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 the Amendment
This action is in response to the submission filed 5 March 2026 for application 17/741,129. Claims 5 and 14 have been cancelled. Currently claims 1-4, 6-13, and 15-22 are pending and have been examined.
Response to the Arguments
Regarding Applicant’s arguments, see pages 8-11, filed 05 March 2026, with respect to Claim Rejections under 35 USC § 101, Applicant argues on Pages 8 and 9 that Claim 1 recites training a naive Bayes classifier on training data derived through k-means clustering, computing a level of relevance using a natural language processing algorithm, computing a classification label using the trained naive Bayes classifier, computing a linear regression over target data on different points in time to generate numerical performance values, and using those generated values to train an LSTM neural network to compute performance forecasts. None of these operations can practically be performed in the human mind. Training a naive Bayes classifier requires iterating over training samples to compute prior probabilities and conditional probability distributions for each feature per class, and then storing those learned parameters for subsequent use in classification. Training an LSTM neural network requires forward propagation through input, forget, and output gate structures, cell state computation, backpropagation through time, gradient descent optimization, and iterative weight matrix updates over multiple epochs. K-means clustering requires iteratively assigning data points to clusters, computing cluster centroids, and minimizing within-cluster variance across multiple iterations until convergence. An NLP algorithm performs tokenization, numerical vectorization of text, and statistical computation of similarity between high- dimensional vector representations.
The Office Action discusses a simplified hypothetical in which a person mentally compares a training transcript to a job title, but this hypothetical strips away the computational operations that these algorithms perform and substitutes a subjective human judgment that bears no resemblance to the claimed operations. These are not generic computer components applied to a mental process. Rather, they are specific computational algorithms, each with defined mathematical and procedural operations, arranged in a defined sequence where the output of each step feeds the input of the next. Under MPEP 2106.04(a)(2)(III), claims fall outside the mental processes grouping when they "cannot practically be performed in the mind," and no step in the claimed pipeline can practically be so performed. Accordingly, the Examiner is respectfully requested to reconsider and withdraw the § 101 rejection of the claims using the "mental processes" rationale.
Examiner’s response: Applicant’s arguments have been fully considered but are not persuasive. Examiner disagrees that the claims are patentable under 101 because firstly, although claim 1 recites using "k-means clustering," "a naive Bayes classifier," "linear regression," and "a long short-term memory (LSTM) neural network, these are identified as additional elements in step 2A, prong 2 that merely apply the abstract idea of computing a numerical value on the computer using these techniques as explained in the previous Non-final rejection dated 26 January 2026. These limitations are not identified as abstract ideas in Step 2A, prong 1. Secondly, the limitations that are considered to be mental processes are only “computing a level of relevance for a target data point in the set of target data points to the target entity based on a comparison of text for the target data point and metadata of the target entity; and computing a classification label for the category of performance based on the set of target data” because under the broadest reasonable interpretation they can be performed in the mind or using a pen and paper. Thirdly,
Although Applicant argues that they are specific computational algorithms, each with defined mathematical and procedural operations, arranged in a defined sequence where the output of each step feeds the input of the next, they are merely applying these algorithms to the data. Lastly, the limitation “computing a linear regression over the set of target data on different points in time to generate a numerical performance value for each of the different points in time” recites mathematical relationships and calculations and therefore falls under “mathematical concepts” of subtract ideas. Furthermore, Applicant has failed to provide any explanation as to why the specific limitations that were identified as abstract in the rejection cannot be abstract. Accordingly, at Step 2A, prong two, after considering all claim elements individually and as an ordered combination, it is determined that the claims do not integrate the judicial exception into a practical application. Hence, the claims are abstract.
Regarding Applicant’s arguments, see pages 9-11, filed 05 March 2026, with respect to Claim Rejections under 35 USC § 101, Applicant argues that while claim 1 includes steps that involve mathematical operations, such as computing a linear regression and computing classification probabilities via a naive Bayes classifier, the claim does not recite these mathematical operations in the abstract or seek to preempt their general use. The linear regression is applied to specific data (target performance data collected from data sources corresponding to a category of performance) at specific points in time, for a specific purpose (generating numerical performance values), and the output feeds directly into a subsequent step (training an LSTM neural network). The naive Bayes classification operates on specific training data that was derived through a preceding k-means clustering step, and its classification output is modulated by relevance values computed through an NLP algorithm. Each mathematical operation is thus constrained to a defined role within a larger data processing pipeline, with specified inputs, outputs, and interdependencies between steps. Under Diamond v. Diehr, 450 U.S. 175 (1981), a claim does not become patent-ineligible merely because it includes a mathematical calculation; rather, the claim must be evaluated as a whole to determine whether it is directed to the mathematical concept itself or integrates the mathematics into a broader process.
The ordered combination of five distinct computational techniques:
" k-means clustering
" naive Bayes classification
" NLP-based relevance computation
" linear regression, and
" LSTM neural network training
constitutes a specific technical pipeline that is not conventional in the art. Each technique's output serves as the input or modulating factor for a subsequent technique: k-means clustering derives training data from reference data, naive Bayes classifiers are trained on that derived data, NLP computes relevance scores that affect how target data is used in classification, linear regression transforms temporal target data into numerical performance values, and those values train an LSTM to compute performance forecasts. This is not a claim to linear regression, Bayes' theorem, or any other mathematical concept in isolation. Instead, it is a claim to a defined sequence of operations where the specific arrangement and data flow between steps produces a result that no individual mathematical operation produces alone. The Section 103 rejection, which requires four references from four different technical domains to reconstruct this combination, itself demonstrates that the specific pipeline is not a conventional arrangement of mathematical tools.
Furthermore, the specification identifies a technical problem in paragraph [0010]:
Current prediction systems fail to accurately and efficiently predict the performance of an entity. The amount of data available for predicting performance is extremely large, thereby making the processing of such data computationally expensive. Furthermore, the large amount of data includes a significant amount of noisy data. As a result, these technical problems make it difficult for current prediction systems to accurately and efficiently predict the performance of an entity.
The claimed pipeline addresses that problem through specific computational means.
Paragraph [0013] discusses the advantages of using the claimed system: The k-means clustering step reduces the computational burden of processing large reference datasets before training. The NLP relevance computation filters or weights target data to reduce the effect of noisy or irrelevant data on classification. The linear regression step transforms raw temporal data into structured numerical values suitable for LSTM training, and the LSTM captures temporal dependencies to generate forecasts that static classification alone cannot produce. These are specific computational operations arranged to solve an identified data processing problem. Under Berkheimer v. HP Inc. (Fed. Cir., Feb. 8, 2018), whether a claimed combination is well-understood, routine, and conventional is a factual question requiring evidence, and the Office Action does not include evidence that this specific five-technique pipeline is conventional. The claim therefore recites significantly more than any individual mathematical concept it may incorporate. Accordingly, the Examiner is respectfully requested to reconsider and withdraw the § 101 rejection of the claims using the "mathematical concepts" rationale.
Examiner’s response: Applicant’s arguments have been fully considered but are not persuasive. Examiner disagrees that the claims are patentable under 101 because although the Applicant argues that the pipeline is specific computational operations arranged to solve an identified data processing problem the inventive concept appears to be in the data processing step that reduces the amount of data available in order to perform the abstract idea of computing performance forecasts. As explained in the previous Non-Final rejection dated 05 March 2026, in this case the improvement is in the abstract idea of computing performance forecasts and not in the computer itself. The computer is merely a tool used in producing the forecasts. As disclosed in MPEP 2106.05(a) it is important to note, the judicial exception alone cannot provide the improvement. And as discussed in MPEP 2106.05(f), mere instructions to implement an abstract idea on a computer as a tool to perform an abstract idea is not indicative of integration into a practical application. Therefore, as explained above and as shown in the detailed rejection below, the claim is not patent eligible. The techniques of k-means clustering, a naive Bayes classifier, linear regression, and a long short-term memory (LSTM) neural network are all well known and conventional and just by using them does not make the limitations “computing a level of relevance for a target data point in the set of target data points to the target entity based on a comparison of text for the target data point and metadata of the target entity; and computing a classification label for the category of performance based on the set of target data”, “computing a linear regression over the set of target data on different points in time to generate a numerical performance value for each of the different points in time” not abstract and even when all the claim elements were considered in combination, these additional elements represent mere instructions to apply an exception and insignificant extra-solution activity of data gathering etc, which cannot provide an inventive concept and do not amount to significantly more. Accordingly, at Step 2B, after considering all claim elements individually and as an ordered combination, it is determined that the claims do not amount to significantly more than the judicial exception.
Regarding Applicant’s arguments, see pages 11-13, filed 05 March 2026, with respect to Claim Rejections under 35 USC § 103, Applicant specifically argues on Page 12 that the Office Action alleges that these elements are taught by Qing on pp. 462, 465-466, and the Abstract.3 The Abstract of Qing states that the paper "proposes a novel solar prediction scheme" using a model "that is trained by using long short-term memory (LSTM) networks." The new model is compared to "persistence algorithm, linear least square regression and multilayered feedforward neural networks (BPNN)." The results show that the "proposed algorithm is %18.34 more accurate than BPNN." Thus, the Abstract of Qing shows that the LSTM network is an alternative to a linear regression, and does not teach or suggest "computing a linear regression ... to generate ... value[s]" and "using the generated ... values to train a ... LSTM ...," as recited by the claims.
Page 462 of Qing reiterates that the "novel prediction scheme" is based on "long short-term memory (LSTM) networks." Page 465 of Qing states that the authors "compared the prediction performance of the proposed LSTM networks algorithm with that of ... the linear least squares regression method (LR) ...." FIG. 8 on page 466 of Qing shows predictions generated by the LSTM and the LR, confirming that Qing is considering these prediction methods in the alternative, and does not teach or suggest "computing a linear regression ... to generate ... value[s]" and "using the generated ... values to train a ... LSTM ...," as recited by the claims. Accordingly, reconsideration and withdrawal of the § 103 rejection of independent claims 1, 10, and 19 is respectfully requested, along with a notice of allowance.
Each of dependent claims 2-4, 6-9, 11-13, 15-18, and 20-22 depends from one of
independent claims 1, 10, and 19, thus incorporating the elements of its respective independent claim. Furthermore, these dependent claims each contain additional patentable subject matter. Accordingly, the undersigned respectfully submits that dependent claims 2-4, 6-9, 11-13, 15-18, and 20-22 are allowable for at least the same reasons presented above in support of independent claims 1, 10, and 19.
Examiner’s response: Applicant’s arguments have been fully considered but are not persuasive.
Examiner disagrees that the cited references do not teach the claims because Reference Qing states on Page 465 (Column 2) in Table 2 showing LSTM forecasting architecture based on Keras the following steps: model.add(Dense(l ,activation = 'linear')). model.compile(loss = 'mse',optimizer = 'adam')
history = model.fit(train_x,train_t,epochs = 50,batch_size = 50). This clearly shows that Table 2 shows activation = linear first corresponding to computing a linear regression and then in the last step it shows train corresponding to using the previous steps to train in the last step of LSTM forecasting architecture. Hence, reference Qing teaches computing a linear regression over the set of target data on different points in time to generate a numerical performance value for each of the different points in time; and using the generated numerical performance values to train a long short-term memory (LSTM) neural network to compute performance forecasts.
Lastly, applicant's arguments with respect to the rejection of dependent claims 2-4, 6-9, 11-13, 15-18, and 20-22 under 35 USC § 103 have been fully considered but they are not persuasive because these claims depend from one of the independent claims 1, 10, or 19 and the combination of cited references teach every element of the amended claims as explained here and shown in detail below.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-4, 6-13, and 15-22 are rejected under 35 U.S.C. 101 because the claimed invention is directed towards abstract ideas without significantly more.
Regarding claims 1-4,6-9, 21, and 22:
According to the first step (Step 1) of the 101 analysis, claims 1-9 are directed to a computer-implemented method (process) and falls within one of the four statutory categories (i.e., process, machine, manufacture, or composition of matter).
Regarding claim 1:
In the next step (Step 2A, prong 1) of the analysis, the limitation of:
computing a level of relevance for a target data point in the set of target data points to the target entity based on a comparison of text for the target data point and metadata of the target entity;
and computing a classification label for the category of performance based on the set of target data.
Under the broadest reasonable interpretation, the above limitations are process steps that covers a mental processes including an observation, evaluation, judgment or opinion that could be performed in the mind or with the aid of pencil and paper but for the recitation of a generic computer component. If a claim, under its broadest reasonable interpretation, covers a mental process but for the recitation of generic computer components, then it falls within the “Mental Process” grouping of abstract ideas.
In the same step (Step 2A, prong 1) of the analysis, the limitation of:
computing a linear regression over the set of target data on different points in time to generate a numerical performance value for each of the different points in time;
Under the broadest reasonable interpretation, the above limitation is a process step that recites mathematical relationships and calculations but for the recitation of generic computer components. If a claim, under its broadest reasonable interpretation covers mathematical concepts but for the recitation of generic computer components, then it falls within the “Mathematical concepts” grouping of abstract ideas.
In the next step (Step 2A, prong 2) of the analysis, the limitations of:
A computer-implemented method performed by a computer system having a memory and at least one hardware processor, the computer-implemented method comprising:
training a naive Bayes classifier using the set of training data;
using a natural language processing algorithm;
using the trained naive Bayes classifier;
and using the generated numerical performance values to train a long short-term memory (LSTM) neural network to compute performance forecasts.
are considered to be additional elements and it does not integrate the abstract idea into a practical application because the additional elements are recited so generically (no details whatsoever are provided other than that it is a computer-implemented method performed by a computer system having a memory and at least one hardware processor, the computer-implemented method comprising: training a corresponding naive Bayes classifier using the set of training data; using a natural language processing algorithm; and using the trained naive Bayes classifier and using the generated numerical performance values to train a long short-term memory (LSTM) neural network to compute performance forecasts) that it represents no more than mere instructions to apply the judicial exception on a computer. As discussed in MPEP 2106.05(f), mere instructions to implement an abstract idea on a computer as a tool to perform an abstract idea is not indicative of integration into a practical application.
In the same step (Step 2A, prong 2) of the analysis, the limitations of:
accessing a set of reference data from a set of data sources corresponding to a category of performance,
deriving a set of training data using a k-means clustering algorithm and the set of reference data;
accessing a set of target data from the set of data sources corresponding to the category of performance.
are considered to be additional elements and as recited represent insignificant extra-solution activity because it is mere data gathering. See MPEP 2106.05(g), discussing limitations that the Federal Circuit has considered to be insignificant extra-solution activity.
In the same step (Step 2A, prong 2) of the analysis, the limitations of:
the set of reference data comprising a set of reference data points for each reference entity in a plurality of reference entities;
the set of target data comprising a set of target data points for a target entity;
are considered to be additional elements because it is simply selecting and describing a particular data source or type of data to be manipulated and the courts have found them to be insignificant extra-solution activities according to MPEP 2106.05(g).
In the last step (Step 2B) of the analysis, the additional elements do not amount to significantly more than the judicial exceptions. As explained with respect to Step 2A Prong Two, a computer-implemented method performed by a computer system having a memory and at least one hardware processor, the computer-implemented method comprising: training a corresponding naive Bayes classifier using the set of training data; using a natural language processing algorithm; and using the trained naive Bayes classifier, is at best the equivalent of merely adding the words “apply it” to the judicial exception. See MPEP 2106.05(f). Mere instructions to apply an exception cannot provide an inventive concept and does not amount to significantly more than the judicial exception.
In the same step (Step 2B) of the analysis, as discussed above the additional elements of:
accessing a set of reference data from a set of data sources corresponding to a category of performance,
deriving a set of training data using a k-means clustering algorithm and the set of reference data;
accessing a set of target data from the set of data sources corresponding to the category of performance;
which is recited at a high level of generality and amounts to extra-solution activity of receiving data i.e. pre-solution activity of gathering data for use in the claimed process. The courts have found limitations directed to obtaining information electronically, recited at a high level of generality, to be well-understood, routine, and conventional (see MPEP 2106.05(d)(II), “receiving or transmitting data over a network”, "electronic record keeping," and "storing and retrieving information in memory"). These limitations therefore remain insignificant extra-solution activity even upon reconsideration, and do not amount to significantly more.
In the same step (Step 2B), the limitation, “the set of reference data comprising a set of reference data points for each reference entity in a plurality of reference entities; the set of target data comprising a set of target data points for a target entity” amounts to insignificant extra solution activity of selecting a particular data source or type of data to be manipulated and is also considered to be well understood, routine, and conventional according to MPEP 2106.05(d)(II) Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); Presenting offers and gathering statistics, OIP Techs., 788 F.3d at 1362-63, 115 USPQ2d at 1092-93. These limitations therefore remain insignificant extra-solution activity even upon reconsideration, and do not amount to significantly more. Even when considered in combination, these additional elements represent mere instructions to apply an exception and insignificant extra-solution activity, which cannot provide an inventive concept. The claim is not patent eligible.
Regarding claim 2:
In the step (Step 2A, prong 1) of the analysis, the limitation of:
wherein the computing of the classification label for the category of performance is further based on the computed level of relevance.
Under the broadest reasonable interpretation, the above limitation is a process step that covers a mental processes including an observation, evaluation, judgment or opinion that could be performed in the mind or with the aid of pencil and paper but for the recitation of a generic computer component. If a claim, under its broadest reasonable interpretation, covers a mental process but for the recitation of generic computer components, then it falls within the “Mental Process” grouping of abstract ideas.
In the next step (Step 2A, prong 2) of the analysis, it does not integrate into a practical application because it does not add any additional elements that integrate the abstract idea into practical application.
In the last step (Step 2B) of the analysis, it does not add any additional elements that amount to significantly more than the abstract idea and thus fails to add an inventive concept. The claim is not patent eligible.
Regarding claim 3:
In the step (Step 2A, prong 1) of the analysis, the limitation of:
wherein the computing of the classification label for the category of performance comprises omitting the target data point from use in the computing of the classification label for the category of performance based on the computed level of relevance and a threshold relevance value.
Under the broadest reasonable interpretation, the above limitation is a process step that covers a mental processes including an observation, evaluation, judgment or opinion that could be performed in the mind or with the aid of pencil and paper but for the recitation of a generic computer component. If a claim, under its broadest reasonable interpretation, covers a mental process but for the recitation of generic computer components, then it falls within the “Mental Process” grouping of abstract ideas.
In the next step (Step 2A, prong 2) of the analysis, it does not integrate into a practical application because it does not add any additional elements that integrate the abstract idea into practical application.
In the last step (Step 2B) of the analysis, it does not add any additional elements that amount to significantly more than the abstract idea and thus fails to add an inventive concept. The claim is not patent eligible.
Regarding claim 4:
In the step (Step 2A, prong 1) of the analysis, the limitation of:
wherein the computing of the classification label for the category of performance comprises weighting the target data point in the computing of the classification label for the category of performance based on the computed level of relevance.
Under the broadest reasonable interpretation, the above limitation is a process step that covers a mental processes including an observation, evaluation, judgment or opinion that could be performed in the mind or with the aid of pencil and paper but for the recitation of a generic computer component. If a claim, under its broadest reasonable interpretation, covers a mental process but for the recitation of generic computer components, then it falls within the “Mental Process” grouping of abstract ideas.
In the next step (Step 2A, prong 2) of the analysis, it does not integrate into a practical application because it does not add any additional elements that integrate the abstract idea into practical application.
In the last step (Step 2B) of the analysis, it does not add any additional elements that amount to significantly more than the abstract idea and thus fails to add an inventive concept. The claim is not patent eligible.
Regarding claim 6:
In the next step (Step 2A, prong 2) of the analysis, the limitation of:
wherein the target data point comprises at least one of: a number of backlog items in an issue tracking system, data of a burndown chart, an amount of code stored in a version control system, a level of complexity of code in the version control system, a number or amount of rewards for contributions for a specific project or task, a current career stage, a speed of promotion to a next career stage, an amount of test coverage of code, a number of trainings completed, or a number of software components.
is considered to be an additional element and it does not integrate the abstract idea into a practical application because the additional element is recited so generically (no details whatsoever are provided other than that it is a method wherein the at least one target data point comprises at least one of: a number of backlog items in an issue tracking system, data of a burndown chart, an amount of code stored in a version control system, a level of complexity of code in the version control system, a number or amount of rewards for contributions for a specific project or task, a current career stage, a speed of promotion to a next career stage, an amount of test coverage of code, a number of trainings completed, or a number of software components) that it represents no more than mere instructions to apply the judicial exception on a computer. As discussed in MPEP 2106.05(f), mere instructions to implement an abstract idea on a computer as a tool to perform an abstract idea is not indicative of integration into a practical application.
In the last step (Step 2B) of the analysis, the additional element does not amount to significantly more than the judicial exceptions. As explained with respect to Step 2A Prong Two, the method wherein the target data point comprises at least one of: a number of backlog items in an issue tracking system, data of a burndown chart, an amount of code stored in a version control system, a level of complexity of code in the version control system, a number or amount of rewards for contributions for a specific project or task, a current career stage, a speed of promotion to a next career stage, an amount of test coverage of code, a number of trainings completed, or a number of software components, is at best the equivalent of merely adding the words “apply it” to the judicial exception. See MPEP 2106.05(f). Mere instructions to apply an exception cannot provide an inventive concept and does not amount to significantly more than the judicial exception. The claim is not patent eligible.
Regarding claim 7:
In the next step (Step 2A, prong 2) of the analysis, the limitation:
causing the classification label for the category of performance to be displayed on a computing device,
is considered to be an additional element and as recited represents insignificant extra-solution activity that is data output, because it is a mere nominal or tangential addition to the claim and is therefore not indicative of integration into a practical application. See MPEP 2106.05(g).
In the last step (Step 2B) of the analysis, the recitation of, causing the classification label for the category of performance to be displayed on a computing device, limitation amounts to insignificant extra solution activity because it is a mere nominal or tangential addition to the claim, amounting to mere data output (see MPEP 2106.05(g)). The courts have similarly found limitations directed to displaying a result, recited at a high level of generality, to be well-understood, routine, and conventional. See (MPEP 2106.05(d)(II), "presenting offers and gathering statistics.", “determining an estimated outcome and setting a price”). These limitations therefore remain insignificant extra-solution activity even upon reconsideration, and do not amount to significantly more. Even when considered in combination, these additional elements represent mere instructions to apply an exception and insignificant extra-solution activity, which cannot provide an inventive concept. The claim is not patent eligible.
Regarding claim 8:
In the step (Step 2A, prong 1) of the analysis, the limitation of:
Wherein: the computer-implemented method further comprises:
computing, a probability value for the classification label corresponding to the first category of performance;
computing, a second classification label for a second category of performance;
computing a single classification label indicating an overall performance of the target entity based on the probability values computed by the first trained naive Bayes classifier and the second trained naïve Bayes classifier;
Under the broadest reasonable interpretation, the above limitation is a process step that covers a mental processes including an observation, evaluation, judgment or opinion that could be performed in the mind or with the aid of pencil and paper but for the recitation of a generic computer component. If a claim, under its broadest reasonable interpretation, covers a mental process but for the recitation of generic computer components, then it falls within the “Mental Process” grouping of abstract ideas.
In the next step (Step 2A, prong 2) of the analysis, the limitations:
the trained naive Bayes classifier is a first trained naive Bayes classifier;
the category of performance is a first category of performance;
by the first trained naïve Bayes classifier,
by a second trained naïve Bayes classifier,
are considered to be additional elements and it does not integrate the abstract idea into a practical application because the additional elements are recited so generically (no details whatsoever are provided other than that the method comprises the trained naive Bayes classifier is a first trained naive Bayes classifier; the category of performance is a first category of performance; by the first trained naïve Bayes classifier, by a second trained naïve Bayes classifier) that it represents no more than mere instructions to apply the judicial exception on a computer. As discussed in MPEP 2106.05(f), mere instructions to implement an abstract idea on a computer as a tool to perform an abstract idea is not indicative of integration into a practical application.
In the same step (Step 2A, prong 2) of the analysis, the limitation:
and causing the single classification label to be displayed on a computing device.
is considered to be an additional element and as recited represents insignificant extra-solution activity that is data output, because it is a mere nominal or tangential addition to the claim and is therefore not indicative of integration into a practical application. See MPEP 2106.05(g).
In the last step (Step 2B) of the analysis, the additional elements do not amount to significantly more than the judicial exceptions. As explained with respect to Step 2A Prong Two, the method comprising the trained naive Bayes classifier is a first trained naive Bayes classifier; the category of performance is a first category of performance; by the first trained naïve Bayes classifier, by a second trained naïve Bayes classifier, is at best the equivalent of merely adding the words “apply it” to the judicial exception. See MPEP 2106.05(f). Mere instructions to apply an exception cannot provide an inventive concept and does not amount to significantly more than the judicial exception. The claim is not patent eligible.
In the same step (Step 2B) of the analysis, the recitation of, and causing the single classification label to be displayed on a computing device, limitation amounts to insignificant extra solution activity because it is a mere nominal or tangential addition to the claim, amounting to mere data output (see MPEP 2106.05(g)). The courts have similarly found limitations directed to displaying a result, recited at a high level of generality, to be well-understood, routine, and conventional. See (MPEP 2106.05(d)(II), "presenting offers and gathering statistics.", “determining an estimated outcome and setting a price”). These limitations therefore remain insignificant extra-solution activity even upon reconsideration, and do not amount to significantly more. Even when considered in combination, these additional elements represent mere instructions to apply an exception and insignificant extra-solution activity, which cannot provide an inventive concept. The claim is not patent eligible.
Regarding claim 9:
In the next step (Step 2A, prong 2) of the analysis, the limitation of:
wherein the target entity comprises a person employed by an organization, and the set of data sources for the category of performance comprises a database of the organization.
is considered to be an additional element and it does not integrate the abstract idea into a practical application because the additional element is recited so generically (no details whatsoever are provided other than that it is a method wherein the target entity comprises a person employed by an organization, and the set of data sources for the category of performance comprises a database of the organization) that it represents no more than mere instructions to apply the judicial exception on a computer. As discussed in MPEP 2106.05(f), mere instructions to implement an abstract idea on a computer as a tool to perform an abstract idea is not indicative of integration into a practical application.
In the last step (Step 2B) of the analysis, the additional element does not amount to significantly more than the judicial exceptions. As explained with respect to Step 2A Prong Two, the method wherein the target entity comprises a person employed by an organization, and the set of data sources for the category of performance comprises a database of the organization, is at best the equivalent of merely adding the words “apply it” to the judicial exception. See MPEP 2106.05(f). Mere instructions to apply an exception cannot provide an inventive concept and does not amount to significantly more than the judicial exception. The claim is not patent eligible.
Regarding claims 10-18:
According to the first step (Step 1) of the 101 analysis, claims 1-9 are directed to a system (manufacture) and falls within one of the four statutory categories (i.e., process, machine, manufacture, or composition of matter).
Regarding claims 10:
The claim 10 recites the substantially same limitations as claim1 except for the “A system comprising: at least one hardware processor; and a non-transitory computer-readable medium storing executable instructions that, when executed, cause the at least one hardware processor to perform operations comprising” limitation and is rejected for same rationale as shown in claim 1.
In the step (Step 2A, prong 2) of the analysis, the limitation of:
A system comprising: at least one hardware processor; and a non-transitory computer-readable medium storing executable instructions that, when executed, cause the at least one hardware processor to perform operations comprising:
is considered to be an additional element and it does not integrate the abstract idea into a practical application because the additional element is recited so generically (no details whatsoever are provided other than that it is a system comprising: at least one hardware processor; and a non-transitory computer-readable medium storing executable instructions that, when executed, cause the at least one hardware processor to perform operations comprising) that it represents no more than mere instructions to apply the judicial exception on a computer. As discussed in MPEP 2106.05(f), mere instructions to implement an abstract idea on a computer as a tool to perform an abstract idea is not indicative of integration into a practical application.
In the last step (Step 2B) of the analysis, the additional element does not amount to significantly more than the judicial exceptions. As explained with respect to Step 2A Prong Two, the system comprising: at least one hardware processor; and a non-transitory computer-readable medium storing executable instructions that, when executed, cause the at least one hardware processor to perform operations comprising, is at best the equivalent of merely adding the words “apply it” to the judicial exception. See MPEP 2106.05(f). Mere instructions to apply an exception cannot provide an inventive concept and does not amount to significantly more than the judicial exception. The claim is not patent eligible.
Regarding claim 11:
Claim 11 is substantially similar to claim 2 and therefore is rejected on similar grounds as claim 2.
Regarding claim 12:
Claim 12 is substantially similar to claim 3 and therefore is rejected on similar grounds as claim 3.
Regarding claim 13:
Claim 13 is substantially similar to claim 4 and therefore is rejected on similar grounds as claim 4.
Regarding claim 14:
Claim 14 is substantially similar to claim 5 and therefore is rejected on similar grounds as claim 5.
Regarding claim 15:
Claim 15 is substantially similar to claim 6 and therefore is rejected on similar grounds as claim 6.
Regarding claim 16:
Claim 16 is substantially similar to claim 7 and therefore is rejected on similar grounds as claim 7.
Regarding claim 17:
Claim 17 is substantially similar to claim 8 and therefore is rejected on similar grounds as claim 8.
Regarding claim 18:
Claim 18 is substantially similar to claim 9 and therefore is rejected on similar grounds as claim 9.
Regarding claims 19-20:
According to the first step (Step 1) of the 101 analysis, claims 1-9 are directed to a non-transitory machine-readable storage medium (manufacture) and falls within one of the four statutory categories (i.e., process, machine, manufacture, or composition of matter).
Regarding claim 19:
Claim 19 recites substantially same limitations as claim 1 except for “A non-transitory machine-readable storage medium tangibly embodying a set of instructions that, when executed by at least one hardware processor, causes the at least one hardware processor to perform operations comprising” and is rejected for same rationale as shown in claim 1.
In the step (Step 2A, prong 2) of the analysis, the limitation of:
A non-transitory machine-readable storage medium tangibly embodying a set of instructions that, when executed by at least one hardware processor, causes the at least one hardware processor to perform operations comprising:
is considered to be an additional element and it does not integrate the abstract idea into a practical application because the additional element is recited so generically (no details whatsoever are provided other than that it is a system comprising: at least one hardware processor; and a non-transitory computer-readable medium storing executable instructions that, when executed, cause the at least one hardware processor to perform operations comprising) that it represents no more than mere instructions to apply the judicial exception on a computer. As discussed in MPEP 2106.05(f), mere instructions to implement an abstract idea on a computer as a tool to perform an abstract idea is not indicative of integration into a practical application.
In the last step (Step 2B) of the analysis, the additional element does not amount to significantly more than the judicial exceptions. As explained with respect to Step 2A Prong Two, the system comprising: at least one hardware processor; and a non-transitory computer-readable medium storing executable instructions that, when executed, cause the at least one hardware processor to perform operations comprising, is at best the equivalent of merely adding the words “apply it” to the judicial exception. See MPEP 2106.05(f). Mere instructions to apply an exception cannot provide an inventive concept and does not amount to significantly more than the judicial exception. The claim is not patent eligible.
Regarding claim 20:
Claim 20 is substantially similar to claim 2 and therefore is rejected on similar grounds as claim 2.
Regarding claim 21:
In step (Step 2A, prong 2) of the analysis, the limitation:
wherein the text for the target data point comprises a transcript of a completed training.
is considered to be an additional element and it does not integrate the abstract idea into a practical application because the additional element is recited so generically (no details whatsoever are provided other than that it is a method wherein the text for the target data point comprises a transcript of a completed training) that it represents no more than mere instructions to apply the judicial exception on a computer. As discussed in MPEP 2106.05(f), mere instructions to implement an abstract idea on a computer as a tool to perform an abstract idea is not indicative of integration into a practical application.
In the last step (Step 2B) of the analysis, the additional element does not amount to significantly more than the judicial exceptions. As explained with respect to Step 2A Prong Two, a method wherein the text for the target data point comprises a transcript of a completed training, is at best the equivalent of merely adding the words “apply it” to the judicial exception. See MPEP 2106.05(f). Mere instructions to apply an exception cannot provide an inventive concept and does not amount to significantly more than the judicial exception. The claim is not patent eligible.
Regarding claim 22:
In step (Step 2A, prong 2) of the analysis, the limitation:
wherein the metadata of the target entity comprises a job title.
is considered to be an additional element and it does not integrate the abstract idea into a practical application because the additional element is recited so generically (no details whatsoever are provided other than that it is a method wherein the metadata of the target entity comprises a job title) that it represents no more than mere instructions to apply the judicial exception on a computer. As discussed in MPEP 2106.05(f), mere instructions to implement an abstract idea on a computer as a tool to perform an abstract idea is not indicative of integration into a practical application.
In the last step (Step 2B) of the analysis, the additional element does not amount to significantly more than the judicial exceptions. As explained with respect to Step 2A Prong Two, a method wherein the metadata of the target entity comprises a job title, is at best the equivalent of merely adding the words “apply it” to the judicial exception. See MPEP 2106.05(f). Mere instructions to apply an exception cannot provide an inventive concept and does not amount to significantly more than the judicial exception. The claim is not patent eligible.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
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.
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 1-4, 6, 7, 9-13, 15, 16, 18-20, and 22 are rejected under 35 U.S.C. 103 as being unpatentable over Prathan et al (Determining the best-fit programmers using Bayes' theorem and artificial neural network, 2020) in view of Fadhil (Hybrid of K-means clustering and naive Bayes classifier for predicting performance of an employee, 2021) and Doolin et al (WO 2016142531 A1) and further in view of Qing et al (Hourly day-ahead solar irradiance prediction using weather forecasts by LSTM, 2017).
Regarding claim 1:
Prathan teaches: A computer-implemented method performed by a computer system having a memory and at least one hardware processor, the computer-implemented method comprising ([Page 434, Column 1, Paragraph 2] Our proposed technique to predict and determine the best-fit programmers for software companies is divided into two phases - Bayes' theorem and artificial neural network (ANN) phases. Note: ANN shows that it is being run on a computer system having a memory and at least one hardware processor):
accessing a set of reference data from a set of data sources corresponding to a category of performance, the set of reference data comprising a set of reference data points for each reference entity in a plurality of reference entities ([Page 433, Column 1, Nomenclature] Parameters and variables - chance a programmer will achieve performance of class i, given that it has attribute value A. Attributes of Company 1 and 2. [Page 438, Column 2, Section 5] The datasets used in this study are the past annual performance appraisal of programmers working in the two Indian software companies (Company I and Company 2) consisting of 470 samples from Company I and 471 samples from Company 2, covering the years 2010-2015. These two datasets were split into two sets of data - 85% used as the training set, and 15% used as the test set. [Page 437] Tables 2 and 3. Note: The list of Attributes of Company 1 and Company 2 recited correspond to a plurality of categories of performance. Company 1 and 2 are the data sources. Programmer correspond to a reference entity);
training a naive Bayes classifier using the set of training data ([Page 434, Column 1, Paragraph 1] In this study, we have incorporated the Bayes' theorem and ANN to study the most important attributes (prognostic attributes) pertaining to the work performance of programmers using the past annual performance appraisal of the programmers and classify the performance of the programmers based on the selected prognostic attributes to predict and determine the best-fit programmers. [Page 435, Column 2, Last Paragraph] The second component involves the development of a technique to predict and determine the best-fit programmers, as shown in Fig. l.b. The technique incorporates the use of Bayes' theorem and ANN. Bayes' theorem is applied on the data stored in the warehouse. ANN with Levenberg-Marquardt (LM) training algorithm and the multilayer perceptron (MLP) neural network. [Page 438, Column 2, Section 5] The datasets used in this study are the past annual performance appraisal of programmers working in the two Indian software companies (Company I and Company 2) consisting of 470 samples from Company I and 471 samples from Company 2, covering the years 2010-2015. These two datasets were split into two sets of data - 85% used as the training set, and 15% used as the test set);
accessing a set of target data from the set of data sources corresponding to the category of performance, the set of target data comprising a set of target data points for a target entity ([Page 434, Column 1, Paragraph 1] This study is aimed at identifying the relevant attributes that programmers should possess as well as proposing a selection technique to determine the best-fit programmer for a software company. To identify the attributes, past annual performance appraisal records of programmers can be analysed. Note: The best-fit programmer corresponds to target entity and attributes correspond to the set of target data of the category of performance);
and computing a classification label for the category of performance based on the set of target data using the trained naive Bayes classifier ([Page 437] Tables 2 and 3. Note: The list of Attributes of Company 1 and Company 2 in column 1 correspond to the category of performance. Good, average, poor correspond to a classification label).
However, Prathan does not explicitly disclose: deriving a set of training data using a k-means clustering algorithm and the set of reference data; computing a level of relevance for a target data point in the set of target data points to the target entity using a natural language processing algorithm based on a comparison of text for the target data point and metadata of the target entity; computing a linear regression over the set of target data on different points in time to generate a numerical performance value for each of the different points in time; and using the generated numerical performance values to train a long short-term memory (LSTM) neural network to compute performance forecasts.
Fadhil teaches, in an analogous system: deriving a set of training data using a k-means clustering algorithm and the set of reference data ([Page 803, Paragraph 1] This data comes from previous works that are mentioned in literature reviews. The clustering and classification models have been established following the data preparation. To assess the efficacy of previous research, a group of attributes has been chosen. The attributes consist of personal information, Education information, and Professional information. These data were used to predict employee performance. In the first phase, the clustering process using the K-Means algorithm is performed first to determine training data).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the computer-implemented method of Prathan to incorporate the teachings of Fadhil to use deriving a set of training data using a k-means clustering algorithm and the set of reference data. One would have been motivated to do this modification because doing so would give the benefit of enhancing the accuracy of the classification model as taught by Fadhil [Page 802, Last Paragraph].
Doolin teaches, in an analogous system: computing a level of relevance for a target data point in the set of target data points to the target entity using a natural language processing algorithm based on a comparison of text for the target data point and metadata of the target entity ([Page 11, Paragraph 3] As shown in Figure 2, determining the relevance score for a profile comprises extracting a value for one or more features from the profile 23. The features for which values are extracted may include, for example, job title, company name, job description, skills, academic background, bio and interests. A profile 23 may be received by the relevance engine 8 in JavaScript Object Notation (JSON) format, where the keys are field names and the feature values are the corresponding data. The data may be simple text, an array of text, or another JSON object. For example, "first name", "last name", "skills", and "bio" are the names of the fields in a profile and the values for "first name", "last name", and "bio" are plain text, with an array of text items for "skills". [Page 11, Paragraph 4] As shown in Figure 2, the relevance engine 8 normalises each feature value to generate a normalized feature value. Normalizing the feature values comprises performing one or more natural language processing (NLP) operations 24 on the feature values, to parse the feature values into one or more tokens. [Page 12, Paragraph 4] As shown in Figure 3, the relevance engine 8 then compares one or more normalized feature values of the normalized profile with values of equivalent features of the user's profile based on user-defmeable settings in a configuration settings file 29 to determine the relevance score 35 of each member profile to the user. [Page 13, Paragraph 1] Similarly, the job title ontology defines relationships between various job titles or profession names. For example, "junior software engineer" and "graduate software engineer" are part of a "software developer and programmer" profession, Note: Job title corresponds to the metadata of the target entity, where target entity corresponds to the programmer. Features like skills, academic background, bio and interests are all attributes that correspond to the set of target data points).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the combined teachings of Prathan and Fadhill to incorporate the teachings of Doolin to compute a level of relevance for a target data point in the set of target data points to the target entity using a natural language processing algorithm based on a comparison of text for the target data point and metadata of the target entity. One would have been motivated to do this modification because doing so would give the benefit of normalizing each feature value to generate a normalized feature value as taught by Doolin [Page 11, Paragraph 4].
Qing teaches, in an analogous system: computing a linear regression over the set of target data on different points in time to generate a numerical performance value for each of the different points in time ([Page 463, Column 2, Last Paragraph] Our goal is to predict the hourly irradiance value Yt using the given feature vector Xt. Classical methods such as linear regression and feedback neural network, are to learn a predictive model which performs an explicit or implicit function relationship between the individual feature vector and the individual output value at a specific hour, which neglect the dependence between consecutive hours of the same day. In this paper, we formulate prediction of the hourly day-ahead irradiance values of the same day as a structured output prediction, i.e, given an input sequence consisting of the hourly feature vectors as Xd = {x8 ,x9 , . . . ,x18 }, produces a output [Page 464, Column 1, Paragraph 1] sequence consisting of the hourly irradiance values at a specific day as Yd= {Ys,Yg, ---, Y18}. [Page 465, Column 2, Table 2: LSTM forecasting architecture based on Keras] model.add(Dense(l ,activation = 'linear')). Note: Table 2 shows activation = linear corresponding to computing a linear regression);
and using the generated numerical performance values to train a long short-term memory (LSTM) neural network to compute performance forecasts ([Abstract] The proposed prediction model is trained by using long short-term memory (LSTM) networks taking into account the dependence between consecutive hours of the same day. We compare persistence algorithm, linear least square regression and multilayered feedforward neural networks using backpropagation algorithm (BPNN) for solar irradiance prediction. [Page 462, Column 2, Paragraph 2] we propose a novel prediction scheme for hourly one-day ahead prediction of solar irradiance based on weather forecasts and long short-term memory (LSTM) networks. [Page 465, Column 2, Table 2: LSTM forecasting architecture based on Keras.] model.add(Dense(l ,activation = 'linear')). model.compile(loss = 'mse',optimizer = 'adam') history = model.fit(train_x,train_t,epochs = 50,batch_size = 50) Note: Table 2 shows activation = linear first corresponding to computing a linear regression and then in the last step it shows train corresponding to using the previous steps to train in the last step of LSTM forecasting architecture).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the combined teachings of Prathan, Fadhill, and Doolin to incorporate the teachings of Qing to compute a linear regression over the set of target data on different points in time to generate a numerical performance value for each of the different points in time; and use the generated numerical performance values to train a long short-term memory (LSTM) neural network to compute performance forecasts. One would have been motivated to do this modification because doing so would give the benefit of performing an explicit or implicit function relationship between the individual feature vector and the individual output value at a specific hour as taught by Qing [Page 465, Column 2, Last Paragraph].
Regarding claim 2:
The system of Prathan, Fadhil, Doolin, and Qing teaches: The computer-implemented method of claim 1 (as shown above).
Prathan further teaches: wherein the computing of the classification label for the category of performance is further based on the computed level of relevance ([Page 435, Column 2, Section 3.2] , Paragraph 1] The first component is data acquisition and processing, as shown in Fig. I .a, involves the extraction of the relevant subsets of data, which include information on the programmers' annual performance appraisal from the two companies. [Page 436, Column 2, Section 4.1 Step 1] - analyse the most relevant attributes using Bayes' theorem. Note: section 4.1 shows computing probabilities and also in Tables 1 and 2 which corresponds to computed level of relevance. Good, poor, and average in Tables 1 and 2 correspond to the classification labels).
Regarding claim 3:
The system of Prathan, Fadhil, Doolin, and Qing teaches: The computer-implemented method of claim 2 (as shown above).
Prathan further teaches: wherein the computing of the classification label for the category of performance comprises omitting the target data point from use in the computing of the classification label for the category of performance based on the computed level of relevance and a threshold relevance value ([Page 437, Column 1, Last Paragraph] The most relevant attributes (prognostic attributes) are thus the extracted values, which are [Page 437, Column 2, 2nd Last Paragraph] higher than 0.5. Based on the values shown in Tables 2 and 3, the high potential attributes (prognostic attributes) of Company I and Company 2 are listed in Tables 4 and 5, respectively. [Page 437, Column 2, Last Paragraph] Table 4 shows the values obtained after applying Bayes' theorem (Table 2), where seven out of the 10 attributes are prognostic attributes. The high potential attributes (prognostic attributes) indicate the probability of a programmer being rated 'good' for each attribute and also being rated 'good' in their work. [Page 438, Column 2, Paragraph 1] Similarly, the probabilities for WAPC (0.640), PSUT (0.590), DCW (0.738), AQWN (0.738), CS (0.705), and CR (0.525), are all higher than 0.5 and hence, they are among the list of prognostic attributes. On the other hand, the probabilities for the three attributes, P (0.230), PI (0.443), and L (0.361) are <0.5, implying that these attributes do not have any impact on a programmer's performance. Note: 0.5 corresponds to the threshold value. The three that are less than 0.5 that do not have any impact on a programmer's performance correspond to omitting the at least on target data point from use).
Regarding claim 4:
The system of Prathan, Fadhil, Doolin, and Qing teaches: The computer-implemented method of claim 2 (as shown above).
Prathan further teaches: wherein the computing of the classification label for the category of performance comprises weighting the target data point in the computing of the classification label for the category of performance based on the computed level of relevance ([Page 434, Column 1, Paragraph 2] This study is aimed at identifying the relevant attributes that programmers should possess as well as proposing a selection technique to determine the best-fit programmer for a software company. To identify the attributes, past annual performance appraisal records of programmers can be analysed. The attributes will then be ranked according to the most important attributes or prognostic attributes (high potential attributes) pertaining to the work performance of programmers. The work performances of programmers have been classified into good, average, and poor performance depending on the software company evaluation approach. Our proposed technique to predict and determine the best-fit programmers for software companies is divided into two phases - Bayes' theorem and artificial neural network (ANN) phases. Bayes' theorem is used to identify the prognostic attributes, while ANN is applied to refine the weightage of each attribute for the purpose of shortlisting candidates who are qualified to be appointed to be programmers).
Regarding claim 6:
The system of Prathan, Fadhil, Doolin, and Qing teaches: The computer-implemented method of claim 1 (as shown above).
Prathan further teaches: wherein the target data point comprises at least one of: a number of backlog items in an issue tracking system, data of a burndown chart, an amount of code stored in a version control system, a level of complexity of code in the version control system, a number or amount of rewards for contributions for a specific project or task, a current career stage, a speed of promotion to a next career stage, an amount of test coverage of code, a number of trainings completed, or a number of software components ([Page 433, Column 1, Attributes of Company 1] PSUT perform satisfactory unit test. Note: This corresponds to an amount of test coverage of code because unit testing is a software development process where individual components of an application, such as functions or methods, are tested in isolation to verify that they function as expected. It's a crucial step in ensuring code quality and reliability by identifying bugs early in the development cycle).
Regarding claim 7:
The system of Prathan, Fadhil, Doolin, and Qing teaches: The computer-implemented method of claim 1 (as shown above).
Prathan further teaches: further comprising: causing the classification label for the category of performance to be displayed on a computing device ([Page 437, Column 2] [Page 438] Figure 5. Note: Tables 2 and 3 are being displayed with the attribute values corresponding to classification labels. Figure 5 also shows values being displayed on a computing device).
Regarding claim 9:
The system of Prathan, Fadhil, Doolin, and Qing teaches: The computer-implemented method of claim 1 (as shown above).
Prathan further teaches: wherein the target entity comprises a person employed by an organization, and the set of data sources for the category of performance comprise a database of the organization ([Page 434, Column 1, Paragraph 1] This study is aimed at identifying the relevant attributes that programmers should possess as well as proposing a selection technique to determine the best-fit programmer for a software company. To identify the attributes, past annual performance appraisal records of programmers can be analysed. Note: Programmer corresponds to target entity and software company corresponds to an organization. [Page 437] Tables 2 and 3. Note: The list of Attributes of Company 1 and Company 2 recited correspond to a plurality of categories of performance. Company 1 and 2 correspond to one or more databases of the organization. Software company corresponds to organization).
Regarding claim 10:
Claim 10 is substantially similar to claim 1 and therefore is rejected on similar grounds as claim 1.
Regarding claim 11:
Claim 11 is substantially similar to claim 2 and therefore is rejected on similar grounds as claim 2.
Regarding claim 12:
Claim 12 is substantially similar to claim 3 and therefore is rejected on similar grounds as claim 3.
Regarding claim 13:
Claim 13 is substantially similar to claim 4 and therefore is rejected on similar grounds as claim 4.
Regarding claim 15:
Claim 15 is substantially similar to claim 6 and therefore is rejected on similar grounds as claim 6.
Regarding claim 16:
Claim 16 is substantially similar to claim 7 and therefore is rejected on similar grounds as claim 7.
Regarding claim 18:
Claim 18 is substantially similar to claim 9 and therefore is rejected on similar grounds as claim 9.
Regarding claim 19:
Claim 19 is substantially similar to claim 1 and therefore is rejected on similar grounds as claim 1.
Regarding claim 20:
Claim 20 is substantially similar to claim 2 and therefore is rejected on similar grounds as claim 2.
Regarding claim 22:
The system of Prathan, Fadhil, Doolin, and Qing teaches: The computer-implemented method of claim 1 (as shown above).
However, the system of Prathan and Fadhil does not explicitly disclose: wherein the metadata of the target entity comprises a job title.
Doolin teaches, in an analogous system: wherein the metadata of the target entity comprises a job title ([Page 13, Paragraph 1] Similarly, the job title ontology defines relationships between various job titles or profession names. For example, "junior software engineer" and "graduate software engineer" are part of a "software developer and programmer" profession).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the combined teachings of Prathan and Fadhill to incorporate the teachings of Doolin wherein the metadata of the target entity comprises a job title. One would have been motivated to do this modification because doing so would give the benefit of normalizing each feature value to generate a normalized feature value as taught by Doolin [Page 11, Paragraph 4].
Claims 8 and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Prathan et al (Determining the best-fit programmers using Bayes' theorem and artificial neural network, 2020), Fadhil (Hybrid of K-means clustering and naive Bayes classifier for predicting performance of an employee, 2021), Doolin et al (WO 2016142531 A1) and Qing et al (Hourly day-ahead solar irradiance prediction using weather forecasts by LSTM, 2017) and further in view of Kamishima et al (The Independence of Fairness-aware Classifiers, 2013).
Regarding claim 8:
The system of Prathan, Fadhil, Doolin, and Qing teaches: The computer-implemented method of claim 1 (as shown above).
Prathan further teaches: and causing the single classification label to be displayed on a computing device ([Page 438, Column 1] [Page 438] Figure 5. Note: Table 4 is being displayed. Figure 5 also shows values being displayed on a computing device).
However, the system of Prathan, Fadhil, Doolin, and Qing does not explicitly disclose: wherein: the trained naive Bayes classifier is a first trained naive Bayes classifier; the category of performance is a first category of performance; and the computer-implemented method further comprises: computing, by the first trained naive Bayes classifier, a probability value for the classification label corresponding to the first category of performance; computing, by a second trained naive Bayes classifier, a second classification label for a second category of performance; computing a single classification label indicating an overall performance of the target entity based on the probability values computed by the first trained naive Bayes classifier and the second trained naive Bayes classifier.
Kamishima teaches, in an analogous system: wherein: the trained naive Bayes classifier is a first trained naive Bayes classifier; the category of performance is a first category of performance; and the computer-implemented method further comprises: computing, by the first trained naive Bayes classifier, a probability value for the classification label corresponding to the first category of performance; computing, by a second trained naive Bayes classifier, a second classification label for a second category of performance; computing a single classification label indicating an overall performance of the target entity based on the probability values computed by the first trained naive Bayes classifier and the second trained naive Bayes classifier ([Page 850, Column 1, Paragraph 1] Though class labels are in fact chosen according to a deterministic Bayes decision rule, our simple version assumes that labels are probabilistically decided, [Page 852, Column 2, Paragraph 1] this method was named “two-naive-Bayes” because it is as if two naive Bayes classifiers are learned depending on each sensitive value. [Page 855, Column 1, Last Paragraph] The first baseline was a naive Bayes classifier that used both non-sensitive and sensitive features, denoted as NB. The second baseline was a naive Bayes classifier that used only non-sensitive features, denoted as NBns. [Page 855, Column 2, Last Paragraph] The second reason is the deterministic Bayes decision rule. Though class labels are in fact chosen according to a deterministic decision rule, our simple version assumes that labels are probabilistically decided. [Page 867, Column 1, Paragraph 6] Pˆr†[Y ∗=1|X=x, S=s] is the probability that an actual label becomes 1 given a specific data (x, s). This probability can be 0 or 1 because labels are deterministically assigned by the decision rule, and becomes 1 if the following condition is satisfied).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the combined teachings of Prathan, Fadhil, Doolin, and Qing to incorporate the teachings of Kamishima wherein: the trained naive Bayes classifier is a first trained naive Bayes classifier; the category of performance is a first category of performance; and the computer-implemented method further comprises: computing, by the first trained naive Bayes classifier, a probability value for the classification label corresponding to the first category of performance; computing, by a second trained naive Bayes classifier, a second classification label for a second category of performance; computing a single classification label indicating an overall performance of the target entity based on the probability values computed by the first trained naive Bayes classifier and the second trained naive Bayes classifier. One would have been motivated to do this modification because doing so would give the benefit of revealing the theoretical background of the two-naive-Bayes method and its connections with other methods as taught by Kamishima [Abstract].
Regarding claim 17:
Claim 17 is substantially similar to claim 8 and therefore is rejected on similar grounds as claim 8.
Claim 21 is rejected under 35 U.S.C. 103 as being unpatentable over Prathan et al (Determining the best-fit programmers using Bayes' theorem and artificial neural network, 2020), Fadhil (Hybrid of K-means clustering and naive Bayes classifier for predicting performance of an employee, 2021), Doolin et al (WO 2016142531 A1), and Qing et al (Hourly day-ahead solar irradiance prediction using weather forecasts by LSTM, 2017) and further in view of Payne et al (US 20120190002 A1).
Regarding claim 21:
The system of Prathan, Fadhil, Doolin, and Qing teaches: The computer-implemented method of claim 1 (as shown above).
However, the system of Prathan, Fadhil, Doolin, and Qing does not explicitly disclose: wherein the text for the target data point comprises a transcript of a completed training.
Payne teaches, in an analogous system: wherein the text for the target data point comprises a transcript of a completed training ([0007, Last 5 lines] a status report of the e-learning plan indicating a status of each selected e-training course of the e-learning plan, a transcript of all completed e-training courses from the e-learning plan, and a certificate of completion for at least one e-training course from the e-learning plan).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the combined teachings of Prathan, Fadhil, Doolin, and Qing to incorporate the teachings of Payne wherein the text for the target data point comprises a transcript of a completed training. One would have been motivated to do this modification because doing so would give the benefit of having a status report of the e-learning plan indicating a status of each selected e-training course of the e-learning plan as taught by Payne [0007].
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Purdy et al (US11138524B2) discloses Cascaded, boosted predictive models trained using distinct sets of exogenous and endogenous features are configured to predict component of performance ratings of entities. From the distinct predicted components, the second entity's rating factor can be determined. A second entity's rating factor represents the specific contribution a second entity makes to his average performance rating, as distinct from the rating that an arbitrary or hypothetical second entity would obtain.
Pfitzmann et al (US 11481682 B2) discloses training the model on a sequence of increasing-sized sets of the training samples and testing performance of the model after training with each set to obtain class-specific performance metrics corresponding to each set size. Embodiments of the present invention may include generating class-specific learning curves from the performance metrics for the plurality of classes. Embodiments of the present invention may include extrapolating the learning curves. Embodiments of the present invention may include optimizing a function of the predicted performance metrics to identify a set of augmentation actions to augment the dataset for further training of the model. Embodiments of the present invention may include providing an output indicative of the set of augmentation actions.
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 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 CHAITANYA RAMESH JAYAKUMAR whose telephone number is (571)272-3369. The examiner can normally be reached Mon-Fri 9am-1pm.
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, Omar Fernandez Rivas can be reached at (571)272-2589. 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.
/C.R.J./Examiner, Art Unit 2128
/OMAR F FERNANDEZ RIVAS/Supervisory Patent Examiner, Art Unit 2128