DETAILED ACTION
Notice of Pre-AIA or AIA Status
1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Double Patenting
2. The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the claims at issue are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); and In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on a nonstatutory double patenting ground provided the reference application or patent either is shown to be commonly owned with this application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The USPTO internet Web site contains terminal disclaimer forms which may be used. Please visit http://www.uspto.gov/forms/. The filing date of the application will determine what form should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to http://www.uspto.gov/patents/process/file/efs/guidance/eTD-info-I.jsp.
3. Claims 1-3, 6-14 and 17-20 are rejected on the ground of nonstatutory obviousness-type double patenting as being unpatentable over claims 1-7, 9-11, 13-14 and 16 of US 12,143,678 B2.
18/915,033 (Instant application),
Claims # 1-3, 6-14 and 17-20
US 12,143,678 B2,
Claims # 1-7, 9-11, 13-14 and 16
1. A method, comprising: determining a placement context for content to be displayed on a user device; generating a feature vector based on the placement context; employing one or more machine learning models to determine a success probability associated with display of the content in accordance with the placement context based on the feature vector; and in response to determining that the success probability exceeds a threshold, providing the content to the user device.
2. The method of claim 1, further comprising: selecting the one or more machine learning models based on the placement context.
3. The method of claim 1, further comprising: employing at least a portion of the feature vector to select the one or more machine learning models.
6. The method of claim 1, further comprising: training the one or more machine learning models using a plurality of input feature vectors mapped to target success possibilities.
7. The method of claim 1, further comprising: training the one or more machine learning models using a set of historical placement patterns and known success possibilities.
8. The method of claim 1, further comprising: training a first machine learning model of the one or more machine learning models using a first set of historical placement patterns corresponding with a first period of time; and training a second machine learning model of the one or more machine learning models using a second set of historical placement patterns corresponding with a second period of time, wherein the second period of time is greater than the first period of time.
9. The method of claim 1, further comprising: training a first machine learning model of the one or more machine learning models using a first set of historical placement patterns corresponding with a first geographical region; and training a second machine learning model of the one or more machine learning models using a second set of historical placement patterns corresponding with a second geographical region, wherein the second geographical region is different from the first geographical region.
10. The method of claim 1, further comprising: obtaining historical placement patterns; identifying a plurality of features from the historical placement patterns; performing feature extraction and feature selection on the plurality of features to generate training features; and training the one or more machine learning models using the training features from the historical placement patterns.
11. The method of claim 1, further comprising: setting the success probability as a likelihood of acquiring a new subscriber in response to a user viewing the content being displayed in the placement context.
12. A computing system, comprising: a memory configured to store computer instructions; and a processor configured to execute the computer instructions to: generate a feature vector for content to be displayed on a user device; employ a machine learning model to determine a success probability associated with display of the content based on the feature vector; and provide the content for display on the user device based on the success probability.
13. The computing system of claim 12, wherein the processor is configured to further execute the computer instructions to: select the machine learning model based on at least a portion of the feature vector;
14. The computing system of claim 12, wherein the processor is configured to further execute the computer instructions to: in response to determining that the success probability exceeds a threshold, identify the content based on at least a portion of the feature vector.
17. The computing system of claim 12, wherein the processor is configured to further execute the computer instructions to: train the machine learning model using a plurality of input feature vectors mapped to target success possibilities.
18. The computing system of claim 12, wherein the processor is configured to further execute the computer instructions to: train the machine learning model using a set of historical placement patterns corresponding with a selected period of time.
19. The computing system of claim 12, wherein the processor is configured to further execute the computer instructions to: train the machine learning model using a set of historical placement patterns corresponding with a selected geographical region.
20. A non-transitory computer-readable storage medium that stores instructions that, when executed by a processor in a computing system, cause the processor to perform actions, the actions comprising: receiving a bid request to display content on a user device; determining a placement context for the secondary content to be displayed on the user device; generating a feature vector based on the placement context; determining a success probability associated with display of the content using the feature vector as input to one or more selected machine learning models; and responding to the bid request based on the success probability.
1. A method, comprising: receiving a request to display secondary content along with primary content on a user device; determining a placement context for the secondary content to be displayed on the user device; generating a feature vector based on the placement context; selecting one or more machine learning models based on the placement context; employing the one or more selected machine learning models to determine a success probability associated with display of the secondary content in accordance with the placement context based on the feature vector; in response to determining that the success probability exceeds a threshold, selecting the secondary content for the request based on the placement context; and providing the selected secondary content to the user device.
7. The method of claim 1, wherein selecting the one or more machine learning models based on the placement context comprises: employing at least a portion of the feature vector to select the one or more machine learning models.
2. The method of claim 1, further comprising: training the one or more machine learning models using a plurality of input feature vectors mapped to target success possibilities.
3. The method of claim 1, further comprising: training the one or more machine learning models using a set of historical secondary placement patterns and known success possibilities.
4. The method of claim 1, further comprising: training a first machine learning model of the one or more machine learning models using a first set of historical secondary placement patterns corresponding with a first period of time; and training a second machine learning model of the one or more machine learning models using a second set of historical secondary placement patterns corresponding with a second period of time, wherein the second period of time is greater than the first period of time.
5. The method of claim 1, further comprising: training a first machine learning model of the one or more machine learning models using a first set of historical secondary placement patterns corresponding with a first geographical region; and training a second machine learning model of the one or more machine learning models using a second set of historical secondary placement patterns corresponding with a second geographical region, wherein the second geographical region is different from the first geographical region.
6. The method of claim 1, further comprising: obtaining historical secondary placement patterns; identifying a plurality of features from the historical secondary placement patterns; performing feature extraction and feature selection on the plurality of features to generate training features; and training the one or more machine learning models using the training features from the historical secondary placement patterns.
9. The method of claim 1, wherein the success probability is a likelihood of acquiring a new subscriber in response to a user viewing the secondary content being displayed in the placement context.
10. A computing system, comprising: a memory configured to store computer instructions; and a processor configured to execute the computer instructions to: receive a request to display secondary content along with primary content on a user device; generate a feature vector for the secondary content to be displayed on the user device; select a machine learning model based on at least a portion of the feature vector; employ the machine learning model to determine a success probability associated with display of the secondary content based on the feature vector; in response to determining that the success probability exceeds a threshold, identify the secondary content for the request based on at least a portion of the feature vector; and respond to the request with the identified secondary content.
11. The computing system of claim 10, wherein the processor is configured to further execute the computer instructions to: train the machine learning model using a plurality of input feature vectors mapped to target success possibilities.
13. The computing system of claim 10, wherein the processor is configured to further execute the computer instructions to: train the machine learning model using a set of historical secondary placement patterns corresponding with a selected period of time.
14. The computing system of claim 10, wherein the processor is configured to further execute the computer instructions to: train the machine learning model using a set of historical secondary placement patterns corresponding with a selected geographical region.
16. A non-transitory computer-readable storage medium that stores instructions that, when executed by a processor in a computing system, cause the processor to perform actions, the actions comprising: receiving a bid request to display secondary content along with primary content on a user device; determining a placement context for the secondary content to be displayed on the user device; generating a feature vector based on the placement context; selecting one or more machine learning models based on the placement context; determining a success probability associated with display of the secondary content in accordance with the placement context using the feature vector as input to the one or more selected machine learning models; in response to determining that the success probability exceeds a threshold value, identifying the secondary content for the bid request based on the placement context; and responding to the bid request with the identified secondary content.
Although the conflicting claims are not identical, they are not patentably distinct from each other because limitations in claims 1-3, 6-14 and 17-20 of the instant application reads on 1-7, 9-11, 13-14 and 16 of US 12,143,678 B2. The claimed limitations recited in the present application are transparently found in US 12,143,678 B2 with obvious wording variations.
Claim Rejections - 35 USC § 102
4. The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale or otherwise available to the public before the effective filing date of the claimed invention.
5. Claim(s) 1 and 4-7 is/are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Ellis et al. (US 2006/0212350 A1, hereinafter referred as “Ellis”).
Regarding claim 1, Ellis discloses a method, comprising:
determining a placement context for content to be displayed on a user device (Abstract discloses ad displayed at a user terminal; and ¶0048 and ¶0170 discloses to select ads for a publisher web page 12 and a user USR, the enhanced online advertising system 174 gathers a large amount information about the page 12, the publisher 76, the advertiser 72, and/or the user 72, and may preferably include many other factors. The gathered and stored information is referred to herein as the ‘context’);
generating a feature vector based on the placement context (¶0179 and ¶0283 discloses machine learning 516 produces a function P that preferably takes as input all of the ‘relevance features’ of the ad 188, the advertiser 72, the target page 12, the publisher 76, and the user USR)… These features are represented as a feature vector 814);
employing one or more machine learning models to determine a success probability associated with display of the content in accordance with the placement context based on the feature vector (¶0179 and ¶0283 discloses the probability of action 86 is estimated using a machine-learned model that takes as input a relevance feature vector 814 (FIG. 29) that measures various attributes of an ad 188 and the context of where the ad 188 is being shown and the user USR it is being shown to); and
in response to determining that the success probability exceeds a threshold (¶0305 discloses the estimated RPAI of each ad 188 is expressed as a confidence interval of minimum and maximum RPAI, shown as: [minRPAI(ad) . . . maxRPAI(ad)], where the confidence is a threshold set by the system, e.g. 95 percent; and ¶0220 discloses the system 174 estimates an ad's RPAI as the probability of a user USR taking at least one action 86 on the ad 188), providing the content to the user device (Abstract discloses ad displayed at a user terminal).
Regarding claim 4, Ellis discloses the method of claim 1, further comprising: selecting the content based on the placement context (¶0183-¶0184 discloses the enhanced online advertising system 174 preferably incorporate both text features and behavioral features, e.g. past behavior of users, to determine and present ads 188 having the greatest predicted value to each user USR).
Regarding claim 5, Ellis discloses the method of claim 1, further comprising: employing at least a portion of the feature vector to select the content for display (¶0179 discloses Machine learning 516 produces a function P that preferably takes as input all of the ‘relevance features’ of the ad 188, the advertiser 72, the target page 12, the publisher 76, and the user USR, and predicts the probability that the user USR will take action 86 on that ad 188 in that context).
Regarding claim 6, Ellis discloses the method of claim 1, further comprising: training the one or more machine learning models using a plurality of input feature vectors mapped to target success possibilities (¶0180 discloses 174, the machine learning algorithm 516 produces P using training data 132, e.g. comprising at least a portion of the tracked actions 86, that typically comprises a very large number, e.g. 100K or more, of examples, where each example corresponds to an ad 188 presented in a context to a user USR, and whether or not the user USR took action 86. Each example is represented as an input feature vector and output of 0 (no action) or 1 (action)).
Regarding claim 7, Ellis discloses the method of claim 1, further comprising: training the one or more machine learning models (¶0180 discloses the machine learning algorithm 516 produces P using training data 132, e.g. comprising at least a portion of the tracked actions 86, that typically comprises a very large number, e.g. 100K or more, of examples, where each example corresponds to an ad 188 presented in a context to a user USR, and whether or not the user USR took action 86) using a set of historical placement patterns and known success possibilities (¶0283 discloses probability of action 86 is estimated using a machine-learned model that takes as input a relevance feature vector 814; ¶0184 discloses the enhanced online advertising system 174 provide a wide variety of relevance features, such as… context identification; and Contextual identification may also be used by the system 174 in regard to the position of available ad space 184, such as at the top, right, left, or bottom of a publisher page 12, such as wherein the position is determined by the system 174 or provided to the system 174 by code associated with the publisher page 12).
Claim Rejections - 35 USC § 103
6. 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 of this title, 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.
7. Claim(s) 12, 14-15 and 17-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ellis in view of Xu et al. (US 2023/0316106 A1, hereinafter referred as “Xu6106”).
Regarding claim 12, Ellis discloses a computing system, comprising:
…generate a feature vector for content to be displayed on a user device (¶0179 and ¶0283 discloses machine learning 516 produces a function P that preferably takes as input all of the ‘relevance features’ of the ad 188, the advertiser 72, the target page 12, the publisher 76, and the user USR)… These features are represented as a feature vector 814);
employ a machine learning model to determine a success probability associated with display of the content based on the feature vector (¶0179 and ¶0283 discloses the probability of action 86 is estimated using a machine-learned model that takes as input a relevance feature vector 814 (FIG. 29) that measures various attributes of an ad 188 and the context of where the ad 188 is being shown and the user USR it is being shown to); and
provide the content for display on the user device (Abstract discloses ad displayed at a user terminal) based on the success probability (¶0305 discloses the estimated RPAI of each ad 188 is expressed as a confidence interval of minimum and maximum RPAI, shown as: [minRPAI(ad) . . . maxRPAI(ad)], where the confidence is a threshold set by the system, e.g. 95 percent; and ¶0220 discloses the system 174 estimates an ad's RPAI as the probability of a user USR taking at least one action 86 on the ad 188).
Ellis doesn’t disclose a memory configured to store computer instructions; and a processor configured to execute the computer instructions.
However, in the same field of endeavor, Xu6106 discloses a memory configured to store computer instructions (¶0246 discloses the memory further includes one or more programs, which are stored in the memory and are configured to be executed by the CPU); and a processor configured to execute the computer instructions (¶0247 discloses at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the method for training a content recommendation model or the content recommendation method provided in the foregoing method embodiments).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Ellis for the purpose of having improved flexibility and adaptability so that new features and placement contexts can be added as needed.
Regarding claim 14, Ellis discloses the computing system of claim 12, …in response to determining that the success probability exceeds a threshold, identify the content based on at least a portion of the feature vector (¶0183-¶0184 discloses the enhanced online advertising system 174 preferably incorporate both text features and behavioral features, e.g. past behavior of users, to determine and present ads 188 having the greatest predicted value to each user USR).
Ellis doesn’t disclose wherein the processor is configured to further execute the computer instructions.
However, in the same field of endeavor, Xu6106 discloses wherein the processor is configured to further execute the computer instructions (¶0247 discloses at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the method for training a content recommendation model or the content recommendation method provided in the foregoing method embodiments).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Ellis for the purpose of having improved flexibility and adaptability so that new features and placement contexts can be added as needed.
Regarding claim 15, Ellis discloses the computing system of claim 12, …to: determine a placement context for content to be displayed on the user device (Abstract discloses ad displayed at a user terminal; and ¶0048 and ¶0170 discloses to select ads for a publisher web page 12 and a user USR, the enhanced online advertising system 174 gathers a large amount information about the page 12, the publisher 76, the advertiser 72, and/or the user 72, and may preferably include many other factors. The gathered and stored information is referred to herein as the ‘context’); and generate a feature vector based on the placement context (¶0179 and ¶0283 discloses machine learning 516 produces a function P that preferably takes as input all of the ‘relevance features’ of the ad 188, the advertiser 72, the target page 12, the publisher 76, and the user USR)… These features are represented as a feature vector 814).
Ellis doesn’t disclose wherein the processor generates the feature vector for the content by being configured to further execute the computer instructions.
However, in the same field of endeavor, Xu6106 discloses wherein the processor generates the feature vector for the content (¶0050 discloses corresponding feature vectors 102 are determined based on the data features 101) by being configured to further execute the computer instructions (¶0247 discloses at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the method for training a content recommendation model or the content recommendation method provided in the foregoing method embodiments).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Ellis for the purpose of having improved flexibility and adaptability so that new features and placement contexts can be added as needed.
Regarding claim 17, Ellis discloses the computing system of claim 12, …train the machine learning model using a plurality of input feature vectors mapped to target success possibilities (¶0180 discloses 174, the machine learning algorithm 516 produces P using training data 132, e.g. comprising at least a portion of the tracked actions 86, that typically comprises a very large number, e.g. 100K or more, of examples, where each example corresponds to an ad 188 presented in a context to a user USR, and whether or not the user USR took action 86. Each example is represented as an input feature vector and output of 0 (no action) or 1 (action)).
Ellis doesn’t disclose wherein the processor is configured to further execute the computer instructions.
However, in the same field of endeavor, Xu6106 discloses wherein the processor is configured to further execute the computer instructions (¶0247 discloses at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the method for training a content recommendation model or the content recommendation method provided in the foregoing method embodiments).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Ellis for the purpose of having improved flexibility and adaptability so that new features and placement contexts can be added as needed.
Regarding claim 18, Ellis discloses the computing system of claim 12,…train the machine learning model (¶0180 discloses machine learning algorithm 516 produces P using training data 132, e.g. comprising at least a portion of the tracked actions 86, that typically comprises a very large number, e.g. 100K or more, of example) using a set of historical placement patterns corresponding with a selected period of time (¶0046 discloses provide an advertising system across a network, that analyzes both publisher content and advertiser content, past user behavior, profile information of users, past rates of performance of ads, time of day and day of week, and/or many other factors to determine relevance of ads to be displayed with publisher content, wherein the relevance is based on a prediction of response by the user).
Ellis doesn’t disclose wherein the processor is configured to further execute the computer instructions.
However, in the same field of endeavor, Xu6106 discloses wherein the processor is configured to further execute the computer instructions (¶0247 discloses at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the method for training a content recommendation model or the content recommendation method provided in the foregoing method embodiments).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Ellis for the purpose of having improved flexibility and adaptability so that new features and placement contexts can be added as needed.
Regarding claim 19, Ellis discloses the computing system of claim 12, …train the machine learning model (¶0180 discloses machine learning algorithm 516 produces P using training data 132, e.g. comprising at least a portion of the tracked actions 86, that typically comprises a very large number, e.g. 100K or more, of example) using a set of historical placement patterns corresponding with a selected geographical region (¶0115 discloses audience information 428a may provide information as to the intended or actual audience of a publisher web site 14, while Geo information 428n may provide information as to the actual location of the user USR, or an intended service region of the publisher web site 14).
Ellis doesn’t disclose wherein the processor is configured to further execute the computer instructions.
However, in the same field of endeavor, Xu6106 discloses wherein the processor is configured to further execute the computer instructions (¶0247 discloses at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the method for training a content recommendation model or the content recommendation method provided in the foregoing method embodiments).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Ellis for the purpose of having improved flexibility and adaptability so that new features and placement contexts can be added as needed.
Regarding claim 20, Ellis discloses the. …the actions comprising:
receiving a bid request to display content on a user device (¶0048 discloses the enhanced online advertising system receives ad request from a publisher associated with placement on a publisher web page; and ¶0101-¶0102 discloses in… the enhanced online advertising system 174, advertisers bid 204 for the cost per action (CPA) prices 252 that they are willing to pay for their respective actions 86);
determining a placement context for the secondary content to be displayed on the user device (Abstract discloses ad displayed at a user terminal; and ¶0048 and ¶0170 discloses to select ads for a publisher web page 12 and a user USR, the enhanced online advertising system 174 gathers a large amount information about the page 12, the publisher 76, the advertiser 72, and/or the user 72, and may preferably include many other factors. The gathered and stored information is referred to herein as the ‘context’);
generating a feature vector based on the placement context (¶0179 and ¶0283 discloses machine learning 516 produces a function P that preferably takes as input all of the ‘relevance features’ of the ad 188, the advertiser 72, the target page 12, the publisher 76, and the user USR)… These features are represented as a feature vector 814);
determining a success probability associated with display of the content using the feature vector as input to one or more selected machine learning models (¶0179 and ¶0283 discloses the probability of action 86 is estimated using a machine-learned model that takes as input a relevance feature vector 814 (FIG. 29) that measures various attributes of an ad 188 and the context of where the ad 188 is being shown and the user USR it is being shown to); and
responding to the bid request based on the success probability (¶0305 discloses the estimated RPAI of each ad 188 is expressed as a confidence interval of minimum and maximum RPAI, shown as: [minRPAI(ad) . . . maxRPAI(ad)], where the confidence is a threshold set by the system, e.g. 95 percent; and ¶0220 discloses the system 174 estimates an ad's RPAI as the probability of a user USR taking at least one action 86 on the ad 188).
Ellis doesn’t disclose a non-transitory computer-readable storage medium that stores instructions that, when executed by a processor in a computing system, cause the processor to perform actions.
However, in the same field of endeavor, Xu6106 discloses a non-transitory computer-readable storage medium that stores instructions that, when executed by a processor in a computing system, cause the processor to perform actions (¶0249 discloses processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the method for training a content recommendation model and the content recommendation method according to any one of the foregoing embodiments).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Ellis for the purpose of having improved flexibility and adaptability so that new features and placement contexts can be added as needed.
8. Claim(s) 2 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ellis in view of Xu et al. (US 2017/0161773 A1, hereinafter referred as “Xu1773”).
Regarding claim 2, Ellis doesn’t disclose the method of claim 1, further comprising: selecting the one or more machine learning models based on the placement context.
However, in the same field of endeavor, Xu1773 discloses selecting the one or more machine learning models based on the placement context (¶0221 discloses control circuitry of a system may implement a hybrid approach that employs multiple machine learning techniques (e.g., ANN, SVM, fourier transforms, Bayesian linear regression, etc.). For example, each of the techniques or combinations thereof may be configured for predictions across a specified period in the future (e.g., 2-4 weeks, 4-6 weeks, etc.). Control circuitry may compare a specified time of an advertising event with one or more thresholds to determine an approach machine learning technique to employ).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Ellis for the purpose of providing accuracy, stability and robustness across various data sets.
9. Claim(s) 3 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ellis in view of Lin et al. (US 8,370,280 B1, hereinafter referred as “Lin”).
Regarding claim 3, Ellis doesn’t disclose the method of claim 1, further comprising: employing at least a portion of the feature vector to select the one or more machine learning models.
However, in the same field of endeavor, Lin discloses employing at least a portion of the feature vector to select the one or more machine learning models (col. 9, lines 22-39 discloses features of a first type (e.g., string data) can be submitted to a first predictive model and features of a second type (e.g., binary data) can be submitted to a second predictive model. This can be advantageous when an occurrence is described in different domains (e.g., time and temperature or color and size).)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Ellis for the purpose of an occurrence is described in different domains (e.g., time and temperature or color and size) (col. 9, lines 31-33).
10. Claim(s) 10 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ellis in view of Huang et al. (CN 115018533 B, hereinafter referred as “Huang”).
Regarding claim 10, Ellis doesn’t disclose the method of claim 1, further comprising: obtaining historical placement patterns;
identifying a plurality of features from the historical placement patterns; performing feature extraction and feature selection on the plurality of features to generate training features; and training the one or more machine learning models using the training features from the historical placement patterns.
However, in the same field of endeavor, Huang discloses obtaining historical placement patterns (pg. 4 discloses second acquisition unit is used to acquire media delivery data of sample clue advertisements); identifying a plurality of features from the historical placement patterns (pg. 4 discloses the media delivery data of sample clue advertisements includes basic information data of sample advertisers, planned sample delivery information data and historical sample delivery information data); performing feature extraction and feature selection on the plurality of features to generate training features (pg. 4 discloses second extraction unit is used to extract features from the basic information data, planned sample delivery information data and historical sample delivery information data of the sample advertiser to obtain a sample feature vector corresponding to the sample lead advertisement); and training the one or more machine learning models using the training features from the historical placement patterns (pg. 4 discloses training unit is used to train the initial conversion quantity estimation model according to the sample feature vector of the sample clue advertisement, the sample constraint feature vector and the conversion quantity estimation classification label corresponding to the sample clue advertisement to generate the conversion quantity estimation model).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Ellis so that the ML model learns for the signal, not noise thus increasing success-probability accuracy.
11. Claim(s) 11 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ellis in view of Collette et al. (US 2014/0006170 A1, hereinafter referred as “Collette”).
Regarding claim 11, Ellis discloses the method of claim 1, further comprising: setting the success probability as a likelihood of [user clicking an advertisement] in response to a user viewing the content being displayed in the placement context (¶0283 discloses the probability of action 86 is estimated using a machine-learned model that takes as input a relevance feature vector 814 (FIG. 29) that measures various attributes of an ad 188 and the context of where the ad 188 is being shown and the user USR it is being shown to).
Ellis doesn’t disclose likelihood of acquiring a new subscriber.
However, in the same field of endeavor, Collette discloses likelihood of acquiring a new subscriber (¶0300 discloses the amount that the advertiser is willing to pay to show the ad to the target consumer is determined based on the consumer's likelihood of subscribing. These questions are answered by the data collected in learning mode).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Ellis for the purpose of generating a recurring value through a repeat customer.
12. Claim(s) 13 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ellis in view of Xu6106 and in further view of Lin.
Regarding claim 13, Ellis doesn’t disclose the computing system of claim 12, wherein the processor is configured to further execute the computer instructions to: select the machine learning model based on at least a portion of the feature vector.
However, in the same field of endeavor, Xu6106 discloses wherein the processor is configured to further execute the computer instructions (¶0247 discloses at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the method for training a content recommendation model or the content recommendation method provided in the foregoing method embodiments).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Ellis for the purpose of having improved flexibility and adaptability so that new features and placement contexts can be added as needed.
Ellis as modified still doesn’t disclose select the machine learning model based on at least a portion of the feature vector.
However, in the same field of endeavor, Lin discloses select the machine learning model based on at least a portion of the feature vector (col. 9, lines 22-39 discloses features of a first type (e.g., string data) can be submitted to a first predictive model and features of a second type (e.g., binary data) can be submitted to a second predictive model. This can be advantageous when an occurrence is described in different domains (e.g., time and temperature or color and size)).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to further modify Ellis for the purpose of an occurrence is described in different domains (e.g., time and temperature or color and size) (col. 9, lines 31-33).
13. Claim(s) 16 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ellis in view of Xu6106 and in further view of Xu1773.
Regarding claim 16, Ellis discloses the computing system of claim 12, …determine a placement context for content to be displayed on the user device (Abstract discloses ad displayed at a user terminal; and ¶0048 and ¶0170 discloses to select ads for a publisher web page 12 and a user USR, the enhanced online advertising system 174 gathers a large amount information about the page 12, the publisher 76, the advertiser 72, and/or the user 72, and may preferably include many other factors. The gathered and stored information is referred to herein as the ‘context’)...
Ellis doesn’t disclose wherein the processor is configured to further execute the computer instructions to:… and selecting the one or more machine learning models based on the placement context.
However, in the same field of endeavor, Xu6106 discloses wherein the processor is configured to further execute the computer instructions (¶0247 discloses at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the method for training a content recommendation model or the content recommendation method provided in the foregoing method embodiments).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Ellis for the purpose of having improved flexibility and adaptability so that new features and placement contexts can be added as needed.
Ellis as modified doesn’t disclose …selecting the one or more machine learning models based on the placement context.
However, in the same field of endeavor, Xu1773 discloses selecting the one or more machine learning models based on the placement context (¶0221 discloses control circuitry of a system may implement a hybrid approach that employs multiple machine learning techniques (e.g., ANN, SVM, fourier transforms, Bayesian linear regression, etc.). For example, each of the techniques or combinations thereof may be configured for predictions across a specified period in the future (e.g., 2-4 weeks, 4-6 weeks, etc.). Control circuitry may compare a specified time of an advertising event with one or more thresholds to determine an approach machine learning technique to employ).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to further modify Ellis for the purpose of providing accuracy, stability and robustness across various data sets.
Allowable Subject Matter
14. Claims 8-9 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to PRIYANK J SHAH whose telephone number is (571)270-3732. The examiner can normally be reached on 10:00 - 6:00 M-F.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, LunYi Lao can be reached on 5712727671. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/PRIYANK J SHAH/Primary Examiner, Art Unit 2621