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
Applicant’s Remarks, filed January 30th, 2026, has been fully considered and entered. Accordingly, claims 1-20 are pending in the case. Claims 1, 8, and 15 are the independent claims.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries set forth in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966), that are applied 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.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1-20 are being rejected under 35 U.S.C. 103 as being unpatentable over Tuschman et al. (US 2019/0102802 A1) in view of Redkar et al. (US 2018/0276553 A1), further in view of Poulin (US 2014/0245207 A1).
Regarding claim 1, Tuschman teaches a method comprising:
maintaining an enterprise environment comprising a plurality of clients and a plurality of associated client data, wherein the plurality of clients includes companies (see Tuschman, Paragraph [0037], “The system 100 may include one or more clients, and three such clients are shown, by way of example, in FIG. 1. An additional system 105 may be included, and this may be similar to one of the client systems 103.” [Figure 1 shows an enterprise environment comprising clients and associated client data.]);
generating one or more statistical models, via a statistical learning and predictive estimation system, by analyzing the client data in the enterprise environment using one or more statistical algorithms (see Tuschman, Paragraph [0020], “generating psychometric models of online users of a population based on automatically machine-collected data about online behavior of such users” [The psychometric models (i.e., statistical models) of users are generated.]);
storing the statistical models in a model database (see Tuschman, Paragraph [0077], “The method may include in 214 storing the generated psychometric profiles (the psychometric models), e.g., in a database.” [The psychometric models (i.e., statistical models) are stored.]);
However, Tuschman does not explicitly teach:
receiving a prediction estimate request from one of the plurality of clients with respect to the associated client data for the client, the prediction estimate request including logistics of managing a project;
Redkar teaches:
receiving a prediction estimate request from one of the plurality of clients with respect to the associated client data for the client, the prediction estimate request including logistics of managing a project (see Redkar, Paragraphs [0017], [0025], “The model management system 120 may receive user statements that include requests for information via one or more communications channels… IT resources 150 may include data sets for any other network or enterprise resources that may be monitored.” [A request (i.e., prediction estimate request that includes logistics of managing a project) may be received.]);
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined Tuschman (teaching predicting psychometric profiles from behavioral data using machine-learning while maintaining user anonymity) in view of Redkar (teaching a system for querying models), and arrived at a method that receives a prediction estimate request. One of ordinary skill in the art would have been motivated to make such a combination for the purposes of providing a result to the user request (see Redkar, Paragraph [0014]). In addition, both the references (Tuschman and Redkar) teach features that are directed to analogous art and they are directed to the same field of endeavor, such as models. The close relation between both of the references highly suggests an expectation of success.
The combination of Tuschman, and Redkar further teaches:
selecting, using a clustering algorithm, a subset of the associated client data based at least on one or more characteristics of the associated client data, wherein the clustering algorithm reduces input dimensions by using metadata about the characteristics of the associated client data including whether an input variable is text data with free user input (see Tuschman, Paragraphs [0040], [0047]-[0048], [0131], [0150], “The behavioral information collected includes data on users' current and past online activity, including users' browsing history of websites and web pages visited, engagement behavior on the websites, search queries, and in-application behavior … the behavioral data in behavioral database 126 may be in raw form. An analysis method is used to reduce dimensionality of the data to summary form. … the at least one machine-learning method includes carrying out unsupervised clustering on an assumed number of clusters, e.g., three clusters, or four clusters, using the psychometric models as features, and examining the so-formed clusters to select the one or more clusters that has the largest proportion or the greatest number of engaged users. These clusters form a learned classification method that can be used to classify users according to engagement, i.e., an engagement model.” [A subset of data is selected using an unsupervised clustering algorithm (i.e., clustering algorithm). The behavioral data (i.e., metadata about the characteristics of the associated client data including whether an input variable is text data with data with free user input) may be reduced by the unsupervised clustering algorithm.]);
selecting, using machine learning techniques, a statistical model from the one or more statistical models based at least on the subset of the associated client data and the one or more characteristics, wherein the statistical model is automatically selected based on the size and complexity of the subset of the associated client data (see Tuschman, Paragraphs [0020], [0031], [0041], [0077], [0205], “Depending on the particular dimension being modeled, there are three types of classifications: binary classification (predicting one of two possible outcomes), multiclass classification (predicting one of more than two outcomes) and regression (predicting a numeric value). One embodiment comprises training a plurality of machine-learning methods, carries out cross-validation, e.g., so-called k-fold cross-validation, and selects a machine-learning method and corresponding model according to a machine-learning method selection criterion. In one embodiment, the selection of the model that provides the best performance according to a performance criterion. The criterion used depends on the type of classification. In one embodiment, 10-fold cross-validation is carried out for selecting the best-performance model.” [A psychometric model (i.e., statistical model) is selected using machine learning based on the dimension being modeled (i.e., size and complexity of the subset of the associated client data).]);
applying the selected statistical model to the subset of the associated client data and the one or more selected characteristics to generate one or more prediction estimates and one or more statistical outliers, wherein the prediction estimates include prediction confidence related to the quality of the statistical model used to perform the prediction (see Tuschman, Paragraphs [0109], [0110], “These pairs of (summary) behavioral data and corresponding psychometric profile for each of N5 users form a training data set for a machine-learning process that determines (“statistically learns”) a prediction method of predicting a psychometric profile, i.e., determining a psychometric model of a user from the (summary) behavioral data of that user, e.g., by trying one or more prediction methods for each dimension and selecting the best prediction method for each dimension. … Once the prediction method is determined, in one embodiment PDAE 108 sends the target population provider system 102 containing the target population and behavioral data thereof an indication 411 that PDAE 108 can carry out large-scale prediction.” Also, see Redkar, Paragraph [0035], “The model registry 320 may include information for each model such as one or more functions that each model offers (e.g., forecasting, regression, classification, anomaly detection, etc.)” [The selected psychometric model (i.e., statistical model) is used to generate data (i.e., predication estimates and statistical outliers).]);
However, the combination of Tuschman, and Redkar do not explicitly teach:
wherein the prediction estimates include prediction confidence related to the quality of the statistical model used to perform the prediction;
Poulin teaches:
wherein the prediction estimates include prediction confidence related to the quality of the statistical model used to perform the prediction (see Poulin, Paragraph [0052], “For FIG. 4A, the search result interface 120 when the rating icon 114 a (of icons 114 a-114 e FIG. 4) is selected is shown for Boeing. Displayed in a screen are results of the prediction that the model forecasts as a “Forecast Delta”, which in this case is negative 3.3323 . . . and a corresponding visual graph on the right that indicates ‘Negative’ and shows a probability or confidence value that is calculated for the model prediction.” [The prediction estimate may include a confidence value (i.e., prediction confidence).]);
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have combined Tuschman (teaching predicting psychometric profiles from behavioral data using machine-learning while maintaining user anonymity) in view of Redkar (teaching a system for querying models), further in view of Poulin (teaching interfaces for predictive models), and arrived at a method that incorporates a prediction confidence. One of ordinary skill in the art would have been motivated to make such a combination for the purposes of improving data analysis (see Poulin, Paragraph [0004]). In addition, the references (Tuschman, Redkar and Poulin) teach features that are directed to analogous art and they are directed to the same field of endeavor, such as models. The close relation between the references highly suggests an expectation of success.
The combination of Tuschman, Redkar, and Poulin further teaches:
and providing a visual arrangement of the one or more prediction estimates including entries showing prediction confidence and one or more statistical outliers to the client, wherein the statistical learning and predictive estimation system is fully integrated with the enterprise environment in order to avoid consuming computation time transforming data from one format to another, wherein the visual arrangement includes a model quality map (see Tuschman, Paragraphs [0110], [0123], [0151], [0215], [0219], “As an example of how such an engagement model may be used, FIGS. 9A and 9B show a graphical display of the results of determining an engagement model of users, using the 32-dimensional psychometric profiles of the example profile shown in FIG. 8. In the test whose results are shown in FIG. 8, there were 300 positive engagements and 42,000 negative engagements. … FIG. 10B shows a map of DMAs in the United States, wherein each DMA can be color coded according to its likelihood of engagement.” Also, see Redkar, Paragraph [0035], “The model registry 320 may include information for each model such as one or more functions that each model offers (e.g., forecasting, regression, classification, anomaly detection, etc.)” Also, see Poulin, Paragraph [0052], “For FIG. 4A, the search result interface 120 when the rating icon 114 a (of icons 114 a-114 e FIG. 4) is selected is shown for Boeing. Displayed in a screen are results of the prediction that the model forecasts as a “Forecast Delta”, which in this case is negative 3.3323 . . . and a corresponding visual graph on the right that indicates ‘Negative’ and shows a probability or confidence value that is calculated for the model prediction. [The prediction estimates including the prediction confidence, and statistical outliers may be displayed. Figure 10B shows a map of designated market areas according the likelihood of engagement (i.e., model quality map).]).
Regarding claim 2, Tuschman in view of Redkar, further in view of Poulin teaches all the limitations of claim 1. Tuschman further teaches:
wherein the one or more statistical models comprise at least one of gradient boosted trees, linear models, and centroids (see Tuschman, Paragraph [0226], “the machine-learning (ML) methods described herein in PDAE 108 use algorithms and utilities provided in Spark and part of Apache Spark's MLlib. Spark's MLlib provides methods usable for binary classification, logistic regression, naive Bayes, and others; for regression, generalized linear regression, survival regression, and others; for decision trees, random forests, and gradient-boosted trees;” [The machine learning algorithms (i.e., statistical models) include gradient boosted trees and generalized linear regression (i.e., linear models).]).
Regarding claim 3, Tuschman in view of Redkar, further in view of Poulin teaches all the limitations of claim 1. Tuschman further teaches:
wherein selecting the statistical model includes using cross-validation (see Tuschman, Paragraph [0209], “As for the binary and multiclass classifiers, several machine-learning methods are used, and the best is selected using cross-validation.” [Cross-validation is used to select the best machine learning algorithms (i.e., statistical models).]).
Regarding claim 4, Tuschman in view of Redkar, further in view of Poulin teaches all the limitations of claim 1. Tuschman further teaches:
wherein input into statistical models includes data and metadata (see Tuschman, Paragraphs [0040], [0055], “The seed users' psychometric profile data and corresponding behavioral data, e.g., as summary behavioral data are used as seed data for at least one machine-learning method to determine a method of predicting a psychometric profile of a person from that person's behavioral data” [The machine learning algorithms (i.e., statistical models) includes data and behavioral data (i.e., metadata).]).
Regarding claim 5, Tuschman in view of Redkar, further in view of Poulin teaches all the limitations of claim 1. Tuschman further teaches:
wherein the one or more predictions include a predicted value of the outcome of the subset of associated client data, a quality flag, and a subset of input fields to produce the one or more predictions (see Tuschman, Paragraphs [0033], [0051], [0205], “the relative likelihood of engagement can be predicted based as a function of psychometric dimensions… Filter program code 128 is operative to filter user records in user ID database 124, for example to exclude or flag those users that meet some pre-determined criteria, e.g., those users that have a relatively low amount behavioral data in the behavioral database 126… Depending on the particular dimension being modeled, there are three types of classifications: binary classification (predicting one of two possible outcomes), multiclass classification (predicting one of more than two outcomes) and regression (predicting a numeric value).” [The likelihood of engagement prediction value is based on a subset of user data, and filtering user records using a flag.]).
Regarding claim 6, Tuschman in view of Redkar, further in view of Poulin teaches all the limitations of claim 1. Tuschman further teaches:
wherein meta-data is used to reduce the number of input dimensions of the associated client data (see Tuschman, Paragraph [0048], “An analysis method is used to reduce dimensionality of the data to summary form.” [The behavioral data (i.e., meta-data) may be used to reduce the input dimensions.]).
Regarding claim 7, Tuschman in view of Redkar, further in view of Poulin teaches all the limitations of claim 1. Tuschman further teaches:
wherein a plurality of one or more statistical algorithms are implemented at a core level of software (see Tuschman, Paragraph [0226], “the machine-learning (ML) methods described herein in PDAE 108 use algorithms and utilities provided in Spark and part of Apache Spark's MLlib.” [The psychometric data analytics engine (i.e., a core level of software) implements the machine learning algorithms (i.e., statistical models).]).
Regarding claims 8-20, Tuschman in view of Redkar, further in view of Poulin teaches all of the limitations of claims 1-7 in method form rather than in system, and non-transitory computer readable medium form. Tuschman also discloses a system [0038], and non-transitory computer readable medium [0038]. Therefore, the supporting rationale of the rejection to claim 1-7 applies equally as well to those limitations of claims 8-20.
Response to Arguments
Applicant’s Arguments, filed January 30th, 2026 have been fully considered, but are not persuasive.
Applicant argues on pages 7-10 of Applicant's Remarks that the cited references do not teach or suggest “wherein the clustering algorithm reduces input dimensions by using metadata about the characteristics of the associated client data including whether an input variable is text data with free user input.” The Examiner respectfully disagrees.
Tuschman discloses in paragraph [0153], “Automatically collected behavioral data on users as used herein means online activity (including activity on its application, network, or exchange). While in many examples embodiments described herein, behavioral data includes data on websites visited by users, behavioral data may include user-generated text in an application, and/or consumer data, and/or user-preference data, and/or first-party data, and/or web-log data. While the analysis method described herein above is for textual analysis of websites visited by users, behavioral data may include or instead be comprised of one or more of images, audio, text messages, emails, blogs produced (or read), data documents, text files, database files, log files, transaction records, purchase orders, and so forth. Thus, while the analysis process described herein comprises analyzing text from online behavior, the analyzing for example including applying unsupervised classification to the text, in other embodiments the analysis process to form the summary behavioral data for a user comprises analyzing at least one image and/or at least one audio element from online behavior of the user, the analyzing for example including applying unsupervised classification to the at least one image and/or at least one audio element.” As shown, behavioral data may include user-generated data (i.e., metadata about the characteristics of the associated client data including whether an input variable is text data with free user input), in which an unsupervised classification (i.e., clustering algorithm) may be applied to the behavioral data.
Therefore, Tuschman teaches reducing dimensionality of the behavioral data such as user-generated text (i.e., metadata about the characteristics of the associated client data including whether an input variable is text data with free user input) using an unsupervised classification method (i.e., clustering algorithm).
For the above reasons, it is believed that the rejections should be sustained.
Conclusion
THIS ACTION IS MADE FINAL. 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 extension fee 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 date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to HUSAM TURKI SAMARA whose telephone number is (571)272-6803. The examiner can normally be reached on Monday - Thursday, Alternate Fridays.
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, Apu Mofiz can be reached on (571)-272-4080. 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.
/HUSAM TURKI SAMARA/Examiner, Art Unit 2161
/APU M MOFIZ/Supervisory Patent Examiner, Art Unit 2161