine and communication between each of the channel engines as discussed above. Desjardins, in view of Lhuillier, does not appear to specify a centralized engine that controls the routing of the recommendation contexts. However, Minor teaches a technique of allocating independent data models via a “master system” for use in each respective channel in at least Figure 1, Figure 2, 0034, 0042-0045, with each of the channels feeding data back into a centralized repository in at least 0026-0033 including using "machine learning" in 0032 and a "mean model" in 0033.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the system of Desjardins, in view of Lhuillier, that has channel management via a single module with individual business logic for each channel with the individual "slave modules" for each communications channel of Minor in order to contain the "complexity of purchasing advertisements in individual channels" within each of the "slave computing modules" (Minor: 0012). Motivation to combine Desjardins, in view of Lhuillier, with Minor comes from the desire to have an "automated way to deal with the complexity and range of online advertising possibilities" (Minor: 0001-0005).
As for, "disconnects the plurality of recommendation engines in response to the centralized recommendation engine being trained, wherein the plurality of channel recommendation engines remain available to their respective communication channels after being disconnected;": Desjardins discloses an interface to control a "plurality of existing communication channels" with each channel having its own business logic as taught in 0071. Lhuillier teaches a technique of multiple channels each with its own channel recommendation engine that receive channel recommendation contexts from other channels which are then incorporated, via machine learning, into a user’s profile for making further recommendations in at least Page 12, Line 1 – Page 13, Line 5 and Page 29, Line 16 – Page 30, Line 25. In Lhuillier, each of the channel recommendation engines remain available to the channel even if those individual channel recommendation engines are not utilized for each recommendation query, which reads on the engines remaining available after they are “disconnected”. Desjardins, in view of Lhuillier, does not appear to specify disabling the channel recommendation engines in response to the central machine learning model being trained. However, Sarfarez teaches a technique of training machine learning models wherein the models may be deployed, activated, deactivated, and un-deployed in at least 0069-0072.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the system of Desjardins, in view of Lhuillier and Minor, with the technique of activating and de-activating machine learning models as taught by Sarfarez. Motivation to combine Desjardins, in view of Lhuillier and Minor, with Sarfarez to test and change machine learning models as required to achieve the best performance.
Claim(s) 2, 10:
receiving, by the recommendation interface computer program, a second result of the second recommendation from the first communication channel;
(Desjardins teaches identifying a user in at least 0051, 0111, and a channel in at least the communications channels 425 in Figure 4 with feedback data returned to the central data aggregation system. The data aggregation module receives "user interactions" in at least 0075. This claim is interpreted as applying the machine learning/central recommendation engine to multiple user interactions over the communications channels.)
and providing, by the recommendation interface computer program, the second result to the centralized recommendation engine, wherein the centralized recommendation engine is configured to train the centralized machine learning engine with the first result.
(Desjardins teaches modifying the marketing actions based upon feedback data analyzed in a central recommendation engine in item 350 of Figure 3 and 0062-0064. The above limitations are interpreted to be claiming an additional set of recommendations similar to that applied to the “first recommendation”.)
Claim(s) 3, 11, 19:
receiving, by the recommendation interface computer program, a third recommendation request context from a second communication channel of the plurality of communication channels,
(Desjardins teaches a plurality of communications channels for presenting content to users in at least 0048, 0057 and teaches that the content may be presented to a user in response to an advertisement request via "web advertising" in at least 0058. Examiner notes that the claim limitations here are interpreted as applying the techniques of claim 1, for example, to subsequent messages sent to user's over a plurality of communications channels. That is, the second and third communications channels are additional channels over which communications are sent to users; communications which are fed back into the central machine learning model to optimize subsequent communications.)
wherein the third recommendation request context comprises an identification of the second communication channel and an identification of a third customer that is interacting with the second communication channel;
(Desjardins teaches identifying a user in at least 0051, 0111, and a channel in at least the communications channels 425 in Figure 4 with feedback data returned to the central data aggregation system.)
routing, by the recommendation interface computer program, the third recommendation request context to a channel recommendation engine for the second communication channel;
receiving, by the recommendation interface computer program, a third recommendation from the channel recommendation engine for the second communication channel, wherein the third recommendation is for a third product or discount;
(Desjardins teaches feedback data 435 being used for "prediction, labelling, and recommendations" and subsequently used by the central system for "Data Aggregation" module in at least Figure 4. Desjardins discloses “real-time product recommendations” in at least 0043, product details in 0072-0074, 0090-0091.)
providing, by the recommendation interface computer program, the third recommendation and the third recommendation request context to the centralized recommendation engine,
(Desjardins teaches that the "data aggregation module" receives feedback from user interactions with the content presented to the user over the communication channels in at least 0075, 0077.)
wherein the centralized recommendation engine is configured to train the centralized machine learning engine with the third recommendation and third first recommendation request context;
(Desjardins teaches the that "data aggregation module" employs an "AI technique" in at least 0078, 0091, and "AI technologies" in at least 0113 in addition to the "machine learning" algorithm usage taught in at least 0062, 0072.)
providing, by the recommendation interface computer program, the third recommendation to the second communication channel, wherein the second communication channel provides the third recommendation to the second customer;
(Desjardins teaches that the "communication channel orchestration interface" uses "business logic" to select content to display to a user in at least 0071.)
receiving, by the recommendation interface computer program, a third result of the third recommendation from the second communication channel; and providing, by the recommendation interface computer program, the third result to the centralized recommendation engine,
(Desjardins teaches identifying a user in at least 0051, 0111, and a channel in at least the communications channels 425 in Figure 4 with feedback data returned to the central data aggregation system. The data aggregation module receives "user interactions" in at least 0075.)
wherein the centralized recommendation engine is configured to train the centralized machine learning engine with the third result.
(Desjardins teaches the that "data aggregation module" employs an "AI technique" in at least 0078, 0091, and "AI technologies" in at least 0113 in addition to the "machine learning" algorithm usage taught in at least 0062, 0072.)
As for, "routing, by the recommendation interface computer program, the third recommendation request context to a channel recommendation engine for the second communication channel;", and "... from the channel recommendation engine for the second communication channel;": The above limitations are interpreted to be claiming an additional set of recommendations similar to that applied to the “first recommendation”. Desjardins does not appear to specify that the data aggregation model operates on a “first recommendation request context” received from a channel recommendation engine. However, Lhuillier teaches a technique of multiple channels each with its own channel recommendation engine that receive channel recommendation contexts from other channels which are then incorporated, via machine learning, into a user’s profile for making further recommendations in at least Page 12, Line 1 – Page 13, Line 5 and Page 29, Line 16 – Page 30, Line 25.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the centralized recommendation system of Desjardins that includes feedback from responses to recommendations made via a number of communication channels with the technique of utilizing the recommendation request context itself (prior to user feedback to the recommendation itself) as taught by Lhuillier. Motivation to combine Desjardins with Lhuillier derives from both references pertaining to multi-channel recommendation systems and from a desire to create “improved recommendations” (Lhuillier: Page 9, Ll. 5-12).
Claim(s) 7, 15 is/are rejected under 35 U.S.C. 103 as being unpatentable over:
Desjardins (Pub. #: US 2021/0319478 A1) in view of
Lhuillier et al. (Pub. #: GB 2444520 A) in view of
Minor et al. (Pub. #: US 2015/0161662 A1) in view of
Sarfarez (Pub. #: US 2021/0241168 A1) in view of
Lopez et al. (Pub. #: US 10,861,052 B1).
Claim(s) 7, 15:
verifying, by the recommendation interface computer program, that the first recommendation has not been presented to the first customer before providing the first recommendation to the first communication channel.
Desjardins teaches optimizing sequential marketing campaigns for presentation to users in at least 0092-0094. Desjardins, in view of Lhuillier, Minor and Sarfarez, does not appear to specify verifying the first recommendation has not been presented to the first customer before providing the recommendation to the communication channel. However, Lopez teaches a technique of controlling the sending of repetitious messages to users via a central manager in at least Col. 6, Line 61 - Col. 7, Line 16.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to modify the system of Desjardins, in view of Lhuillier, Minor and Sarfarez, with the technique of channel management that controls the repetition of messages to users Lopez. Motivation to do so comes from the desire to track and adapt communications to users over the most "effective channel(s) for a particular user, action, or campaign" (Lopez: Col. 4, Ll. 45-49) and to "ensure that the campaign content is no longer displayed to that user once the user takes the corresponding action" (Lopez: Col 7, Ll. 8-16).
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