DETAILED ACTION
Notice of 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 .
Status of the Application
Claims 1-20 were pending and were rejected in the previous office action. No claims were amended, added, or canceled in the 1/27/2026 response. Claims 1-20 remain pending and are examined in this office action.
Response to Arguments
35 USC § 101:
Applicant’s arguments regarding the § 101 rejection of claims 1-20 (pgs. 14-18, remarks filed 1/27/2026) have been fully considered, but they are not persuasive.
Applicant first argues that the claims do not recite a mental process because the claims cannot be practically performed in the human mind and require “using a machine learning model” (pgs. 14-16, remarks).
However, the examiner respectfully disagrees. First, in response to the argument that the examiner’s Step 2A Prong One analysis has oversimplified the claims, the examiner reminds applicant that “An abstract idea can generally be described at different levels of abstraction.” Apple, Inc. v. Ameranth, Inc., 842 F.3d 1229, 1240-41 (Fed. Cir. 2016), and the test at Step 2A Prong One is whether or not the claims recite limitations that fall under one of the enumerated groupings of judicial exceptions (i.e. abstract ideas). While the limitation “using a machine learning model” is an additional element to be consider at Step 2A Prong Two/Step 2B and is not part of the abstract idea, simply reciting an additional element to “apply it” does not change that the claim still including limitations that can be practically performed via the human mind (with or without physical aid such as pen and paper) and recite a mental process at Step 2A Prong One. See MPEP 2106.04(a)(2)(III)(C) showing “Claims can recite a mental process even if they are claimed as being performed on a computer.” As seen in the previous office action (pgs. 3-4, 11/5/2025 non-final rejection), independent claim 1 (and similar claims 11 and 20) recite limitations for “issuing…a query,” “receiving…a response to the query, the response including first natural language data comprising data associated with a plurality of users and second natural language data comprising textual data,” “augmenting…the first natural language data…to generate a set of augmented profiles for the plurality of users,” “augmenting…the second natural language data…to generate a set of augmented textual data,” and “generating…a difference between the set of augmented profiles and the set of augmented textual data” that were identified as reciting an abstract idea (the abstract idea being characterized as “retrieving first natural language data associated with users and second natural language data including textual data, augmenting the first natural language to generate user profiles, and augmenting second natural language data to generate augmented textual data, and generating a difference between the first and second augmented sets of data”).
Furthermore, as described in MPEP 2106.04(a)(2)(III), “[T]he "mental processes" abstract idea grouping is defined as concepts performed in the human mind, and examples of mental processes include observations, evaluations, judgments, and opinions.” and “If a claim recites a limitation that can practically be performed in the human mind, with or without the use of a physical aid such as pen and paper, the limitation falls within the mental processes grouping, and the claim recites an abstract idea.” The limitations recited by the independent claims 1, 11 and 20 above cover concepts (e.g. observation, evaluation, judgment, and opinion) that can reasonably be performed in the human mind or by the human mind with the aid of simple tools such as pen and paper. For example, the paired steps for “issuing a query…” and then “receiving…a response to the query…” step amounts to observation, while the “augmenting…to generate a set of augmented profiles for the plurality of users…,” “augmenting…to generate a set of augmented textual data…,” and “generating a difference between the set of augmented profiles and the set of augmented textual data…” steps would be considered evaluations, judgments, and opinions that could be carried out either with the human mind alone or with the aid of simple tools such as pen and paper. Therefore, as the processes above described by the representative independent claims 1, 11, and 20 can be characterized as mental processes (i.e. observation, evaluation, judgment, and opinion), but for the recitation of generic computer components in the claims, the claims fall under the “mental processes” category of judicial exceptions (i.e. abstract ideas).
Applicant further argues, at Step 2A Prong Two, that the claims integrate any recited abstract idea into a practical application in view of Ex Parte Desjardins, and also discussing Recentive Analytics (pgs. 16-18, remarks).
The examiner respectfully disagrees. Applicant’s comparisons to Ex Parte Desjardins are not applicable to the instant application, as the claims do not recite an improvement specific to machine learning models themselves (e.g. improvement the performance of a machine learning model while preventing catastrophic forgetting), but instead simply recite “using” a set of three machine learning models recited at a high level of generality without detailing any of the technical mechanisms of how these machine learning models operate (e.g. they are recited as a “black box” that simply receives inputs and generates outputs). Applicant has not identified specific limitations or inventive concepts in the claims that recite a technological solution to a technological problem – and the alleged solution for “extracting meaningful semantic insights…” is not a technological improvement. Extracting meaningful insights simply describes data analysis, which is a fundamentally human task (e.g. a mental process) that is not specific to computers or to a specific field of technology. The task being difficult to accomplish manually due to the volume of data is not a technological problem, and nonetheless, this alleged problem is simply overcome by use of generic computer implementation. See Bancorp Servs., L.L.C. v. Sun Life Assurance Co. of Can. (U.S.), 687 F.3d 1266, 1278 (Fed. Cir. 2012) ("[T]he fact that the required calculations could be performed more efficiently via a computer does not materially alter the patent eligibility of the claimed subject matter."). Similarly, “[C]laiming the improved speed or efficiency inherent with applying the abstract idea on a computer” does not integrate a judicial exception into a practical application or provide an inventive concept. Intellectual Ventures I LLC v. Capital One Bank (USA), 792 F.3d 1363, 1367, 115 USPQ2d 1636, 1639 (Fed. Cir. 2015). Also see Recentive Analytics, Inc. v. Fox. Corp. showing "Finally, the claimed methods are not rendered patent eligible by the fact that (using existing machine learning technology) they perform a task previously undertaken by humans with greater speed and efficiency than could previously be achieved.”
Furthermore, the claims do not recite some “specific ordered pipeline” with specialized machine learning models such that the claims recite an inventive concept – they instead recite an abstract idea (i.e. retrieving first natural language data associated with users and second natural language data including textual data, augmenting the first natural language to generate user profiles, and augmenting second natural language data to generate augmented textual data, and generating a difference between the first and second augmented sets of data) being carried out on generic computers and machine learning models. The claims do not recite a single technical detail about how the first, second, or third machine learning models operate beyond “using a first machine learning model,” “using a second machine learning model,” and “using a third machine learning model,” which provides nothing more than mere instructions to “apply it” generic computer elements. See Recentive Analytics, Inc. v. Fox Corp., 134 F.4th 1205 (Fed. Cir. 2025), showing “[P]atents that do no more than claim the application of generic machine learning to new data environments, without disclosing improvements to the machine learning models to be applied, are patent ineligible under § 101.”
Therefore, the previous § 101 rejection of claims 1-20 is maintained.
35 USC § 103:
Applicant’s arguments with respect to the current § 103 rejections of claims 1-20 (pgs. 11-14, remarks filed 1/27/2026) have been fully considered, but they are not persuasive.
Applicant argues that Guru (US 20220092514 A1) does not teach the recited “augmenting” steps recited in claims 1, 11 and 20 (pgs. 12-13, remarks), and that Miguel does not cure the deficiencies of Guru (pgs. 12-13, remarks).
However, the examiner respectfully disagrees. In response to applicant's argument that the references fail to show certain features of the invention, it is noted that the features upon which applicant relies (i.e., “enriched data structures containing augmented skill information” and “enriched, semantically meaningful profiles from raw natural language data”) are not recited in the rejected claim(s). Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993). Furthermore, “The test for obviousness is what the combined teachings of the references would have suggested to one of ordinary skill in the art.” (MPEP 2143.01). Notably, the claims also do not specify what data these “augmented profiles” and “augmented textural data” include, but instead the elements are broadly recited.
As seen in the previous office action (pgs. 12-14, non-final office action mailed 11/5/2025), the examiner cites to Guru as teaching an employee data component 212 using machine learning techniques to extract skills from employee data (e.g. resume data) and output/augment the data as employee vector data (Guru: ¶ 0029-0030 showing employee data component receive incoming data input 302, e.g. resume information, extracts relevant skills using machine learning techniques, and represents the data as multi-dimensional vectors, i.e. generates augmented profiles for the plurality of employees; also see ¶ 0043 showing “skill component 111, through employee data component 212, generates a multi-dimensional vector that represents the job skills of the employee. For example, a multi-dimensional vector is created for employee_A 350 that is based on the skillset (e.g., data analytical skill, programming computing skill, etc.)”). The generation of multi-dimensional vectors representing employee job skills, i.e. additional data, with the original extracted employee data (which includes resumes/text that one of ordinary skill would understand to as a type of natural language data) using machine learning, clearly reads on the broadest reasonable interpretation of “augmenting, by the computing device, the first natural language data using a first machine learning model to generate a set of augmented profiles for the plurality of users.”
The previous office action further states that Guru further teaches a separate job data component 211 using machine learning techniques to extract skills from job description data and output/augment the data as a job description vector (Guru: ¶ 0028, ¶ 0030, ¶ 0032 job data component receives input data, e.g. job data/job description, and extracts data related to the job profile as multi-dimensional vectors; also ¶ 0044 “Skill component 111 generates job description vector (step 406). In an embodiment, skill component 111, through job data component 211, generates a multi-dimensional vector that represents the job description”). Inputting textual data from a job description and using machine learning techniques to generate a job description vector, i.e. additional data, using the textual data, clearly reads on the broadest reasonable interpretation of “augmenting, by the computing device, the second natural language data using a […] machine learning model to generate a set of augmented textual data.”
Thus, Guru merely lacks an explicit distinction that the employee data component and job data component use two different machine learning models. However, Miguel (US 20210240787 A1) teaches a system for using multiple machine learning models to match users to job opportunities including augmenting a first set of input data (which may include natural language/text data) to output a first set of augmented data corresponding to users, using a first machine learning model (Miguel: ¶ 0128-0129 see using first machine learning model ensemble 341 which includes an NLP algorithm, a clustering algorithm, and a boosted trees algorithm), and augmenting a second set of input data, which may include natural language/text data) to output a second set of augmented data using a second machine learning model (Miguel: ¶ 0128-0129 showing using an intermediate, i.e. second, machine learning model ensemble 342 which includes natural language processing, clustering, and vectorization). Therefore, Miguel cures the deficiencies of Guru above by teaching the use of multiple machine learning models for generating augmented data.
Applicant further argues that the examiner does not provide sufficient motivation to combine the prior art references above and is based upon improper hindsight reasoning (pg. 13, remarks). However, the examiner respectfully disagrees. First, in response to applicant's argument that the examiner's conclusion of obviousness is based upon improper hindsight reasoning, it must be recognized that any judgment on obviousness is in a sense necessarily a reconstruction based upon hindsight reasoning. But so long as it takes into account only knowledge which was within the level of ordinary skill at the time the claimed invention was made, and does not include knowledge gleaned only from the applicant's disclosure, such a reconstruction is proper. See In re McLaughlin, 443 F.2d 1392, 170 USPQ 209 (CCPA 1971).
As indicated in the previous office action (pgs. 13-14, non-final rejection mailed 11/5/2025) – It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have included the use of multiple different machine learning models for augmenting input data sets as part of matching users with job opportunities of Miguel in the system for matching an individual to a job role requirement of Guru (such that the machine learning techniques of the job data component and the employee data component as taught by Guru are performed using distinct first and second machine learning models) with a reasonable expectation of success of arriving at the claimed invention, with the motivation to “use machine learning models to collaboratively learn from a community of users and provide predictions to individuals based on data provided from a larger group. This enables sensitive information of the group of users to be used to provide predictions for another user, while at the same time keeping the sensitive information hidden. The platform can process the information via machine learning models to predict better opportunities for a user” (Miguel: ¶ 0016). Further, it would also have been obvious to one of ordinary skill in the art before the effective filing date of the invention to do so, since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable.
The previous rationale provides a clear teaching, suggestion, or motivation for combining the teachings of the prior art, as described in MPEP 2143(I)(G) – (“use machine learning models to collaboratively learn from a community of users and provide predictions to individuals based on data provided from a larger group. This enables sensitive information of the group of users to be used to provide predictions for another user, while at the same time keeping the sensitive information hidden. The platform can process the information via machine learning models to predict better opportunities for a user”). Furthermore, under the rationale provided in MPEP 2143(I)(A), it would have also been obvious to one of ordinary skill in the art before the effective filing date of the invention to combine the teachings of Guru with those of Miguel, since the claimed invention is merely a combination of old elements (the use of machine learning techniques for augmentation of Guru with the use of multiple machine learning models for generating augmented data of Miguel), and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable.
The examiner also respectfully disagrees that the combination of Guru and Miguel do not teach “generating, by the computing device, a difference between the set of augmented profiles and the set of augmented textual data using a third machine learning model” (pgs. 12-13, remarks). Applicant focuses attacks the Miguel reference individually as not teaching the entire limitation, whereas the limitation is taught by the combination of both references in the rejection. One cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
Guru already teaches determining a difference between the set of augmented profile and the set of augmented textual data using machine learning techniques, but merely does not explicitly teach a third machine learning model. As seen below, Miguel teaches a third machine learning model that generates an output according to data from first and second machine learning models, and thus cures the deficiencies of Guru. More specifically, and as cited in the previous office action, Guru teaches generating, by the computing device, a difference between the set of augmented profiles and the set of augmented textual data using machine learning techniques (Guru: ¶ 0045, ¶ 0015-0016 showing determining difference between the multi-dimension person vectors and the multi-dimension job description vectors; ¶ 0034 “analysis component 213 of the present invention provides the capability of, 1) perform gap analysis by calculating vector distance and 2) process the gap analysis using NLP/machine learning techniques (i.e., generate skill gap summary report/data)”), and thus uses machine learning techniques to analyze the first and second machine learning outputs, but merely lacks an explicitly teaching of a third distinct machine learning model analyzing the outputs of the first and second machine learning models.
However, Miguel teaches receiving the outputs of first and second machine learning models as input to a third machine learning model and analyzing the outputs of first and second machine learning models by the third machine learning model (Miguel: ¶ 0128-0129, with ¶ 0129 showing “he income history data 351 output from the intermediate machine learning model ensemble 341 and the skills data 352 output from the intermediate machine learning model ensemble 342 are input into an output machine learning model 362. In addition, user data 333 (resume information, etc.) of the user and of other users, activity data 334 (e.g., in-app clicks, loads, etc.) of the user and other users, and schedule data 335 (time entries, work schedules, etc.) of the user and other users are also input into the output machine learning model 362. Based on these inputs, the output machine learning model 362 generates a recommend job 372 or list of jobs (e.g., such as shown in FIG. 2B in items 231-236) for the user A"). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have included the use of a third machine learning model for analyzing the outputs of first and second machine learning models of Miguel in the system for matching an individual to a job role requirement of Guru/Miguel (such that analysis component machine learning techniques as taught by Guru are performed by a distinct third machine learning model with data received from first/second machine learning models) for the same reasons described above.
Therefore, the examiner maintains that the combined teachings of Guru and Miguel render the limitations of independent claims 1, 11, and 20 obvious to one of ordinary skill in the art.
Applicant further argues, regarding claim 10, that Fang’s (US 20180130024 A1) disclosure of training a machine learning model does not teach generating a third machine learning model (pgs. 13-14, remarks filed 1/27/2026). However, the examiner respectfully disagrees. First, the claim broadly recites “generating” based on without any further steps indicating what generating the model requires. Under the broadest reasonable interpretation, generating can include any potential functions involved in generation of a machine learning model. Applicant’s specification at ¶ 0046 also discloses that generating the third machine learning model includes training the model, so it is unclear how training a machine learning model as taught by Fang would not teach generating a machine learning at the level of breadth claimed, based on the level of ordinary skill in the art, and in view of applicant’s own disclosure. The examiner maintains one of ordinary skill in the art of machine learning would understand that using input data for training machine learning models reads on generating a machine learning model.
Therefore, the previous § 103 rejections of claims 1-20 are maintained.
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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e. an abstract idea) without significantly more.
Step 1:
Claims 1-10 recite “A method…” (i.e. a process); claims 11-19 recite “A non-transitory computer-readable storage medium…” (i.e. an article of manufacture); and claim 20 recites “A device comprising: a processor; and a storage medium for tangibly storing thereon program logic for execution by the processor…” (i.e. a machine). These claims fall under one of the four categories of statutory subject matter and as a result, pass Step 1 of the subject matter eligibility test. However, “Determining that a claim falls within one of the four enumerated categories of patentable subject matter recited in 35 U.S.C. 101 (i.e., process, machine, manufacture, or composition of matter) in Step 1 does not end the eligibility analysis, because claims directed to nothing more than abstract ideas (such as a mathematical formula or equation), natural phenomena, and laws of nature are not eligible for patent protection.” See MPEP 2106.04. Accordingly, the examiner continues the subject matter eligibility analysis below.
Step 2A Prong One:
Independent claim 1 recites limitations (and independent claims 11 and 20 recite similar limitations) for:
issuing…a query;
receiving…a response to the query, the response including first natural language data comprising data associated with a plurality of users and second natural language data comprising textual data;
augmenting…the first natural language data…to generate a set of augmented profiles for the plurality of users;
augmenting…the second natural language data…to generate a set of augmented textual data;
generating…a difference between the set of augmented profiles and the set of augmented textual data
The limitations of independent claims 1, 11 and 20 above are determined to recite an abstract idea (i.e. retrieving first natural language data associated with users and second natural language data including textual data, augmenting the first natural language to generate user profiles, and augmenting second natural language data to generate augmented textual data, and generating a difference between the first and second augmented sets of data) for the reasons discussed in the following continued Step 2A Prong One analysis. Note that “An abstract idea can generally be described at different levels of abstraction.” Apple, Inc. v. Ameranth, Inc., 842 F.3d 1229, 1240-41 (Fed. Cir. 2016).
As described in MPEP 2106.04(a)(2)(III), “[T]he "mental processes" abstract idea grouping is defined as concepts performed in the human mind, and examples of mental processes include observations, evaluations, judgments, and opinions.” and “If a claim recites a limitation that can practically be performed in the human mind, with or without the use of a physical aid such as pen and paper, the limitation falls within the mental processes grouping, and the claim recites an abstract idea.” The limitations recited by the representative independent claims 1, 11 and 20 above cover concepts (e.g. observation, evaluation, judgment, and opinion) that can reasonably be performed in the human mind or by the human mind with the aid of simple tools such as pen and paper. For example, the paired steps for “issuing a query…” and then “receiving…a response to the query…” step amounts to observation, while the “augmenting…to generate a set of augmented profiles for the plurality of users…,” “augmenting…to generate a set of augmented textual data…,” and “generating a difference between the set of augmented profiles and the set of augmented textual data…” steps would be considered evaluations, judgments, and opinions. Therefore, as the processes above described by the representative independent claims 1, 11, and 20 can be characterized as mental processes (i.e. observation, evaluation, judgment, and opinion that could be carried out either with the human mind alone or with the aid of simple tools such as pen and paper), but for the recitation of generic computer components in the claims, the claims fall under the “mental processes” category of judicial exceptions (i.e. abstract ideas).
Step 2A Prong Two:
Independent claims 1, 11 and 20 recite the following additional elements:
“by a computing device” and “at the computing device” of claim 1
“A non-transitory computer-readable storage medium for tangibly storing computer program instructions capable of being executed by a computer processor, the computer program instructions defining steps of…” of claim 11
“A device comprising: a processor; and a storage medium for tangibly storing thereon program logic for execution by the processor, the program logic comprising: logic, executed by the processor, for…” of claim 20
“to a database” of claims 1, 11 and 20
“using a first machine learning model,” “using a second machine learning model,” and “using a third machine learning model” of claims 1, 11 and 20
“transmitting, by the computing device, a visualization of an output from the third machine learning model” of claim 1, and “transmitting a visualization of an output from the third machine learning model” of claims 11 and 20
The judicial exception (i.e. abstract idea) recited in claims 1, 11, and 20 is not integrated into a practical application because the claims recite mere instructions to apply the abstract idea (i.e. retrieving first natural language data associated with users and second natural language data including textual data, augmenting the first natural language to generate user profiles, and augmenting second natural language data to generate augmented textual data, and generating a difference between the first and second augmented sets of data) using generic computers/computer components (i.e. “by a computing device” and “at the computing device,” of claim 1, “A non-transitory computer-readable storage medium for tangibly storing computer program instructions capable of being executed by a computer processor, the computer program instructions defining steps of…” of claim 11, “A device comprising: a processor; and a storage medium for tangibly storing thereon program logic for execution by the processor, the program logic comprising: logic, executed by the processor, for…” of claim 20, “to a database” of claims 1, 11, and 20, “using a first machine learning model,” “using a second machine learning model,” and “using a third machine learning model” of claims 1, 11 and 20, and “transmitting, by the computing device, a visualization of an output from the third machine learning model” of claim 1, and “transmitting a visualization of an output from the third machine learning model” of claims 11 and 20).
The limitations for issuing (sending) a query is sent to a database, receiving a response by a computing device, and transmitting a visualization of an output, all recite the use of generic computers in their ordinary capacity (e.g. receiving or transmitting data) to apply the abstract idea.” See MPEP 2106.05(f), showing “[C]laims that amount to nothing more than an instruction to apply the abstract idea using a generic computer do not render an abstract idea eligible. Alice Corp.”
Further, the use of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea does not integrate a judicial exception into a practical application or provide significantly more, but instead also indicates that the claims recite mere instructions apply the abstract idea using a generic computer or computer components. Furthermore, applicant is merely “using” first, second and third machine learning models as generic machine learning models to apply the abstract idea, without any indication that the claimed invention improves upon the machine learning technology or any other technical field. See Recentive Analytics, Inc. v. Fox. Corp., showing “[P]atents that do no more than claim the application of generic machine learning to new data environments, without disclosing improvements to the machine learning models to be applied, are patent ineligible under § 101.” Further see Recentive Analytics, Inc. v. Fox. Corp. showing "Finally, the claimed methods are not rendered patent eligible by the fact that (using existing machine learning technology) they perform a task previously undertaken by humans with greater speed and efficiency than could previously be achieved.”
Therefore, because the claims, considered as a whole, do not recite anything that integrates the abstract idea into a practical application, the claims are directed to an abstract idea.
Step 2B:
Claims 1, 11, and 20 do not include additional elements, whether considered alone or as an ordered combination, that are sufficient to amount to significantly more than the judicial exception (i.e. abstract idea) because as mentioned above, the claims recite mere instructions to apply the abstract idea (i.e. retrieving first natural language data associated with users and second natural language data including textual data, augmenting the first natural language to generate user profiles, and augmenting second natural language data to generate augmented textual data, and generating a difference between the first and second augmented sets of data) using generic computers/computer components (i.e. “by a computing device” and “at the computing device,” of claim 1, “A non-transitory computer-readable storage medium for tangibly storing computer program instructions capable of being executed by a computer processor, the computer program instructions defining steps of…” of claim 11, “A device comprising: a processor; and a storage medium for tangibly storing thereon program logic for execution by the processor, the program logic comprising: logic, executed by the processor, for…” of claim 20, “to a database” of claims 1, 11, and 20, “using a first machine learning model,” “using a second machine learning model,” and “using a third machine learning model” of claims 1, 11 and 20, and “transmitting, by the computing device, a visualization of an output from the third machine learning model” of claim 1, and “transmitting a visualization of an output from the third machine learning model” of claims 11 and 20).
The limitations for issuing (sending) a query is sent to a database, receiving a response by a computing device, and transmitting a visualization of an output, all recite the use of generic computers in their ordinary capacity (e.g. receiving or transmitting data) to apply the abstract idea.
Furthermore, applicant is merely “using” first, second and third machine learning models as generic machine learning models to apply the abstract idea, without any indication that the claimed invention improves upon the machine learning technology or any other technical field. See Recentive Analytics, Inc. v. Fox. Corp., showing “[P]atents that do no more than claim the application of generic machine learning to new data environments, without disclosing improvements to the machine learning models to be applied, are patent ineligible under § 101.” Further, see Recentive Analytics, Inc. v. Fox. Corp. showing "Finally, the claimed methods are not rendered patent eligible by the fact that (using existing machine learning technology) they perform a task previously undertaken by humans with greater speed and efficiency than could previously be achieved.”
Considering the additional elements above as an ordered combination does not alter the analysis above such that the claims would amount to significantly more. Therefore, claims 1, 11, and 20 are directed to an abstract idea without significantly more.
Dependent Claims 2-10 and 12-19:
Dependent claims 2-10 and 12-19 are directed to the same abstract idea as respective independent claims 1 and 11 above as they do not recite anything that integrates the abstract idea into a practical application or amounts to significantly more than the abstract idea.
Dependent claims 2, 4, 12, and 14 recite additional steps describing the abstract idea (“receiving…third natural language data …augmenting…the third natural language data to generate an additional set of augmented textual data; and generating…a difference between at least one of: the set of augmented profiles and the additional set of augmented textual data; or the set of augmented textual data and the additional set of augmented textual data” of claims 2 and 12; and “extracting the first natural language data…from at least one of: a digital user profile; a digital user resume; a digital user certificate; a digital user record; or a digital project record” of claims 4 and 14) and recite mere instructions to further apply the abstract step using generic computers/computer components (i.e. “at the computing device” and “by the computing device,” of claims 2 and 4, “using the third machine learning model” of claim 2, and “the computer program instructions further defining steps of…” of claims 12 and 14). Similar to above, the “third machine learning model” of claim 2 merely uses an existing/generically recited machine learning model to apply the abstract idea. Additionally, to the extent “a digital user profile; a digital user resume; a digital user certificate; a digital user record; or a digital project record” recite additional elements, they at best generally link the performance of the abstract idea to a particular technological environment (digital data) that is consistent with applying the abstract within a computer environment.
Dependent claims 3, 5-9, 13, 15-19 do not add any additional elements but merely further describe the abstract idea above by reciting limitations for: describing data included in the first natural language data (claims 3, 7, 13 and 17); wherein the output comprises a semantic skills insight (claims 5 and 15); “wherein the semantic skills insight comprises at least one of…” (claims 6 and 16); describing the semantic insight relating to the specific user (claims 8 and 18); describing the one or more skills of the specific user (claims 9 and 19); and
Claim 10 recites “generating the third machine learning model based on the first natural language data and the second natural language data” which reads on training the machine learning model. However, these are generic machine learning functions for implementing machine learning technology in its ordinary capacity, and do not improve machine learning technology itself, but instead amount to nothing more than the use of generic machine learning to apply the abstract idea, by at best, describing data used to generate the machine learning model. See Recentive Analytics, Inc. v. Fox. Corp., showing “The requirements that the machine learning model be “iteratively trained” or dynamically adjusted in the Machine Learning Training patents do not represent a technological improvement.”
Therefore, claims 1-20 are ineligible under § 101.
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.
Claims 1-9 and 11-20 are rejected under 35 U.S.C. 103 as being unpatentable over US 20220092514 A1 to Guru et al. (Guru) in view of US 20210240787 A1 to Miguel et al. (Miguel).
Claim 1: Guru teaches:
A method (Guru: ¶ 0004-0006, ¶ 0040-0046 showing method(s) and process flowchart) comprising:
issuing, by a computing device, a query to a database; receiving, at the computing device, a response to the query (Guru: ¶ 0041-0042 showing “HR (Human Resources) department would like to identify skill gaps of employees within the organization to determine what training program are needed and/or career goal path. Thus, HR department of ACME Inc. gathers data from internal database that includes employee's resume, feedback and job descriptions of all jobs within the company” and “a talent acquisition portion of HR department would like to search for suitable candidate based on the various resumes received for an open job role, junior data scientist of ACME Inc.,” i.e. querying for data from a database; with ¶ 0025-0026, ¶ 0040-0041 showing skill component 111 of server 110 accesses and retrieves job data and employee data from a database 116), the response including first natural language data comprising data associated with a plurality of users (Guru: ¶ 0026, ¶ 0029, ¶ 0041 showing employee (or job candidate) data includes N number of employee resumes, evaluations, etc.; reads on “natural language” data which includes semantic/text information) and second natural language data comprising textual data (Guru: ¶ 0026, 0028, ¶ 0041 showing job data includes a job description including job requirements and job role, which reads on “natural language” as it includes textual/semantic data; for more explicit “textual data,” see ¶ 0030-0031 showing it is words that are converted to vectors, e.g. word2vec);
With respect to the limitations:
augmenting, by the computing device, the first natural language data using a first machine learning model to generate a set of augmented profiles for the plurality of users;
augmenting, by the computing device, the second natural language data using a second machine learning model to generate a set of augmented textual data;
Guru teaches i) an employee data component 212 using machine learning techniques to extract skills from employee data (e.g. resume data) and output/augment the data as employee vector data (Guru: ¶ 0029-0030 showing employee data component receive incoming data input 302, e.g. resume information, extracts relevant skills using machine learning techniques, and represents the data as multi-dimensional vectors, i.e. generates augmented profiles for the plurality of employees; also see ¶ 0043 showing “skill component 111, through employee data component 212, generates a multi-dimensional vector that represents the job skills of the employee. For example, a multi-dimensional vector is created for employee_A 350 that is based on the skillset (e.g., data analytical skill, programming computing skill, etc.)”), and ii) a separate job data component 211 using machine learning techniques to extract skills from job description data and output/augment the data as a job description vector (Guru: ¶ 0028, ¶ 0030, ¶ 0032 job data component receives input data, e.g. job data/job description, and extracts data related to the job profile as multi-dimensional vectors; also ¶ 0044 “Skill component 111 generates job description vector (step 406). In an embodiment, skill component 111, through job data component 211, generates a multi-dimensional vector that represents the job description”) – but Guru merely lacks an explicit distinction that the employee data component and job data component use two different machine learning models.
However, Miguel teaches a system for using machine learning models to match users to job opportunities including augmenting a first set of input data (which may include natural language/text data) to output a first set of augmented data corresponding to users, using a first machine learning model (Miguel: ¶ 0128-0129 see using first machine learning model ensemble 341 which includes an NLP algorithm, a clustering algorithm, and a boosted trees algorithm), and augmenting a second set of input data, which may include natural language/text data) to output a second set of augmented data using a second machine learning model (Miguel: ¶ 0128-0129 showing using an intermediate, i.e. second, machine learning model ensemble 342 which includes natural language processing, clustering, and vectorization).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have included the use of multiple different machine learning models for augmenting input data sets as part of matching users with job opportunities of Miguel in the system for matching an individual to a job role requirement of Guru (such that the machine learning techniques of the job data component and the employee data component as taught by Guru are performed using distinct first and second machine learning models) with a reasonable expectation of success of arriving at the claimed invention, with the motivation to “use machine learning models to collaboratively learn from a community of users and provide predictions to individuals based on data provided from a larger group. This enables sensitive information of the group of users to be used to provide predictions for another user, while at the same time keeping the sensitive information hidden. The platform can process the information via machine learning models to predict better opportunities for a user” (Miguel: ¶ 0016). Further, it would also have been obvious to one of ordinary skill in the art before the effective filing date of the invention to do so, since the claimed invention is merely a combination of old elements (the use of machine learning techniques for augmentation of Guru with the use of multiple machine learning models for generating augmented data of Miguel), and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable.
With respect to the limitation(s):
generating, by the computing device, a difference between the set of augmented profiles and the set of augmented textual data using a third machine learning model; and
Guru teaches generating, by the computing device, a difference between the set of augmented profiles and the set of augmented textual data using machine learning techniques (Guru: ¶ 0045, ¶ 0015-0016 showing determining difference between the multi-dimension person vectors and the multi-dimension job description vectors; ¶ 0034 “analysis component 213 of the present invention provides the capability of, 1) perform gap analysis by calculating vector distance and 2) process the gap analysis using NLP/machine learning techniques (i.e., generate skill gap summary report/data)”), and thus uses machine learning techniques to analyze the first and second machine learning outputs and determine a difference between multi-dimensional person vectors and multi-dimensional job description vectors, but merely lacks an explicitly teaching of a third distinct machine learning model analyzing the outputs of the first and second machine learning models.
However, Miguel teaches receiving the outputs of first and second machine learning models as input to a third machine learning model and analyzing the outputs of first and second machine learning models by the third machine learning model (Miguel: ¶ 0128-0129, with ¶ 0129 showing “he income history data 351 output from the intermediate machine learning model ensemble 341 and the skills data 352 output from the intermediate machine learning model ensemble 342 are input into an output machine learning model 362. In addition, user data 333 (resume information, etc.) of the user and of other users, activity data 334 (e.g., in-app clicks, loads, etc.) of the user and other users, and schedule data 335 (time entries, work schedules, etc.) of the user and other users are also input into the output machine learning model 362. Based on these inputs, the output machine learning model 362 generates a recommend job 372 or list of jobs (e.g., such as shown in FIG. 2B in items 231-236) for the user A"). It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have included the use of a third machine learning model for analyzing the outputs of first and second machine learning models of Miguel in the system for matching an individual to a job role requirement of Guru/Miguel (such that analysis component machine learning techniques as taught by Guru are performed by a distinct third machine learning model with data received from first/second machine learning models) for the same reasons described in the limitations above.
Guru, as modified above (such that the “generating” step involves a “third” distinct machine learning model comparing outputs of “first” and “second” distinct machine learning models as per Miguel), further teaches:
transmitting, by the computing device, a visualization of an output from the third machine learning model (Guru: ¶ 0037, ¶ 0039, ¶ 0046-0051, incl. Tables 1 and 2, showing generating and sending or outputting skill gap summary/report)
Claim 2: Guru/Miguel teach claim 1. Guru, as modified above, further teaches:
further comprising: receiving, at the computing device, third natural language data comprising additional textual data (Guru: ¶ 0034 showing there may be 1000 job roles, i.e. the process may be repeated for each job role included in the job data; therefore, see ¶ 0026, 0028, ¶ 0041 as above showing receiving job data including job descriptions);
augmenting, by the computing device, the third natural language data to generate an additional set of augmented textual data (Guru: ¶ 0028, ¶ 0030, ¶ 0032 job data component of the server receives input data, e.g. job data/job description, and extracts data related to the job profile as multi-dimensional vectors; also ¶ 0044 “Skill component 111 generates job description vector (step 406). In an embodiment, skill component 111, through job data component 211, generates a multi-dimensional vector that represents the job description”); and
generating, by the computing device using the third machine learning model, a difference between at least one of: the set of augmented profiles and the additional set of augmented textual data (Guru: ¶ 0045, ¶ 0015-0016, ¶ 0034 showing determining difference between the multi-dimension person vectors and the multi-dimension job description vectors; ¶ 0034 “analysis component 213 of the present invention provides the capability of, 1) perform gap analysis by calculating vector distance and 2) process the gap analysis using NLP/machine learning techniques (i.e., generate skill gap summary report/data)” and showing “[I]n an organization, there will be “N” employees and “D” job description. Each employee will have “n” skills and descriptions will have “d” skills. Computing skills gap analysis for each employee will be an (N*D) operation. For example, if an organization has 4 million employees and 1000 Job Roles then the number of computation will be 4 million*1000 Computations (that is thousand times of four million)”); or
the set of augmented textual data and the additional set of augmented textual data
Claim 3: Guru/Miguel teach claim 1. Guru, as modified above, further teaches:
wherein the first natural language data comprises at least one of: a user job description; user-generated content; a description of a learning course completed by a user; or a description of a project completed by a user (Guru: ¶ 0026, ¶ 0029, ¶ 0041 resume data; ¶ 0031 showing “if a person has passed a PMP certification,” i.e. learning course; ¶ 0029 “Data input 302 represents incoming data to be fed into the system. 302 represents N (variable representing unknown number of employee's resume) of resume of employees. Data for 302 can also include evaluation reports (e.g., performance assessment, quarterly/yearly review, etc.), social media profiles, manager feedback, peer review feedback, client feedback, deliverables met, and performance feedback, etc. Thus, the data (from all sources) can provide an accurate representation of employee skills and its proficiency level based on the experience captured”; i.e. at least a job description, user-generated content, and description of projects (deliverables) completed)
Claim 4: Guru/Miguel teach claim 1. Guru, as modified above, further teaches:
further comprising extracting the first natural language data, by the computing device, from at least one of: a digital user profile (Guru: ¶ 0026, ¶ 0029 data related to employee profile extracted, including social media profile); a digital user resume (Guru: ¶ 0016, ¶ 0029, ¶ 0041-0042 showing resume used to extract skills data); a digital user certificate; a digital user record (Guru: ¶ 0015 “extract key skills like function skills (Java, Python, etc.), soft skills (e.g., project Management, consulting, etc.) using trained entity recognition model from employee profiles”); or a digital project record (Guru: ¶ 0026, ¶ 0029 showing “deliverables met”)
Claim 5: Guru/Miguel teach claim 1. Guru, as modified above, further teaches:
wherein the output comprises a semantic skills insight (Guru: ¶ 0037, ¶ 0039, ¶ 0046-0051 showing the output includes a skill gap summary, including semantic insight such as “The candidate exhibits sound knowledge of computer science fundamentals, object-oriented programming, excellent debugging, and in-depth analytical reasoning skills. He/she has a thorough command over any of the third-generation programming language—Python. He/she may have solid understanding of data structures and some of the most commonly used algorithms. He/she needs to improve on his/her design and coding practices and a desire to develop new bold ideas. He/she has demonstrated excellent verbal and written communication skills. He/she needs to improve on his/her ability to work independently and multi-task effectively.”)
Claim 6: Guru/Miguel teach claim 5. Guru, as modified above, further teaches:
wherein the semantic skills insight (Guru: ¶ 0034, ¶ 0037, ¶ 0039, ¶ 0041, ¶ 0046-0051 generating skill gap summary report/data) comprises at least one of:
a semantic insight relating to a skill covered by a current workforce (Guru: ¶ 0047 describing skills of a particular employee/candidate for a job role, e.g. “The candidate exhibits sound knowledge of computer science fundamentals, object-oriented programming, excellent debugging, and in-depth analytical reasoning skills…”);
a semantic insight relating to a level of proficiency in a skill covered by the current workforce (Guru: ¶ 0047 describing skills of a particular employee/candidate for a job role, e.g. “The candidate exhibits sound knowledge of computer science fundamentals, object-oriented programming, excellent debugging, and in-depth analytical reasoning skills…”);
a semantic insight relating to a skills gap corresponding to a skill that is not covered by the current workforce (Guru: ¶ 0041 “identify skill gaps of employees within the organization to determine what training program are needed” and ¶ 0039 “the gap analysis report can be used by HR to identify employees that needs training or identify a career path progression for an employee,” and ¶ 0047 “He/she needs to improve on his/her ability to work independently and multi-task effectively.”);
a summary of skills corresponding to jobs being recruited via open job postings (Guru: ¶ 0051 and Table 2 showing required skills for an open job role and summary of how candidates meet those skills);
a summary of skills that the current workforce is losing via attrition;
a summary of how specific skills are represented in the current workforce;
a suggestion for transitioning the current workforce from a current set of skills to a new set of skills; or
a summary of skills-based industry hiring trends
Claim 7: Guru/Miguel teach claim 1. Guru, as modified above (to include a third machine learning model as per Miguel above), further teaches:
wherein: the first natural language data comprises data associated with a specific user within the plurality of users (Guru: ¶ 0026, ¶ 0029, ¶ 0032 showing each employee profile corresponding to data specific to an employee; see ¶ 0041 showing data corresponding to employee A);
and the output from the third machine learning model comprises a semantic insight relating to the specific user (Guru: ¶ 0034, ¶ 0037, ¶ 0046-0052 showing the skill gap summary generated by the analysis component using machine learning is specific to employee A)
Claim 8: Guru/Miguel teach claim 7. Guru, as modified above, further teaches:
wherein the semantic insight relating to the specific user relates to at least one of:
one or more skills of the specific user (Guru: ¶ 0046-0049 showing the skill gap summary pertains to skill specific to the employee);
a job history of the specific user;
a current job corresponding to the specific user;
one or more skills areas where the specific user is weak according to a skills metric (Guru: ¶ 0046-0049 showing the skill gap summary pertains to skill gap where employee needs improvement, e.g. “ He/she needs to improve on his/her design and coding practices and a desire to develop new bold ideas” and “He/she needs to improve on his/her ability to work independently and multi-task effectively.”); or
one or more learning courses completed by the specific user
Claim 9: Guru/Miguel teach claim 8. Guru, as modified above, further teaches:
wherein the one or more skills of the specific user comprise at least one of:
a current skill of the specific user (Guru: ¶ 0046-0049 showing the skill gap summary pertains to skills specific to the employee);
a skill that the specific user lacks (Guru: ¶ 0046-0049 showing the skill gap summary pertains to skill gap where employee needs improvement, e.g. “He/she needs to improve on his/her design and coding practices and a desire to develop new bold ideas” and “He/she needs to improve on his/her ability to work independently and multi-task effectively.”); or
a skill predicted to be beneficial for the user to obtain (Guru: ¶ 0046-0049 showing the skill gap summary pertains to skill gap where employee needs improvement, e.g. “ He/she needs to improve on his/her design and coding practices and a desire to develop new bold ideas” and “He/she needs to improve on his/her ability to work independently and multi-task effectively.”)
Claim 11: See the rejection of claim 1 above teaching analogous limitations. Guru further teaches A non-transitory computer-readable storage medium for tangibly storing computer program instructions capable of being executed by a computer processor, the computer program instructions defining steps (Guru: ¶ 0006, ¶ 0060-0062, ¶ 0065 showing computer readable medium storing program instructions executed by a processor to implement to perform acts according to the embodiment of the present invention).
Claim 12: See the rejection of claim 2 above.
Claim 13: See the rejection of claim 3 above.
Claim 14: See the rejection of claim 4 above.
Claim 15: See the rejection of claim 5 above.
Claim 16: See the rejection of claim 6 above.
Claim 17: See the rejection of claim 7 above.
Claim 18: See the rejection of claim 8 above.
Claim 19: See the rejection of claim 9 above.
Claim 20: See the rejection of claim 1 above teaching analogous limitations. Guru further teaches A device comprising: a processor; and a storage medium for tangibly storing thereon program logic for execution by the processor (Guru: ¶ 0052-0062 computing device including processor, and memory storing program instructions executed by the processor).
Claim 10 is rejected under 35 U.S.C. 103 as being unpatentable over US 20220092514 A1 to Guru et al. (Guru) in view of US 20210240787 A1 to Miguel et al. (Miguel), and further in view of US 20180130024 A1 to Fang et al. (Fang).
Claim 10: Guru/Miguel teach claim 1. With respect to the following limitation, the combination of Guru/Miguel, as seen in claim 1 above, teaches the use of a third machine learning model, but they do not teach generating the machine learning model based on the first natural language data and the second natural language data. However, Fang teaches:
further comprising generating the third machine learning model based on the first natural language data and the second natural language data ((Fang: ¶ 0025 “[A] machine learning model based on a synonymous search technique can be trained to score resumes in regard to a job pipeline…A training set of data for the machine learning model can include labels relating to, for example, progress of a job candidate in her job related interactions with the company. The training set of data for the machine learning model can include features of structured data from resumes or other sources relating to, for example, information regarding significant educational or professional achievements”; also see ¶ 0035 “the machine learning models can be trained with a training set of data including a resume corpus of any suitable number (e.g., 2 million, 500, etc.) of resumes (or curricula vitae) …information relating to educational history and professional experience of job candidates from other data sources apart from resumes, such as online data sources and professional networking websites…structured data from the resumes (or other data sources) can be selected as features for training. The structured data can include information relating to, for example, skills, past projects, and experience reflected in the resumes”)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have included training the model on a large training set of resume and job data of Fang in the system for matching an individual to a job role requirement of Guru/Miguel with a reasonable expectation of success of arriving at the claimed invention, with the motivation to “optimize the machine learning model for identifying highly qualified job candidates for the job pipeline” (Fang: ¶ 0025). It would also have been obvious to one of ordinary skill in the art before the effective filing date of the invention to do so, since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable.
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 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 Hunter Molnar whose telephone number is (571)272-8271. The examiner can normally be reached Monday - Friday, 7:30 - 4:00 EST.
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, Jeffrey Zimmerman can be reached at (571)272-4602. 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.
/HUNTER MOLNAR/Examiner, Art Unit 3628
Prosecution Insights