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 .
Status of Claims
Claims 1, 4 – 5, 9 – 10, 13 – 14, and 18 - 19 have been amended and are hereby entered.
Claims 6 and 15 were cancelled.
Claims 1 - 5, 7 - 14 and 16 - 20 are pending and have been examined.
This action is made FINAL.
Response to Arguments
Applicant's arguments filed October 22, 2025 have been fully considered but they are not persuasive.
Regarding the applicant's arguments against the 101 rejection of pending claims on pages 10-18: Applicant’s arguments directed to 101 analysis were considered. However, these arguments are not persuasive and the Examiner respectfully disagrees for the following reasons:
For Step 2A-Prong 1 starting in p. 11: The Applicant argues that the pending claims are not directed to any of the abstract ideas identified because the “claimed invention is inextricably tied to a machine (processor), which is one of the statutory categories of invention and cannot be considered as methods of organizing human activity”. However, the Examiner finds this argument unpersuasive and respectfully disagrees. Because such incorrect absolute assertion is not part of the 101 analysis dictated by the MPEP 2106 and the two criteria for subject matter eligibility (the claimed must fall in one of the statutory categories and the claimed invention also must qualify as patent-eligible subject matter, i.e., the claim must not be directed to a judicial exception unless the claim as a whole includes additional limitations amounting to significantly more than the exception). Moreover, the broadest reasonable interpretation (BRI) of the claim be established before examining a claim for eligibility, thus these claims can be interpreted as encompassing an abstract idea (see Synopsys, Inc. v. Mentor Graphics Corp., 839 F.3d 1138, 1147-49, 120 USPQ2d 1473, 1480-81 (Fed. Cir. 2016)). Finally, the determination of whether a claim satisfies the criteria for subject matter eligibility includes steps for products (i.e. systems or machine) and processes (i.e. methods) and in step 2A the claim as a whole is further evaluated to whether be directed to a judicial exception or not. Thus, claims that were found to recite one of the statutory categories are further required under the 101 analysis to be evaluated to find if the claim is directed to an abstract idea or judicial exception (see MPEP 2106.04).
For Step 2A-Prong 2 and Step 2B starting in p. 14: The Applicant alleges that the claims integrate, the judicial exception identified, into a practical application and points out different paragraphs from the Applicant specifications to show the “improvement” that the claimed invention reflects by providing “a solution to technical problems of shortlisting a plurality of candidates from a large data group of candidates” using an AI model that can further be trained/re-trained and by “providing an efficient and optimized system for the hiring of the candidates through virtual interviews, resulting in efficient hiring of suitable candidates that have strong technical or non-technical skills matching with job requirements” (see p.17 from Remarks for Step 2B). However, the Examiner finds this argument is unpersuasive and respectfully disagrees. Because the abstract ideas identified in the claims are not being integrated into a practical application or does not amount to significantly more than the judicial exception (e.g. abstract idea(s)) itself when considering the additional elements individually and in combination. Rather, the claim limitations invoked the use of a computer as a tool to perform an abstract idea (see MPEP 2106.04(d)(I) and MPEP 2106.05(f)). Thus, not providing an inventive concept at Step 2B. Moreover, the claim steps further describe the end result without providing details on how this alleged “improvement” to the computer functioning and/or to the existing technology for recruitment platform systems using AI models is achieved. Specifically, the claims clearly recited the use of the “AI model” that is broadly claimed to achieve the end result of efficiently shortlisting candidates based on weighted scores that were obtained from the assessment of each candidate and their virtual interaction, and that are periodically updated for accurate assessments. Finally, “claiming 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” (see MPEP 2106.05(f)(2); TLC communications).
Thus, for all the reasons stated above, the Examiner respectfully disagrees, and maintains 35 USC § 101 rejection for these pending claims.
Regarding to Applicant's arguments of rejection under 35 USC § 102(a)(1) for the pending claims on pages 18 – 22: Applicant’s arguments with respect to claim(s) 1, 10 and 19 have been considered but are moot because the new grounds of rejection now rely on the combination of the prior art of Preuss and Hardtke as a 35 USC § 103 rejection for obviousness. Moreover, and for clarity purposes, the Applicant arguments are not persuasive because the Applicant is focusing on each prior art teaching, rather than focusing on the actual language claimed in each claim limitation mentioned in pp. 18 – 22 from Remarks and how their corresponding limitation steps are different from the prior art teachings while considering the broadest reasonable interpretation (BRI) of the claims. For these same reasons, Applicant’s arguments with respect to Preuss, maintained herein, are unpersuasive as Applicant's arguments fail to comply with 37 CFR 1.111(b) and the Examiner maintains that Preuss in view of Hardtke reasonably teach the independent claims’ limitations. Please, refer to the Claim Rejections - 35 USC § 103 section for further details.
Regarding to Applicant's arguments of rejection under 35 USC § 103 for the pending claims on page 23: Applicant’s arguments regarding the dependent claims not being taught by the prior art are not persuasive since the Applicant failed to provide specific reasons as to how the language of the claims patentably distinguishes them from the Preuss and Hardtke references. Therefore, the Examiner respectfully disagrees, and maintains the updated 35 USC § 103 rejection for these pending claims.
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 - 5, 7 - 14 and 16 - 20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The analysis of this claimed invention recited in the claims begins in view of independent claim 1, the most representative claim of the independent claims set 1, 10 and 19, as follows:
At Step 1: Claims 1 – 5 and 7 - 9, falls under statutory category of a system, claims 10 – 14 and 16 – 18 are directed to a process while claims 19 – 20 are directed to an article of manufacture.
At Step 2A Prong 1: Claim 1 (representative of claims 10 and 19) recites an abstract idea in the following limitations:
retrieve a job description;
analyze…the job description to generate assessment data, wherein the assessment data comprises:
a plurality of assessment parameters, wherein the plurality of assessment parameters includes technical skills associated with the job description and non- technical skills associated with the job description, and
a weightage score that is associated with each of the plurality of assessment parameters;
retrieve a plurality of resumes associated with a plurality of candidates;
train…to identify specific keywords from the plurality of resumes, wherein the specific keywords are associated with the assessment data;
generate a plurality of initial scores for the plurality of resumes, based on the identification of the specific keywords from the plurality of resumes;
shortlist candidates of the plurality of candidates for a virtual interaction, based on the plurality of initial scores that are greater than a threshold value, wherein each initial score of the plurality of initial scores is associated with each shortlisted candidate of the shortlisted candidates;
initiate the virtual interaction for the shortlisted candidates;
retrieve response data of each shortlisted candidate associated of the shortlisted candidates with the assessment data, based on the initiation of the virtual interaction for the shortlisted candidates;
generate…a score for the response data based on the assessment data, wherein the score comprises an audio sentiment score;
obtain user input based at least on the score of the response data; and
update the weightage score associated with at least one of the plurality of assessment parameters based on the user input, wherein…is re-trained based on the updated weightage score.
Generally, and as disclosed in the specification in ¶0004 – 5, this claimed invention provides the assessment of “skills of candidate through virtual interaction using an AI model” that is “trained” and further generates a “score”, updates a “weightage score” and gets re-trained based on it. However, the abstract idea(s) of a certain method of organizing human activity (See MPEP 2106.04(a)(2), subsection II) are/is recited in claim 1 in the form of “commercial or legal interactions”. Specifically, the abstract idea is recited in the steps of “generate…a score for the response data based on the assessment data”, “shortlist candidates of the plurality of candidates for a virtual interaction…”and “obtain user input based at least on the score of the response data”. Because generating a score to shortlist candidates based on assessment data (see ¶0064 and ¶0067 from Specifications), then shortlist the candidates for a virtual interaction (i.e. for virtual interviews) and receiving user input related to the candidate response for the assessment taken after the virtual interaction (see ¶0070 and ¶0098 from Specifications) at least encompasses interactions related to business relations such as job interviews for recruitment (i.e. providing recruitment services).
Additionally, the steps of “generate a plurality of initial scores for the plurality of resumes…”, “shortlist candidates of the plurality of candidates for a virtual interaction…” and “generate…a score for the response data based on the assessment data” fall under the abstract idea of mental processes that can be practically be performed in the human mind or in pen and paper (See MPEP 2106.04(a)(2), subsection III). Because generating a plurality of initial scores for resumes to shortlist their respective candidates for a virtual interaction (i.e. a virtual interview) and generate a score for the candidate’s response data based on assessment data (e.g. assessment parameters and weightage score) after the virtual interaction requires evaluation and judgement.
At Step 2A Prong 2: For independent claims 1, 10 and 19, The judicial exception(s) or abstract idea previously identified is not integrated into a practical application (see MPEP 2106.04 (d)). The claims recite the additional element(s) of one or more processors, a memory coupled (from claim 1); a[n] AI model and AI model is re-trained (from claims 1, 10 and 19). These additional elements, individually and in combination, and while considering the claims as a whole, are merely used as a tool to perform the abstract idea (See MPEP 2106.05(f)). Specifically, steps directed in part to the use of an “AI model” to “analyze” job descriptions for assessment data generation, “train the Al model to identify specific keywords from the plurality of resumes…” and “generate, using the AI model, a score for the response data based on the assessment data” and “update the weightage score …wherein the AI model is re-trained based on the updated weightage score” are recited as being performed by the computer and an AI model used. The computer and the AI model used are recited at a high level of generality that is being used as a tool to perform the generic computer functions for generating the score for shortlisting candidates. Thus, these steps mentioned above are further describing and applying the abstract idea without placing any limits on how the technological components are being improved, while distinguishing in the claim language, the performing limitations from functions that generic computer components can perform.
Finally, the steps of “retrieve” resume, job description and response data, “initiate the virtual interaction…”, “obtain” input data and “update the weightage score…” in the representative claim are really nothing more than links to computer for implementing the use of 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 (refer to MPEP 2106.05 f (2)). Thus, in these limitation steps, the computer is used to perform an abstract idea, as discussed above in Step 2A, Prong One, such that it amounts to no more than mere instructions to apply the exception using a generic computer, and an AI model broadly claimed in this particular case.
Therefore, this analysis is indicative of the fact that even when viewed in combination, the claims’ additional elements do not integrate the abstract idea or judicial exception into a practical application.
Step 2B: For independent claims 1, 10 and 19, these claims do not provide an inventive concept. The recited additional elements of the claim(s) are the following: one or more processors, a memory coupled (from claim 1); a[n] AI model and AI model is re-trained (from claims 1, 10 and 19). These additional elements are not sufficient to amount significantly more than the judicial exception or abstract idea (see MPEP 2106.05). Because, as indicated in Step 2A Prong 2, these additional element(s) claimed are merely, instructions to “apply” the abstract ideas, which cannot provide an inventive concept. Also, the recitation of a computer to perform the claim limitations amounts to no more than mere instructions to apply the exception using a generic computer component. Thus, even when considered in combination, these additional elements represent mere instructions to implement an abstract idea or other exception on a computer, which do not provide an inventive concept at Step 2B.
For dependent claims 2 - 5, 7 - 9, 11- 14 and 16 - 20, the same analysis is incorporated. Due to their dependency to the independent claims analyzed, these claims cover or fall under the same abstract idea(s) of a method of organizing human activity and mental processes. They describe additional limitations steps of:
Claims 2 - 5, 7 - 9, 11- 14 and 16 - 20: further describes the abstract idea of the method for assessing a virtual interaction and further specifies the assessment data, corresponding response features, assessment parameters and response data (e.g. input/video/audio data) with corresponding response segments. Further specifies the analysis of response segments and resume, the generation of a segment score, the score for the response data, an initial score to shortlist the candidate and add the candidate to a list. Thus, being directed to the abstract idea group of “managing personal behavior or relationships or interactions between people” as it is encompassing interactions related to business relations and requires evaluation and judgement.
Step 2A Prong 2 and Step 2B: For dependent claims 5 and 14, these claims recite the additional elements directed to AI model is further trained. This additional element not actively and positively recited, as well as broadly recited for such further “training”, invokes a computer using the AI model, to be merely used as a tool to perform or “apply” the abstract idea(s) to the existing process of generating the score for shortlisting a candidate. Thus, amounting to no more than mere instructions to “apply” the exception using a generic computer component (MPEP 2106.05(f) and (f)(2)). Accordingly, for the same reasons stated above, these additional element(s) claimed cannot provide an inventive concept at Step 2B.
Finally, the additional elements previously mentioned above, are nothing more than descriptive language about the elements that define the abstract idea, and these claims remain rejected under 101 as well.
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.
Claims 1 - 5, 7 - 14 and 16 - 20 are rejected under 35 U.S.C. 103 as being unpatentable over Preuss (U.S. Pub No. 20210233031 A1) in view of Hardtke (U.S. Pub No. 20140122355 A1).
Regarding claims 1, 10 and 19:
Preuss teaches:
one or more processors; and a memory coupled to the one or more processors, the memory having stored therein instructions executable by the one or more processors to configure the system to: (See Fig. 6 (600 and 602): teaches a memory and a CPU. See ¶0078 for more details.)
retrieve a job description; (In ¶0028; Fig. 1 (129): teaches that “in response to receiving position information data inputs from an employer 104, the data management engine 132 can store the data inputs as a portion of the position information 129 for the respective position” and the “the data management engine 132 can link the questions from the interview question data 112 that were asked of the candidate 102 to the position information 129 along with question scores 120, candidate overall scores 122, and baseline data 114 to candidate profile data 118” which is directed to retrieving “position information” or job description.)
analyze, using a[n] Al model, the job description to generate assessment data, wherein the assessment data comprises: (In ¶0025 – 26; Fig. 1 (108 and 148): teaches that “the employers 104 interact with one or more system-generated user interface screens to identify interview questions and define ideal competencies, attributes, and personality traits of an ideal employee that can be used by the system to automatically assess how well suited a particular candidate is for a job”. Further “the automated reaction assessment system 108 may include one or more engines or processing modules 130, 132, 134, 136, 138, 140, 142, 146 that perform processes associated with generating personality aspect mappings to questions for available positions” including “detect personality aspects from a personality model using a trained natural language classifier” which is directed to the AI model generating assessment data. Refer to ¶0029 – 30, wherein the system can include “an interview management engine 144 that controls front-end processing of a candidate interview process” and can “can generate prompts for the candidate 102 to provide a “yes” or “no” answer at different speeds (quickly, slowly, normal speed) or to just provide a nonverbal response.”. For more details about the AI model and its training refer to ¶0046 – 47.)
a plurality of assessment parameters, wherein the plurality of assessment parameters includes technical skills associated with the job description and non- technical skills associated with the job description, and (In ¶0040: teaches an example wherein “the response processing engine 148 can identify the next question from the stored interview question data 112” and these questions can include “a question directed toward work experience and/or interpersonal communication skills” which is directed to having assessment parameters of technical skills and non- technical skills associated with the job description.)
a weightage score that is associated with each of the plurality of assessment parameters; (In ¶0028; Fig. 1 (108, 114, 112, 120, 122, 136 and 132): teaches that an “automated reaction assessment system 108” includes a “data management engine 132 that organizes, stores, and controls access to data in data repository 110”, this data includes “data inputs as a portion of the position information 129 for the respective position”. Also, the system can “link the questions from the interview question data 112 that were asked of the candidate 102 to the position information 129 along with question scores 120, candidate overall scores 122, and baseline data 114 to candidate profile data 118 that provides biographical and experiential information about the candidate (e.g., demographic information, contact information, education experience, work experience and locations) as well as benchmarked trustworthiness and continuous response scores for the candidate 102 and other groups of candidates sharing similar attributes (e.g., language and cultural similarities)”. Such scores and baseline data (see ¶0040 for an example) are directed to assessment parameters and corresponding weightage scores, in accordance to their definitions as given in ¶0073 from Applicant’s disclosure. Refer to ¶0042 wherein the system also “determines whether the candidate responses meet predetermined quality standards that can be depended on by employers 104 for making hiring decisions” as another example of at least assessment parameters with standardized scores or weightage scores.)
retrieve a plurality of resumes associated with a plurality of candidates; (In ¶0025; Fig. 1 (544); Fig. 2 (203): teaches that the system can retrieve and “store resumes 546” that are stored in “one or more databases 544”. See ¶0039 – 40 for more details.)
initiate the virtual interaction for the shortlisted candidates; (In ¶0027; Fig. 1 (100, 108, 130 and 158): teaches that “the automated reaction assessment system 108 may include a user management engine 130 that may include one or more processes associated with providing an interface to interact with one or more users (e.g., individuals employed by or otherwise associated with employers 104 as well as candidates 102) within the video assessment environment 100”. For example, “the user management engine 130 can control connection and access to the automated reaction assessment system 108 by the candidates 102 and employers 104 via authentication interfaces at one or more external devices 158.”)
retrieve response data of each shortlisted candidate of the shortlisted candidates associated with the assessment data, based on the initiation of the virtual interaction for the shortlisted candidates; (In ¶0030; Fig. 1 (108, 144, 146 and 116); Fig. 4: teaches that the system captures “candidate video responses to one or more interview questions” via the “interview management engine”.)
generate, using the AI model, a score for the response data based on the assessment data, wherein the score comprises an audio sentiment score; (In ¶0043; Fig. 1 (138): teaches that the system “can also include a classification engine 138 that is configured to apply one or more trained machine learning algorithms to classify one or more detected facial expression features and/or prosodic features as being associated with a positive (yes), neutral, or negative (no) response to a close-ended question”. The “classification engine 138 can also be trained to associate detected facial expression features associated with a response value in a predetermined range (e.g., −1 to 1) such that detections of the same facial expression may be assigned different scores based on the shape, size, and/or magnitude of movement of detected facial features associated with a respective expression” which is directed to generate a score for the response data, in light of ¶0086 from Applicant’s disclosure. Refer to ¶0045 wherein the “classification engine 138 can also include a natural language classifier configured to detect one or more personality traits of a candidate 102 in transcripts of responses to open-ended questions”. As for the audio sentiment score it is interpreted as the “prosodic score” (see ¶0043 – 44) which includes “audio data” extracted for the “prosodic features” that further include sentiment in the applicant’s responses and these features are then used in the “prosodic score” (see ¶0050), in accordance to ¶0086 – 87 from Applicant disclosure.)
obtain user input based at least on the score of the response data; and update the weightage score associated with at least one of the plurality of assessment parameters based on the user input, wherein the AI model is re-trained based on the updated weightage score. (In ¶0047; Fig. 1 (138, 124 and 142): teaches that the “AI training engine 142 can augment with training data sets 124 with new data for training the nonverbal feature detection classifiers after receiving and processing each set of candidate video responses to a close-ended question set” wherein the “training data sets” include “amplifying feedback provided by employers 104 regarding the accuracy of the scores determined from the detected nonverbal features” which is directed to the user input that is used to update weightage scores for the AI model training as defined in ¶0070 and ¶0097 – 99 from Applicant’s disclosure. Refer to ¶0043 – 44 wherein the “classification engine 138, in some examples, can be trained by artificial intelligence (AI) training engine 142 with training data sets 124 of facial features that can be updated over time as the system 108 processes interview responses for candidates 102.” Refer to ¶0051 for details about the use of “intra-individual norming data and benchmarking data” that can be used in “processing future responses” and see ¶0060)
Preuss teaches an “AI training engine” that “can also compile customized training data sets 124 for training a natural language classifier to detect personality traits of the candidate 102 within transcripts of open-ended questions” and such training data sets can include “amplifying feedback provided by employers 104 regarding the accuracy of the scores determined from the detected nonverbal features” (see ¶0046 – 47; Preuss). Preuss does not explicitly teach the abilities of training the AI model to identify specific resume keywords, generate an initial score per resume and shortlist the candidate for the virtual interaction based on the initial scores exceeding a threshold value. However, Hardtke teaches:
train the Al model to identify specific keywords from the plurality of resumes, wherein the specific keywords are associated with the assessment data; (In ¶0105; Fig. 2 (210 and 203); Fig. 6A: teaches that the system can “analyze data from a large scale comparison of resumes to job openings using a method selected from” at least “machine learning; neural networks and other multi-layer perceptrons; support vector machines;” wherein the analysis of data from resumes includes resume keywords or “job features” (see ¶0047 – 48 and ¶0085) that can be further “tagged” to emphasize employer preferences over a “preferred job type for the candidate” to calculate the “suitability score” only for resumes that include the certain job features from a specific job description (see ¶0054). At least one of the AI models chosen can be trained as later described in examples wherein a “training set” of the “most important features of a successful match between a candidate and a job opening could be determined” (see ¶0130 and ¶0133).)
generate a plurality of initial scores for the plurality of resumes, based on the identification of the specific keywords from the plurality of resumes; (In ¶0052; Fig. 3 (200 – 220): teaches that the system after identifying “a plurality of job features” and “a plurality of candidate features 210 in the resume and the profile” (see ¶0046 – 48) can take “each resume that has been uploaded in turn and proceeds to calculate a suitability score 220 (also, simply, a “score” herein)” that is based on “a match between the plurality of candidate features in the resume along with any features that have been extracted from the candidate's profile or social media or other external data, and the plurality of job features in the description of the job opening” which is directed to the initial score as defined in ¶0076 and ¶0093 from Applicant’s disclosure. Refer to ¶0118, ¶0138 and ¶0147 for the training algorithms used to calculate the “function of fitness or suitability of a candidate for a job opening”.)
shortlist candidates of the plurality of candidates for a virtual interaction, based on the plurality of initial scores that are greater than a threshold value, wherein each initial score of the plurality of initial scores is associated with each shortlisted candidate of the shortlisted candidates; (In ¶0064; Fig. 2 (230 and 240): teaches that the system then “communicates 240 a notification of one or more selected resumes to an employer, or other third party submitter of the description, if a selected resume has a score that exceeds the first threshold fit for the description of a job opening provided by that employer”. Further, “whenever an employer is provided with a list of candidates whose suitability scores exceed a first or a second threshold, the employer is able to review the candidates' resumes, profiles, and any other available data, and make a decision on whether to invite one or more of the candidates to formally apply for the job opening, or to come straight to an interview” (see ¶0070).)
It would have been obvious to one of ordinary skill in the art before the earliest effective filing date of the claimed invention to modify Preuss to provide the abilities of training the AI model to identify specific resume keywords, generate an initial score per resume and shortlist the candidate for the virtual interaction based on the initial scores exceeding a threshold value, as taught by Hardtke in order to “properly ascertain a good set of features within both a candidate's resume and a description of a job opening that would lead to more reliable matching” (¶0004; Hardtke) and because for “employers, it is critical to be presented quickly with candidates who should be invited to interview” (¶0002; Hardtke).
Regarding claims 2, 11 and 20:
Preuss, as shown in the rejection above, discloses the limitations of claims 1, 10 and 19, respectively.
Preuss further teaches:
wherein the assessment data further comprises one or more questions and corresponding one or more response features associated with each of the plurality of assessment parameters. (In ¶0020; Fig. 1 (112): teaches this limitation for having assessment data since the system can receive “responses from the interviewee for all applicable questions in a set of one or more close-ended interview questions” which then the “system can generate an open-ended question for the interviewee to respond to.” Further the “system can apply a trained speech-to-text algorithm and natural language classifier to determine how well the candidate fits one or more ideal personality characteristics for the available position.” As an example, the system “can apply the open-ended question processing, scoring, and data quality assessment techniques to identify personality aspects”. Refer to ¶0028 for details about the system linking “questions from the interview question data 112 that were asked of the candidate 102 to the position information 129 along with question scores 120, candidate overall scores 122, and baseline data 114 to candidate profile data 118” as well as “benchmarked trustworthiness and continuous response scores for the candidate 102 and other groups of candidates sharing similar attributes (e.g., language and cultural similarities).”)
Regarding claims 3 and 12:
Preuss, as shown in the rejection above, discloses the limitations of claims 2 and 11, respectively.
Preuss further teaches:
wherein the response data comprises one or more response segments corresponding to the one or more questions, (In ¶0033: teaches that the system “captures the video data associated with the response, processes the data, which can include separating an audio portion from a video portion of the data file”. Then from these data portions, “verbal/nonverbal response features” can be captured and can include “one or more video frames indicating a speed and direction of the response (e.g., yes/no/neutral)” for the verbal response features and for the “nonverbal response features 126” includes “one or more video frames indicating facial expressions and/or prosodic features of the response” wherein the “captured facial expression features” can further include “visual indications of moods and/or behaviors such as happiness, sadness, surprise, neutral, anger, contempt, and disgust” (see ¶0035).)
and wherein the one or more response features comprises positive response features and negative response features for the corresponding one or more questions. (In ¶0038: teaches an example wherein “if the candidate responds “yes” to a question asking whether the candidate 102 meets minimum education requirements, the facial expression features indicate happiness, and the prosodic features indicate confidence (steady tone, steady inflection in voice), then the trustworthiness score can be toward a higher end of the stanine scale” which is directed to positive features in this example, and as response features and their definition given in ¶0079 – 80, ¶0083 from Applicant’s disclosure. As for “any open-ended questions presented to the candidate 102, the question scoring engine 140 can apply the question scoring techniques” (Also see ¶0046 for more details). Refer to ¶0049 wherein the system can calculate “baseline state scores” based on the “mood survey data” that are further used for “baseline scores” wherein “baseline state score captures an overall mood or emotional state of the candidate during the interview and can be based on detected facial expression and other mood-based features” wherein “the mood-based features can include detected facial expression features and other physiological features that can indicate a level of calm/peacefulness or agitation/nervousness/anger of the candidate (e.g., respiratory rate can indicate a level of agitation or nervousness)” which is another example of positive/negative response features. Refer to ¶0023 wherein the system with the help of a NL classifier can detect “positive and negative polarizations of the personality aspects from the personality model within an interview question transcript”.)
Regarding claims 4 and 13:
Preuss, as shown in the rejection above, discloses the limitations of claims 3 and 12, respectively.
Preuss further teaches:
wherein the one or more processors is further configured to: analyze, using the AI model, each of the one or more response segments based on the corresponding response features; (In ¶0043: teaches “classification engine 138 that is configured to apply one or more trained machine learning algorithms to classify one or more detected facial expression features and/or prosodic features as being associated with a positive (yes), neutral, or negative (no) response to a close-ended question” directed to at least one response segment.)
generate, using the AI model, a segment score for each of the one or more response segments based on the analysis; and (In ¶0043: teaches that the system’s “classification engine 138 can also be trained to associate detected facial expression features associated with a response value in a predetermined range (e.g., −1 to 1) such that detections of the same facial expression may be assigned different scores based on the shape, size, and/or magnitude of movement of detected facial features associated with a respective expression” directed to a segment score as defined in ¶0086 from Applicant’s disclosure. Refer to ¶0044 wherein the system can also be trained to associate detected “prosodic features” with “variations of positive (yes), neutral, and negative (no) responses in captured video frames” directed to another segment score. Therefore, these facial expression/prosodic features and their predetermined scores identified by the “classification engine 138” can be further determined by “the question scoring engine 140 and/or baseline calculation engine 134” to calculate respective “prosodic scores” and/or “facial expression scores” for a candidate based on the “prosodic” or “facial expression” features (see ¶0043 – 44).)
generate, using the AI model, the score for the response data based on the segment score for each of the one or more response segments. (In ¶0052; Figs. 2 – 3 (218 and 220): teaches that the system via the “question scoring engine 140 and/or baseline calculation engine 134” while using the facial expression/prosodic features and their predetermined scores identified by the “classification engine 138”, can calculate respective “prosodic scores” and/or “facial expression scores” for a candidate. In terms of “the baseline calculation engine” as disclosed in ¶0049, this engine “discards outlier video frames and calculates a baseline facial expression score and a baseline prosody score for the candidate from averages of the retained frames”. As an example, “the baseline scores 218, 220 fall in a range from −1 to 1, where −1 represents a negative response, 0 represents a neutral response, and 1 represents a positive response.” In terms of “question scoring engine” as disclosed in ¶0052, the “question scoring engine 346 can use the direction and latency of the response 342, extracted prosody and facial expression features 344, and baseline facial expression and prosody scores 218, 220 to determine a continuous response score 348 and nonverbal reaction score 350 for each question” wherein the “continuous response score 348” is directed to the score generated based on the segment scores from each response segment. As an example, “the continuous response score is the product of the direction (−1 if the response is no, and 1 for all other responses) multiplied by a speed of the response, which can be normalized to a scale between 0 and 1”.)
Regarding claims 5 and 14:
Preuss, as shown in the rejection above, discloses the limitations of claims 1 and 10, respectively.
Preuss further teaches:
wherein the AI model is further trained based on the assessment data. (In ¶0019 – 20: teaches “the system can apply a specially trained machine learning classifier that can detect one or more response attributes from the captured video data” as well as applying “a trained speech-to-text algorithm and natural language classifier to determine how well the candidate fits one or more ideal personality characteristics for the available position”. Refer to ¶0028 for more details of the training data used in the AI model which includes the video interview data captured.
Regarding claims 6 and 15:
Preuss, as shown in the rejection above, discloses the limitations of claims 1 and 10, respectively.
Preuss does not explicitly teach the abilities of retrieving a candidate’s resume to analyze it and generate an initial score via a trained AI model and shortlist the candidate for the virtual interaction. However, Hardtke teaches:
wherein the one or more processors is further configured to: retrieve a resume of the candidate; (In ¶0025; Fig. 1 (544); Fig. 2 (203): teaches that the system can retrieve and “store resumes 546” that are stored in “one or more databases 544”. See ¶0039 – 40 for more details.)
analyze, using the trained AI model, the resume of the candidate to generate an initial score; and (In ¶0052; Fig. 3 (200 – 220): teaches that the system after identifying “a plurality of job features” and “a plurality of candidate features 210 in the resume and the profile” (see ¶0046 – 48) can take “takes each resume that has been uploaded in turn and proceeds to calculate a suitability score 220 (also, simply, a “score” herein)” that is based on “a match between the plurality of candidate features in the resume along with any features that have been extracted from the candidate's profile or social media or other external data, and the plurality of job features in the description of the job opening” which is directed to the initial score as defined in ¶0076 and ¶0093 from Applicant’s disclosure. Refer to ¶0118, ¶0138 and ¶0147 for the training algorithms used to calculate the “function of fitness or suitability of a candidate for a job opening”.)
shortlist the candidate for the virtual interaction based on the initial score. (In ¶0064; Fig. 2 (230 and 240): teaches that the system then “communicates 240 a notification of one or more selected resumes to an employer, or other third party submitter of the description, if a selected resume has a score that exceeds the first threshold fit for the description of a job opening provided by that employer”. Further, “whenever an employer is provided with a list of candidates whose suitability scores exceed a first or a second threshold, the employer is able to review the candidates' resumes, profiles, and any other available data, and make a decision on whether to invite one or more of the candidates to formally apply for the job opening, or to come straight to an interview” (see ¶0070).)
It would have been obvious to one of ordinary skill in the art before the earliest effective filing date of the claimed invention to modify Preuss to provide the abilities of retrieving a candidate’s resume to analyze it and generate an initial score via a trained AI model and shortlist the candidate for the virtual interaction, as taught by Hardtke in order to have a “suitability score [that] serves both sides of the hiring process, both allowing candidates to find their optimal job, as well as employers to find their optimal candidates, and thereby engenders productivity in the successful employment of the most-suited individuals as well as efficiency in locating those individuals from among large applicant pools” (¶0010; Hardtke).
Regarding claims 7 and 16:
The combination of Preuss and Hardtke, as shown in the rejection above, discloses the limitations of claims 1 and 10, respectively.
Preuss teaches the score of the response data as the “response value” or “continuous response score 348” when aggregated (see ¶0043 and ¶0052, respectively; Preuss). However, Preuss does not explicitly teach the ability of adding a candidate to a list of shortlisted candidates based on the response data score and the initial score of the plurality of initial scores. However, Hardtke further teaches:
wherein the one or more processors is further configured to add a candidate to a list of shortlisted candidates based on the score of the response data and an initial score of the plurality of initial scores. (In ¶0063): teaches that “where a first threshold score has been set, the computer system identifies 230 for each of the one or more descriptions of job openings those resumes in the first list whose score exceeds the first threshold fit, and flags those resumes as selected resumes”. Also, the “suitability score” can be an aggregation of “external data” retrieved that is “relevant to the candidate's resume” (e.g. past/recent assessment/response data from an interview, etc.) that includes the combination of “feature scores” not necessarily “found within a job description”, such as the case of the score of the response data claimed which is covered by the primary reference (see ¶0050, ¶0080 and ¶0082 – 83). Thus, these two scores claimed were interpreted as an aggregated score for purposes of selecting candidates in a list of shortlisted candidates.)
It would have been obvious to one of ordinary skill in the art before the earliest effective filing date of the claimed invention to modify Preuss to provide the ability of adding a candidate to a list of shortlisted candidates based on the response data score and the initial score of the plurality of initial scores, as taught by Hardtke in order to “properly ascertain a good set of features within both a candidate's resume and a description of a job opening that would lead to more reliable matching” (¶0004; Hardtke) and because for “employers, it is critical to be presented quickly with candidates who should be invited to interview” (¶0002; Hardtke).
Regarding claims 8 and 17:
The combination of Preuss and Hardtke, as shown in the rejection above, discloses the limitations of claims 7 and 16, respectively.
Preuss teaches the obtaining user input (e.g. feedback) as the “amplifying feedback provided by employers regarding the accuracy of the scores” for the “detected nonverbal features” that were further determined (see ¶0047; Preuss). However, Preuss does not explicitly teach the ability of obtaining user input specifically based on the shortlisted candidates list and the candidate’s resume. However, Hardtke further teaches:
wherein the one or more processors is further configured to obtain the user input further based on the list of the shortlisted candidates and the resume of the candidate. (In ¶0104: teaches that when “a set of resumes that have been ranked by individuals whose primary profession is recruiting” to determine the “features” such as “errors” in their respective “feature scores” to determine “whether the feature is a quantity that indicates a good match between the candidate and the job opening” which is a “weighting coefficient” used to “determine the contribution of a feature score to the suitability score” and adapt it when outputting the candidate lists, in accordance to ¶0070 from Applicant’s disclosure. Also, another example of obtaining user input under the Broadest Reasonable Interpretation (BRI), is when the employer may “choose a value for the first threshold so that they see more or fewer resumes at their discretion”, this way the employer can set/adjust the threshold (e.g. directed to the user input) to their liking for the final results of the list (see ¶0059). Refer to ¶0067 wherein the employer can choose and elect different types of notification settings regarding to “notifications of job openings for which they have high scores at some frequency of their choosing” Finally, refer to ¶0117 wherein “features and filters can be customized for an individual employer” with “explicit feedback” to “identify correlations between the resumes of different candidates as well as between resumes and job descriptions to predict the top candidates for a given opening, and customize the suitability score specifically for an employer's requirements”.)
It would have been obvious to one of ordinary skill in the art before the earliest effective filing date of the claimed invention to modify Preuss to provide the ability of obtaining user input specifically based on the shortlisted candidates list and the candidate’s resume, as taught by Hardtke in order to “properly ascertain a good set of features within both a candidate's resume and a description of a job opening that would lead to more reliable matching” (¶0004; Hardtke) and because for “employers, it is critical to be presented quickly with candidates who should be invited to interview” (¶0002; Hardtke).
Regarding claims 9 and 18:
Preuss, as shown in the rejection above, discloses the limitations of claims 1 and 10, respectively.
Preuss further teaches:
wherein the response data comprises at least one of: user input data, video data, and audio data. (In ¶0033: teaches that “once an interview question response is detected, the data acquisition engine 146 captures the video data associated with the response, processes the data, which can include separating an audio portion from a video portion of the data file, and saving the processed video data in the data repository 110 as captured video response data 116.” Refer to ¶0026 wherein “the system 108 can convert video submissions to text transcripts” and to ¶0050 wherein “the interview management engine 212 can extract detectable features from the video and/or audio data of the interview question responses, which can include prosodic features and facial expression features from one or more frames of the response”.)
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Preuss - b (U.S. Pub No. 20210233030 A1) is pertinent because it is about “systems and methods for performing automated candidate video assessments include receiving, from a remote computing device of a first party via a network, a candidate video submission for an available position.”
Jose (U.S. Pub No. 20240104509 A1) is pertinent because it “relate to a recruitment system and more particularly relate to a system and a method for generating interview insights in an interviewing process to improve the efficiency and effectiveness of the interviewing process.”
Jersin (U.S. Pub No. 20190197487 A1) is pertinent because it “generally relates to computer technology for solving technical challenges in determining query attributes (e.g., locations, skills, positions, job titles, industries, years of experience and other query terms) for search queries. More specifically, the present disclosure relates to creating a stream of candidates based on attributes such as locations and titles, which may be suggested to a user, and refining the stream based on user feedback”
Goodwin (U.S. Patent No. 8548929 B1) is pertinent because it “relates generally to the field of employment candidate data management, and more particularly to methods and systems for employment candidate recruitment, pre-screening, scheduling, testing, tracking, staffing, and resume data management.”
Liu (U.S. Pub No. 20190220824 A1) is pertinent because it is “relates to automated systems for matching resumes from job applicants to job posting requirements based on machine learning techniques, and providing interviewing and hiring recommendations.”
Mayers (U.S. Pub No. 20240256999 A1) is pertinent because it “relates generally to machine learning systems for classifying behavioral content, and in particular to systems and methods for identifying and/or evaluating behavioral content to determine behavioral attributes and skills associated therewith.”
Ghosh (U.S. Pub No. 20210150486 A1) is pertinent because it is “a method can include generating a dynamically refinable set of job requirements by searching a plurality of networked systems and extracting data therefrom based on comparing data contained in the networked systems to specifications determined by natural language processing of user input. The method also can include ranking a job candidate based on comparing attributes of the job candidate to corresponding job requirements contained in the dynamically refinable set of job requirements.”
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Ivonnemary Rivera Gonzalez whose telephone number is (571)272-6158. The examiner can normally be reached Mon - Fri 9:00AM - 5:30PM.
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, Nathan Uber can be reached at (571) 270-3923. 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.
/IVONNEMARY RIVERA GONZALEZ/Examiner, Art Unit 3626
/NATHAN C UBER/Supervisory Patent Examiner, Art Unit 3626
Prosecution Insights