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
In view of the Appeal Brief filed on 12/10/2025, PROSECUTION IS HEREBY REOPENED. New grounds of rejection are set forth below.
To avoid abandonment of the application, appellant must exercise one of the following two options:
(1) file a reply under 37 CFR 1.111 (if this Office action is non-final) or a reply under 37 CFR 1.113 (if this Office action is final); or,
(2) initiate a new appeal by filing a notice of appeal under 37 CFR 41.31 followed by an appeal brief under 37 CFR 41.37. The previously paid notice of appeal fee and appeal brief fee can be applied to the new appeal. If, however, the appeal fees set forth in 37 CFR 41.20 have been increased since they were previously paid, then appellant must pay the difference between the increased fees and the amount previously paid.
A Supervisory Patent Examiner (SPE) has approved of reopening prosecution by signing below: /NATHAN C UBER/ Supervisory Patent Examiner, Art Unit 3626
DETAILED ACTION
Claims 1-20 are pending in this action.
Response to Amendments/Arguments
Applicant’s arguments regarding the 35USC101 rejection have been considered but are unpersuasive. Applicant argues the claims do not follow rules or instructions by humans, and that, “The pending claims do not recite a method of playing any sort of game, human activity related to grooming or appearance, or instructions for how to perform some tasks. As such the Examiner’s argument that the claims recite the abstract idea of “certain methods of organizing human activity” is erroneous”. Applicant subsequently states, “the claims merely involve mathematical formulas or calculations”, and that, “the claimed matching attributes of a user to attributes of a job posting does not specifically recite a mathematical relationship between variables or numbers. There is no recitation of a ration, equivalency, or an equation”. Applicant then avers, “applying a loss function” does not recite a specific relationship between variables or numbers, a ratio, equivalency or equation. Also generating a score based on a similarity is not based on any specific mathematical relationship either. As such the three claimed features identified by the Examiner do not recite a mathematical relationship”. Applicant then states, “the claimed invention improves computer related technology or technical field because it is configured to “address the technical problems of the prior art by improving the manner in which a machine learning model is trained to score attributes, and improving the manner in which a query rewriter derives a query, based on scored attributes, for use in selecting candidate content items”. Examiner points out that Claim 1 is considered an abstract idea because the limitations as claimed, pertains to certain methods of organizing human activity including managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules or instructions) whereby attributes of a user are provided and matched to a query, a loss function is applied to generate a corresponding score for the plurality of attributes based on a similarity, and a subset of job postings are presented to a user as recommendations based on the ranking scores of the job postings, which is directed to managing interactions between people including following rules or instructions. Claim 1 is also directed to mathematical concepts (mathematical relationships, mathematical formulas or equations, mathematical calculations) since the claim limitations require applying by a query writer a loss function to generate a corresponding score for the plurality of attributes based on a similarity between a user and a job posting having two or more common attributes, and deriving a ranking score for each job posting; hence directed to mathematical concepts. Further, as noted in the Final action, applicant’s query re-writer, applying a loss function to the machine learning model is merely used as a tool to further process received data based on rules logic- see applicant’s disclosure, ¶9: “a machine learning model for use in ranking attributes to be used in a query of a candidate selection technique”; ¶14: “machine learning model is optimized using a loss function that expresses similarity between an end-user and a content item as a match between at least "kc" attributes”; ¶17: “a pre-trained machine learned model receives as input the attributes of the end-user, and then outputs a score for each attribute”; ¶24: “after the machine learned model 310 has generated scores for each of the attributes of the end-user, the query processor 312 of the query rewriter 308 derives a query 314 including a term for each attribute having a score that exceeds a predetermined threshold. The query 314 is derived as a weighted-OR query, with individual weighting factors being assigned to each term, and a threshold score for the query”; ¶29: “the query rewriter selects those attributes that have scores exceeding some predetermined threshold, for use in a weighted OR query for selecting the candidate content items (e.g., job postings)”, and amounts to no more than applying the judicial exception using generic computing components, and linking the use of the judicial exception to a computing environment. Hence, the additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea.
Applicant’s arguments regarding the 35USC103 rejection, specifically the secondary reference Xue are persuasive, the rejection is withdrawn. Examiner has modified the rejection with Halabi reference to teach applicant’s query re-writer. Moreover, Examiner has modified the rejection to further explain how the limitations are being interpreted and addressed each of applicant’s claims as noted below in this Non-Final action.
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 an abstract idea without significantly more. Claims 1-7 are directed to a process (an act, or series of acts or steps); claims 8-14 are directed to a system and claims 15-20 are directed to a non-transitory computer readable storage medium. Thus claims 1-20 fall within one of the four statutory categories.
Step 2A-Prong 1:
Claim 1 recites in part, “…training with training data, a machine learning model to receive as input a plurality of attributes associated with an end-user of an online service and to generate as output a score for each attribute, the score for an attribute indicating the predictive power of the attribute when used as a term of a query for use in selecting candidate job postings to recommend to the end-user, the training data comprising a plurality of attributes i) obtained for an end-user and job posting pair for which the end-user has previously applied for a job, associated with the job posting, and ii) having at least "k" attributes for the end-user matching corresponding attributes of the job posting; subsequent to training the machine learning model: receiving a request to generate a plurality of job posting recommendations for a first end-user of the online service; responsive to receiving the request, obtaining a plurality of attributes associated with the first end-user; providing the plurality of attributes associated with the first end-user as input to a query rewriter, the query rewriter including the trained machine learning model; applying, by the query rewriter, a loss function to the trained machine learning model to optimize the trained machine learning model to generate the score for the plurality of attributes based on a similarity that two or more attributes to be shared in common between the first end-user and one of the plurality of job postings; generating, by the trained machine learning model, for each attribute the score indicating the predictive power of the attribute when used as a term of a query for use in selecting candidate job postings to recommend to the end-user; deriving using the query rewriter a query for fetching candidate job postings, the query i) including a term for each attribute in the plurality of attributes for which the machine learning model generated a corresponding score that exceeds a predetermined threshold, and ii) when executed against a plurality of job postings, fetches as candidate job postings those job postings in the plurality of job postings that have at least "k" attributes matching attributes expressed in the terms of the query; executing the query to fetch a plurality of candidate job postings; processing the plurality of candidate job postings to derive a ranking score for each job posting; and based at least in part on the ranking scores of the plurality of candidate job postings, selecting a subset of the plurality of job postings for presentation as recommendations to the first end-user.”
The underlined limitations above demonstrate independent claim 1 is directed toward the abstract idea for receiving and matching attributes of a user to a query, generating corresponding attribute scores and presenting a subset of ranked job postings as recommendations to a user based on a similarity between a user and a job posting having two or more common attributes in a computing environment.
Applicant’s specification discusses a recommendation system for online job postings with a candidate selection technique using machine learning to rank and select attributes of a user to be used with a query for selecting the candidate job postings. The specification also discloses ranking attributes of an end-user so that a query can be generated to return relevant content items (job postings) for recommendation. (¶1- ¶4, ¶14). Claim 1 is considered an abstract idea because the (underlined) limitations as claimed, pertains to certain methods of organizing human activity including managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules or instructions) whereby attributes of a user are provided and matched to a query, a loss function is applied to generate a corresponding score for the plurality of attributes based on a similarity, and a subset of job postings are presented to a user as recommendations based on the ranking scores of the job postings, which is directed to managing interactions between people including following rules or instructions. With the exception of generic computing components, the limitations are merely using computing components as a tool to perform the abstract idea. The limitations are also directed to mathematical concepts (mathematical relationships, mathematical formulas or equations, mathematical calculations) since the claim limitations require applying by a query writer a loss function to generate a corresponding score for the plurality of attributes based on a similarity between a user and a job posting having two or more common attributes, and deriving a ranking score for each job posting; hence directed to mathematical concepts. Therefore, the claim recites an abstract idea--see MPEP 2106.04(II).
Step 2A-Prong 2: This judicial exception is not integrated into a practical application because the additional elements “a machine learning model, ””query rewriter”, “loss function” [claim 1]; “processor”, “computer-readable instructions”, “memory storage device storing computer-readable instructions”, “system” [claim 8]; “non-transitory computer-readable storage medium”, “processing circuit” [claim 15] for (receiving, generating, obtaining, providing, deriving, executing, processing and selecting) data gathering and analysis and merely to provide instructions for managing information, and to implement the abstract idea recited above utilizing “a machine learning model, ””query rewriter”, “loss function” [claim 1]; “processor”, “computer-readable instructions”, “memory storage device storing computer-readable instructions”, “system” [claim 8]; “non-transitory computer-readable storage medium”, “processing circuit” [claim 15] as a tool to perform the abstract idea, and generally links the abstract idea to a particular technological environment. See MPEP 2106.05 (f-h).
Independent claim 1 fails to operate “a machine learning model, ””query rewriter”, “loss function” [claim 1]; “processor”, “computer-readable instructions”, “memory storage device storing computer-readable instructions”, “system” [claim 8]; “non-transitory computer-readable storage medium”, “processing circuit” [claim 15] (which is merely a nominal recitation of a standard computer technology, database and hardware/software components) in any exceptional manner, and there is no evidence in the disclosure to suggest achieving an actual improvement in the computer functionality itself, or improvement in any specific computer technology other than utilizing ordinary computational tools to automate and perform the abstract idea for receiving and matching attributes of a user to a query, generating corresponding attribute scores and presenting a subset of ranked job postings as recommendations to a user based on a similarity between a user and a job posting having two or more common attributes in a computing environment —see MPEP 2106.05(a). Accordingly, applicant has not shown an improvement or practical application under the guidance of MPEP section 2106.04(d) or 2106.05(a). Applicant’s limitations as recited above do nothing more than supplement the abstract idea using generic computer components performing generic computer functions (receiving, generating, obtaining, providing, applying, deriving, executing, processing and selecting) such that it amounts to no more than mere instruction to apply the exception using a generic computer component-see MPEP 2106.05(f) and linking the use of the judicial exception to a particular technological environment as discussed in MPEP 2106.05(h). Independent claims 8 and 15 recite substantially similar limitations as independent claim 1 and therefore also recite the same abstract idea.
Dependent claims 2-7, 9-14 and 16-20 fail to cure the deficiencies of the above noted independent claim from which they depend and are therefore rejected under the same grounds. The dependent claims further recite the abstract idea without imposing any meaningful limits on practicing the abstract idea. Dependent claims 2, 9 and 16 recite in part, “wherein the loss function comprises”; claims 3 and 10 recite in part, “deriving the query as a ..”; claims 4 and 11 recite in part, “assigning to the query a query threshold score”; claims 5, 12 and 19 recite in part, “wherein the threshold score is”; claims 6, 13 and 20 recite in part, “wherein the value of k is set to”; claims 7 and 14 recite in part, “wherein obtaining”; claim 17 recites in part, “ wherein the query for fetching candidate job postings comprises”; claim 18 recites in part, “wherein the processing circuit is further configured to”, which is still directed toward the abstract idea identified previously and are no more than mere instructions to apply the exception using a computer or with computing components. Therefore, the abstract idea fails to integrate into any practical application. Thus, under Step 2A-Prong Two the claims are directed to an abstract idea.
Step 2B: The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because as discussed above, with respect to integration of the abstract idea into a practical application, the additional element “a machine learning model”, “query rewriter”, “loss function” [claim 1]; “processor”, “computer-readable instructions”, “memory storage device storing computer-readable instructions”, “system” [claim 8]; “non-transitory computer-readable storage medium”, “processing circuit” [claim 15] amounts to no more than mere instructions to apply the exception using a generic computer component and linking the use of the judicial exception to a computing environment which does not integrate a judicial exception into a practical application nor provide an inventive concept (significantly more than the abstract idea). In this case, the “a machine learning model, “query rewriter”, “loss function” [claim 1]; “processor”, “computer-readable instructions”, “memory storage device storing computer-readable instructions”, “system” [claim 8]; “non-transitory computer-readable storage medium”, “processing circuit” [claim 15] are generically used to further process and store received data- see ¶51: computer system within which a set of instructions may be executed for causing the machine to perform any one or more of the methodologies discussed herein, according to an example embodiment. Specifically, FIG. 7 shows a diagrammatic representation of the machine 900 in the example form of a computer system, within which instructions 916 (e.g., software, a program, an application, an applet, an app, or other executable code) for causing the machine 900 to perform any one or more of the methodologies discussed herein may be executed. For example, the instructions 916 may cause the machine 900 to execute any one of the methods or algorithms described herein… The machine 900 may comprise, but not be limited to, a server computer, a client computer, a PC, a tablet computer, a laptop computer… “.
Further applicant’s query re-writer, applying a loss function to the machine learning model is merely used as a tool to further process received data based on rules logic- see applicant’s disclosure, ¶9: “a machine learning model for use in ranking attributes to be used in a query of a candidate selection technique”; ¶14: “machine learning model is optimized using a loss function that expresses similarity between an end-user and a content item as a match between at least "kc" attributes”; ¶17: “a pre-trained machine learned model receives as input the attributes of the end-user, and then outputs a score for each attribute”; ¶24: “after the machine learned model 310 has generated scores for each of the attributes of the end-user, the query processor 312 of the query rewriter 308 derives a query 314 including a term for each attribute having a score that exceeds a predetermined threshold. The query 314 is derived as a weighted-OR query, with individual weighting factors being assigned to each term, and a threshold score for the query”; ¶29: “the query rewriter selects those attributes that have scores exceeding some predetermined threshold, for use in a weighted OR query for selecting the candidate content items (e.g., job postings)”, and amounts to no more than applying the judicial exception using generic computing components, and linking the use of the judicial exception to a computing environment.
Hence, the additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea.
Accordingly, even when considered as a whole, the claims do not transform the abstract idea into a patent-eligible invention since the claim limitations do not amount to a practical application or significantly more than an abstract idea for receiving and matching attributes of a user to a query, generating corresponding attribute scores and presenting a subset of ranked job postings as recommendations to a user based on a similarity between a user and a job posting having two or more common attributes in a computing environment.
Hence, claims 1-20 are directed to non-statutory subject matter and are rejected under 35 USC 101. See 2019 PEG and MPEP 2106.
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 for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1, 3, 4, 7, 8, 10, 11, 14, 15, 17 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Heath et al., (WO 2023/154064 A1) in view of Tang et al., US Patent Application Publication No US 2017/0300230 A1.
With respect to claims 1, 8 and 15,
Heath discloses,
A computer-implemented method comprising: training with training data, a machine learning model to receive as input a plurality of attributes associated with an end-user of an online service and to generate as output a score for each attribute, (¶7: “The method may include executing by at least one processor on a computer, a job matching algorithm stored on a non-transitory computer-readable memory medium that performs online job opening matching between a job-seeker profile and one or more job opening profiles”; ¶8: “the job matching algorithm may perform matching, to determine a weighted match score, corresponding plurality of attributes in the qualifications between the job-seeker profile and one or more job opening profiles from an employer database”; ¶25: “Job-seeker profile’s corresponding attributes in qualifications and characteristics are multi-dimensionally mapped to those in the job opening profile to calculate an overall score against a threshold score for successful matching”; ¶27: “the job matching algorithm 150a may retrieve from an applicant database 122, job-seeker’s qualifications and characteristics from job-seeker’s profile, and retrieve employer’s job qualifications and characteristics from one or more job opening profiles. In implementation, a plurality of attributes in the qualifications and a plurality of attributes in the characteristics may be abstracted from job-seeker’s answers to a list of questions using the job-seeker’s communication device 123 (e.g., smart phone, laptop computer, desk top computer, tablet, etc. The plurality of attributes in the qualifications and the plurality of attributes in the characteristics may be stored in the applicant database 122 under the job-seeker’s profile”; ¶35: “, the elastic analysis may be performed automatically by the job matching algorithm 150a or by training an artificial intelligence (AI) component 151 (shown in FIG.1B)”; ¶40: “The AI system 151 may be trained to model an elastic analysis based on a designated range to test a sensitivity of the qualifications match or fitness match. For example, certain job requires minimal human interactions but heavy on analytical skills (e.g., research, engineering), so the attributes on social skills may have a lighter weight, but the technical attributes may be adjusted heavier in weight”;¶76: “Insights and data from the employer behavior and actions block 582 may be used by the employer application 520 as a feedback loop to incrementally and continuously improve itself via machine learning and AI algorithms”)
the score for an attribute indicating the predictive power of the attribute when used as a term of a query for use in selecting candidate job postings to recommend to the end-user (¶25: “Job-seeker profile’s corresponding attributes in qualifications and characteristics are multi-dimensionally mapped to those in the job opening profile to calculate an overall score against a threshold score for successful matching. The job- seeker’s profile and the job opening’s profile may also be re-matched using an elastic analysis to model job seeker’s strengths and growth potential for a same or a different job, through adjusting individual weights of each attribute of the qualifications and characteristics of one or both of the job-seeker’s profile and the job opening profile to optimize risks or successes, and adaptability for the job-seeker in taking the job opening”; ¶38: “In the weighted average summing, attributes which are more important may be assigned a higher weight value while other attributes may be assigned a less weight or a zero weight”; ¶41: “the job-seeker may elect to answer additional profile questions to enable the job matching algorithm 150 to automatically adjust a functional weight assignment to one or more attributes in a respective category of the qualifications or characteristics in the job-seeker profile for current job opening re-matching or for a new job opening matching”)
the training data comprising a plurality of attributes i) obtained for an end-user and job posting pair for which the end-user has previously applied for a job, associated with the job posting, and ii) having at least "k" attributes for the end-user matching corresponding attributes of the job posting (¶8: “the job matching algorithm may perform: (a) adjusting one or both of: the weighted match score and the weighted fit score to re-determine an adjusted overall score; and (b) performing, according to the adjusted overall score, one or both of: a re-matching to the one or more job opening profiles according to the first threshold score, and a new matching to a new job opening profile from the employer database, according to a second threshold score, wherein the new job opening profile has not been previously matched …when after multiple iterations the job matching algorithm still does not advance the job-seeker to a next recruitment or hiring decision on behalf of the employer, the job matching algorithm may make recommendations as what actions the job-seeker may take to qualify for a different job that had not been previously matched to, after receiving certain training, certification or pursuit of a degree. In effect, the job matching algorithm tool may engage both the employers and the job-seekers as a coach by providing a tool that performs broad and elastic analysis to match the best available resources to the best opportunities with good fit”; ¶38: “In the weighted average summing, attributes which are more important may be assigned a higher weight value while other attributes may be assigned a less weight or a zero weight”; ¶51: “job matching algorithm 150a may include training an artificial intelligence (AI) system to perform the adjusting of the weighted match score and the adjusting of the weighted fit score for the re-matching to the one or more job opening profiles and the new matching to the new job opening profile as shown in Figs 1A to 1C. When the re-matching or the new matching is unsuccessful, the job matching algorithm 150a may invite the job- seeker to consider other jobs, and in some cases the job matching algorithm 150a may query to explore other career paths based on the elastic analysis in matching to more job openings from an employer database 124 under a curation criterion over a defined range of the adjusted weighted match score and over a defined range of the adjusted weighted fit score”)
subsequent to training the machine learning model: receiving a request to generate a plurality of job posting recommendations for a first end-user of the online service; (¶8: “the job matching algorithm may make recommendations as what actions the job-seeker may take to qualify for a different job that had not been previously matched to, after receiving certain training, certification or pursuit of a degree”; ¶59: “The AI system 200 may dynamically present job seeker’s elements of their profile, such as the qualifications attributes and characteristics attributes (with or without an associated weight) to complete answers to questions based on previously filled in information”; ¶60: “In block 216, the job-seeker may set his/her job search and matching criteria. For example, the job-seekers may have the option to set the criteria by which jobs and occupations are presented to them based on match scores, fit scores, geographic location, city radius or range of travel distance for job, full-time or part-time work, demographics, and other custom criteria defined by the job-seeker”; ¶62: “If a successful match is found, that is, the scores and curation criteria are met for both employer and job seeker, then blocks 224 and 226 in the system 200 may proceed to the next level decision to automatically invite the job seeker to apply to the job to begin the hiring process. Alternately, when no match is found, block 228 in the AI system may recommend other jobs and career paths for consideration. When the scores and curation criteria are not met for both employer and job seeker, the system’s AI 200 may automatically recommend other jobs and career paths for the job seeker to consider based on their score and curation criteria.”)
responsive to receiving the request, obtaining a plurality of attributes associated with the first end-user (¶59: “the AI system 200 may guide a job-seeker to create an account in filling out their profile to be stored in an applicant database 122. The AI system may assist the job-seeker in filling out their profile to maximize their match and fit scores for jobs, occupations, and career paths… The AI system 200 may dynamically present job seeker’s elements of their profile, such as the qualifications attributes and characteristics attributes (with or without an associated weight) to complete answers to questions based on previously filled in information”; ¶60: “the job-seeker may set his/her job search and matching criteria. For example, the job-seekers may have the option to set the criteria by which jobs and occupations are presented to them based on match scores, fit scores, geographic location, city radius or range of travel distance for job, full-time or part-time work, demographics, and other custom criteria defined by the job-seeker. ¶61: “The AI system 200 may use the job matching algorithm in block 220 to perform live matching of the job-seeker profile from the applicant database 122, to the one or more job openings profile from the employer database 124 in blocks 210 and 218. The curation criteria evaluation in block 222 may evaluate the job opening profile and the job-seeker to determine a weighted match score, and a weighted fit score according to the job opening curation criteria to determine which events (i.e., which matched attributes) to trigger next in the recruiting and hiring process”).
generating by the trained machine learning model, for each attribute the score indicating the predictive power of the attribute when used as a term of a query for use in selecting candidate job postings to recommend to the end-user; (¶32: “the plurality of attributes in the qualifications and the plurality of attributes in the characteristics of the job-seeker profile may be abstracted from job-seeker’s answers to a list of questions by the Job Seeker Qualifications block 152 and the Job Seeker Characteristics block 156 respectively, in the job matching algorithm 150a”; ¶34; ¶38: “it may be informative to include or to accommodate certain amount of contributions from the qualifications and characteristics in the job-seeker profile when calculating the overall score, for a more complete picture in selecting the best matched and best fit talent among a list of matched job-seekers in the analysis”; ¶39: “the job matching algorithm 150 may be configured to perform an elastic analysis by selecting a portion (or entire portion) of the attributes in both the qualifications and the characteristics to have the corresponding assigned weights be adjusted (to increase or decrease) to provide outcome re-evaluations in order to optimize risks, successes, and adaptability for the best matched job-seeker when taking the job opening”; ¶40-¶45; ¶42: “the employer may specify a range of weight adjustment in each attribute when posting the job opening profile (use default weight when no range is provided). The employer may benefit from using the elastic analysis information to adjust a selected portion of the attributes according to their importance in order to glean a larger pool of talents, to weigh the strengths versus the weaknesses of each candidate prior to proceeding to a next level decision for an interview or for hiring”; ¶59: “the AI system 200 may guide a job-seeker to create an account in filling out their profile to be stored in an applicant database 122. The AI system may assist the job-seeker in filling out their profile to maximize their match and fit scores for jobs, occupations, and career paths… The AI system 200 may dynamically present job seeker’s elements of their profile, such as the qualifications attributes and characteristics attributes (with or without an associated weight) to complete answers to questions based on previously filled in information”; ¶61: “The AI system 200 may use the job matching algorithm in block 220 to perform live matching of the job-seeker profile from the applicant database 122, to the one or more job openings profile from the employer database 124 in blocks 210 and 218. The curation criteria evaluation in block 222 may evaluate the job opening profile and the job-seeker to determine a weighted match score, and a weighted fit score according to the job opening curation criteria to determine which events (i.e., which matched attributes) to trigger next in the recruiting and hiring process”; ¶62: “ If a successful match is found, that is, the scores and curation criteria are met for both employer and job seeker, then blocks 224 and 226 in the system 200 may proceed to the next level decision to automatically invite the job seeker to apply to the job to begin the hiring process. Alternately, when no match is found, block 228 in the AI system may recommend other jobs and career paths for consideration. When the scores and curation criteria are not met for both employer and job seeker, the system’s AI 200 may automatically recommend other jobs and career paths for the job seeker to consider based on their score and curation criteria”; ¶74: “Insights and data from job-seeker behavior and actions block 550 may be used by the job seeker application 520 as a feedback loop to incrementally and continuously improve itself via machine learning and AI algorithms)
a processor configured to execute computer-readable instructions; and a memory storage device storing computer-readable instructions, which, when executed by the processor, cause the system to: (¶7: “The method may include executing by at least one processor on a computer, a job matching algorithm stored on a non-transitory computer-readable memory medium that performs online job opening matching between a job-seeker profile and one or more job opening profiles”)
Heath discloses all of the above limitations, Heath does not distinctly describe the following limitations, but Tang however as shown discloses,
providing the plurality of attributes associated with the first end-user as input to a query rewriter, the query rewriter including the trained machine learning model (¶60: “The query rewrite model learner 740 in this example may retrieve the parallel training data from the training data database 735 and generate a query rewrite model based on the parallel training data, e.g. via a machine learning algorithm”; ¶64: “the query rewrite model learner 740 in this example includes a word alignment determiner 910, a null word inserter 915, a phrase extractor 920, a phrase score determiner 930, and a query rewrite model generator 940”; ¶65: “he training data may include a plurality of query-title pairs each of which comprises a query previously submitted by a user and a title of a document or a URL searched out in response to the query. For each query-title pair, the word alignment determiner 910 can determine one or more alignment candidates”)
applying, by the query rewriter, a loss function to the trained machine learning model to optimize the trained machine learning model to generate the score for the plurality of attributes based on a similarity that two or more attributes to be shared in common between the first end-user and one of the plurality of job postings (¶176, ¶77)
deriving using the query rewriter, a query for fetching candidate job postings, the query i) including a term for each attribute in the plurality of attributes for which the machine learning model generated a corresponding score that exceeds a predetermined threshold (¶81)
and ii) when executed against a plurality of job postings, fetches as candidate job postings those job postings in the plurality of job postings that have at least "k" attributes matching attributes expressed in the terms of the query; executing the query to fetch a plurality of candidate job postings; (¶¶80)
processing the plurality of candidate job postings to derive a ranking score for each job posting; and based at least in part on the ranking scores of the plurality of candidate job postings, selecting a subset of the plurality of job postings for presentation as recommendations to the first end-user (¶90: “the search result ranking determiner 1360 may generate search results merely based on the second ranked list of searched documents, i.e. the search results are all generated based on one or more rewritten queries”)
Heath discloses a method/system for online job matching/placement including but not limited to a job matching algorithm to determine a weighted match score, corresponding plurality of attributes in the qualifications between the job-seeker profile and one or more job opening profiles from an employer database.
Tang is directed to methods, systems, and programming for rewriting a query and providing search results. Tang further teaches that a search engine may determine whether the query needs to be rewritten. If so, the search engine may communicate with the query rewrite engine to rewrite the query and generate the search results based at least partially on the rewritten query.
Heath and Tang are directed to the same field of endeavor since they are both related to performing search queries, providing matching results in a computing environment. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the method/system for online job matching/placement of Heath with the techniques for rewriting a query as taught by Tang since it allows for improving search performance via query rewriting model (Abstract, ¶2, ¶35-¶37).
With respect to claims 3, 10 and 17,
Heath and Xue disclose all of the above limitations, Xue further discloses,
wherein deriving a query for fetching candidate job postings comprises: deriving the query as a weighted OR query (¶42: “The model 320 determines which query type to be applied to the job search query. In example embodiments, the model 320 takes the extracted feature vector (derived from the extracted features) and applies logistic regression (e.g., applying a dot product with the learned weight vector, then a Sigmoid function)”; ¶43: “Once the model 320 determines the query type, the job search query (e.g., the one or more features) is passed to either the title query rewriter 304 or the compound query rewriter 306 based on the output. More specifically, a job search query that is determined to be best implemented as a title field search is passed to the title query rewriter 304, while a job search query that is determined to be best implemented as a compound search is passed to the compound query rewriter 306”)
wherein the query for fetching candidate job postings comprises a weighted OR query ((¶42: “The model 320 determines which query type to be applied to the job search query. In example embodiments, the model 320 takes the extracted feature vector (derived from the extracted features) and applies logistic regression (e.g., applying a dot product with the learned weight vector, then a Sigmoid function)”; ¶43: “Once the model 320 determines the query type, the job search query (e.g., the one or more features) is passed to either the title query rewriter 304 or the compound query rewriter 306 based on the output. More specifically, a job search query that is determined to be best implemented as a title field search is passed to the title query rewriter 304, while a job search query that is determined to be best implemented as a compound search is passed to the compound query rewriter 306”)
Heath and Xue are directed to the same field of endeavor since they are both related to performing job search queries and matching results in a computing environment. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the method/system for online job matching/placement of Heath with the techniques/search system for classifying job search queries as taught by Xue since it allows for improved precision for determining the query type (the one or more features) by applying logistic regression (e.g., applying a dot product with the learned weight vector, then a Sigmoid function) to the extracted feature vector (Fig 2, Fig 3, ¶42-¶45).
With respect to claim 4, 11 and 18,
Heath and Xue disclose all of the above limitations, Xue further discloses,
wherein deriving the query as a weighted OR query comprises: assigning to the query a query threshold score; and assigning to each term in the query a weighting factor (¶42: the model 320 outputs a single number between zero and one, which is the probability. An output or probability of greater than or equal to 0.5 results in a title field query type, while an output or probability less than 0.5 results in a compound field query type”; “¶43: “The model 320 determines which query type to be applied to the job search query. In example embodiments, the model 320 takes the extracted feature vector (derived from the extracted features) and applies logistic regression (e.g., applying a dot product with the learned weight vector, then a Sigmoid function). Accordingly, the model 320 outputs a single number between zero and one, which is the probability. An output or probability of greater than or equal to 0.5 results in a title field query type, while an output or probability less than 0.5 results in a compound field query type”)
Heath and Xue are directed to the same field of endeavor since they are both related to performing job search queries and matching results in a computing environment. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the method/system for online job matching/placement of Heath with the techniques/search system for classifying job search queries as taught by Xue since it allows for improved precision for determining the query type (the one or more features) by applying logistic regression (e.g., applying a dot product with the learned weight vector, then a Sigmoid function) to the extracted feature vector (Fig 2, Fig 3, ¶42-¶45).
Heath further discloses,
wherein, upon evaluating the query against a particular job posting, a score is derived for the particular job posting by adding the weighting factors of each term of the query associated with an attribute that is present in the job posting; (¶37: “The Match Score block 160 may be configured to evaluate qualifications match by performing a default calculation on the first weighted match score by summing respectively, the third default functional weight values of all matched attributes in the qualifications between the job-seeker profile and the one or more job opening profiles (or the new job opening profile). Likewise, the Fit Score block 162 may be configured to evaluate characteristics match by performing a default calculation on the second weighted fit score by summing respectively, the fourth default functional weight values of all matched attributes in the characteristics between the job-seeker profile and the one or more job opening profiles or the new job opening profile”; ¶38; ¶58: “the employer may add job-seeker curation criteria. For example, the employers may have the option to set job criteria, to sort, aggregate, and clarify job seeker requirements based on match scores, fit scores, and other custom criteria defined by the employer”; ¶61: “The curation criteria evaluation in block 222 may evaluate the job opening profile and the job-seeker to determine a weighted match score, and a weighted fit score according to the job opening curation criteria to determine which events (i.e., which matched attributes) to trigger next in the recruiting and hiring process”; ¶62: “ If a successful match is found, that is, the scores and curation criteria are met for both employer and job seeker, then blocks 224 and 226 in the system 200 may proceed to the next level decision to automatically invite the job seeker to apply to the job to begin the hiring process”
fetching the particular job posting as a candidate job posting when the score for the particular job posting exceeds the query threshold score. (¶8: “When the overall score exceeds a first threshold score, a successful job opening match may be found in order to advance to a next recruitment or hiring decision, such as followed by a successful interview before presenting an offer; otherwise, the job matching algorithm may perform: (a) adjusting one or both of: the weighted match score and the weighted fit score to re-determine an adjusted overall score; and (b) performing, according to the adjusted overall score, one or both of: a re-matching to the one or more job opening profiles according to the first threshold score, and a new matching to a new job opening profile from the employer database, according to a second threshold score, wherein the new job opening profile has not been previously matched”; ¶25: “job-seeker profile’s corresponding attributes in qualifications and characteristics are multi-dimensionally mapped to those in the job opening profile to calculate an overall score against a threshold score for successful matching” ¶34; ¶37: “The Match Score block 160 may be configured to evaluate qualifications match by performing a default calculation on the first weighted match score by summing respectively, the third default functional weight values of all matched attributes in the qualifications between the job-seeker profile and the one or more job opening profiles (or the new job opening profile). Likewise, the Fit Score block 162 may be configured to evaluate characteristics match by performing a default calculation on the second weighted fit score by summing respectively, the fourth default functional weight values of all matched attributes in the characteristics between the job-seeker profile and the one or more job opening profiles or the new job opening profile”; ¶62: “If a successful match is found, that is, the scores and curation criteria are met for both employer and job seeker, then blocks 224 and 226 in the system 200 may proceed to the next level decision to automatically invite the job seeker to apply to the job to begin the hiring process”)
With respect to claims 7 and 14,
Heath and Xue disclose all of the above limitations, Heath further discloses,
wherein obtaining the plurality of attributes associated with the first end-user comprises obtaining at least one attribute from an end-user profile of the first end-user, and obtaining at least one attribute from logged data relating to an interaction the first end-user has had with content via the online service. (¶8: “the job matching algorithm may perform: (a) adjusting one or both of: the weighted match score and the weighted fit score to re-determine an adjusted overall score; and (b) performing, according to the adjusted overall score, one or both of: a re-matching to the one or more job opening profiles according to the first threshold score, and a new matching to a new job opening profile from the employer database, according to a second threshold score, wherein the new job opening profile has not been previously matched …when after multiple iterations the job matching algorithm still does not advance the job-seeker to a next recruitment or hiring decision on behalf of the employer, the job matching algorithm may make recommendations as what actions the job-seeker may take to qualify for a different job that had not been previously matched to, after receiving certain training, certification or pursuit of a degree. In effect, the job matching algorithm tool may engage both the employers and the job-seekers as a coach by providing a tool that performs broad and elastic analysis to match the best available resources to the best opportunities with good fit”; ¶59: “the AI system 200 may guide a job-seeker to create an account in filling out their profile to be stored in an applicant database 122. The AI system may assist the job-seeker in filling out their profile to maximize their match and fit scores for jobs, occupations, and career paths… The AI system 200 may dynamically present job seeker’s elements of their profile, such as the qualifications attributes and characteristics attributes (with or without an associated weight) to complete answers to questions based on previously filled in information”)
Claims 2, 5, 6, 9, 12, 13, 16, 19 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Heath, Xue, in further view of Kobayashi, (JP3562572B2).
With respect to claims 2, 9 and 16,
Heath and Xue disclose all of the above limitations, the combination of Heath and Xue does not distinctly describe the following limitations, however Kobayashi as shown discloses,
wherein the loss function comprises an approximation of a modified Heaviside Step function, (¶43: “The model 320 determines which query type to be applied to the job search query. In example embodiments, the model 320 takes the extracted feature vector (derived from the extracted features) and applies logistic regression (e.g., applying a dot product with the learned weight vector, then a Sigmoid function). Accordingly, the model 320 outputs a single number between zero and one, which is the probability. An output or probability of greater than or equal to 0.5 results in a title field query type, while an output or probability less than 0.5 results in a compound field query type”; ¶58: “The time window used in the present invention may include a symmetric Gaussian function, a delta function, a step function, and a heaviside step function”; Fig 13, ¶61: “This time window is useful for keeping track of certain things, i.e. topics that have occurred after a certain time and / or date. Also, a time window, which is zero after a particular time and / or date, and a constant weighting factor before this particular date, i.e. a "heaviside step function" ¶66: track only those that occurred before a particular date, a Heaviside "step function" that takes the value 1 before the particular date and has a value of zero after the particular date is effective. If the exact date is somewhat unclear, or if it is necessary to take into account the time delay between the receipt and posting of the news report, also a "soft curve" that changes more slowly from 1 to zero Can also be used. … step functions can of course be modified”)
wherein the modified Heaviside Step function evaluates to one, when a job posting has at least two attributes matching attributes of the end-user, and the modified Heaviside Step function evaluates to zero, when a job posting has less than two attributes matching attributes of the end-user (¶37: “FIG. 6 shows an embodiment of a combined matrix obtained by adding a date / time stamp to a document matrix. This associative matrix includes elements used as date / time stamps after the elements originally included in the document matrix. These elements are given the number 1 as shown in FIG. 6; however, other positive real numbers may be present as elements if the weighting factors are used to form a combined matrix”; ¶42: “weight factors and / or weight functions can be used for both keywords and time parameters when forming a connection matrix”;¶61: FIG. 13 (c) shows a time window of a step function with a weighting factor of zero before a particular time and / or date and a constant weighting factor for things after that particular time and / or date. . This time window is useful for keeping track of certain things, i.e. topics that have occurred after a certain time and / or date. Also, a time window, which is zero after a particular time and / or date, and a constant weighting factor before this particular date, i.e. a "heaviside step function", is for example the searcher, whose This can be used when confirming that the occurrence has occurred on a specific day and / or a specific time”; ¶66: “a Heaviside "step function" that takes the value 1 before the particular date and has a value of zero after the particular date is effective”)
Kobayashi discloses a method/system for detecting and tracking items/ classes of documents including attribute data associated with a time parameter in a database. Kobayashi further discloses a heaviside step function for tracking data, the associated weighting factor and confirmation that an occurrence has occurred. Kobayashi teaches generating keywords for performing detection and tracking based on a document and time parameters whereby weight factors and/or weight functions can be used for both keywords and time parameters when forming a connection matrix and/or binding matrix. Examiner asserts that the claim would have been obvious since all of the claimed elements were known in Kobayashi, and one skilled in the art before the effective filing date of applicant’s invention could have combined the elements as claimed with the known techniques of Kobayashi with no change in their respective functions and the combination would yield nothing more than predictable results to one of ordinary skill in the art. Narrowing the data item to job postings would have been predictable based on the method/system for detecting and tracking items/ classes of documents, the associated weighting factors/functions for both keywords and time parameters as taught by Kobayashi. Heath, Xue and Kobayashi are directed to the same field of endeavor since they are related to matching/tracking data in a computing environment. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the method/system for online job matching/placement of Heath with the techniques/search system for classifying job search queries of Xue and the techniques for detecting and tracking items/classes of documents as taught by Kobayashi since it allows for utilizing a step function with a weighting factor, matching and tracking data; and ranking documents, hence, improving the accuracy of a search and better relate to time/date dependent topics.
(Fig 6, Fig 13C, ¶37, ¶42, ¶43, ¶47, ¶61, ¶66)
With respect to claims 5, 12 and 19,
Heath and Xue disclose all of the above limitations, the combination of Heath and Xue does not distinctly describe the following limitations, however Kobayashi as shown discloses,
wherein the threshold score is set to the value of "k" and the weighting factor assigned to each term in the query is one (Fig 2, ¶32: “the date / time stamp is added to the document matrix to form a combined matrix used for the Doc / kwd query, new matter detection and tracking of step 205”; ¶34: “FIG. 3 is a diagram illustrating a document matrix. This matrix includes rows composed of documents 1 (doc1) to document n (docn), and each document has an element obtained from a keyword (kwd1 ... kwdn) included in a specific document. Contains. The number of documents and the number of keywords are not limited in the present invention and depend on the size of the document and the database. As shown in FIG. 3, the elements of the document matrix are indicated by the numeral 1, but another positive real number can also be used when forming the document matrix using weighting factors”; ¶35: “FIG. 4 shows a typical time window applied to a date / time stamp and weighting the date / time stamp. The time window shown in FIG... sub.0Has a maximum height W. Where T.sub.0Corresponds to a particular date / time of a particular document, and height corresponds to a weight factor. This type of window has a specific point T.sub.0T around.sub.0Can be used to detect new items and / or classes using a weight factor that decreases with time interval from”; ¶37: “FIG. 6 shows an embodiment of a combined matrix obtained by adding a date / time stamp to a document matrix. This associative matrix includes elements used as date / time stamps after the elements originally included in the document matrix. These elements are given the number 1 as shown in FIG. 6; however, other positive real numbers may be present as elements if the weighting factors are used to form a combined matrix”; ¶42: weight factors and / or weight functions can be used for both keywords and time parameters when forming a connection matrix”; ¶61: FIG. 13 (c) shows a time window of a step function with a weighting factor of zero before a particular time and / or date and a constant weighting factor for things after that particular time and / or date. This time window is useful for keeping track of certain things, i.e. topics that have occurred after a certain time and / or date. Also, a time window, which is zero after a particular time and / or date, and a constant weighting factor before this particular date, i.e. a "heaviside step function", is for example the searcher, whose This can be used when confirming that the occurrence has occurred on a specific day and / or a specific time”; ¶66: “a Heaviside "step function" that takes the value 1 before the particular date and has a value of zero after the particular date is effective”)
Kobayashi discloses a method/system for detecting and tracking items/ classes of documents including attribute data associated with a time parameter in a database. Kobayashi discloses a method/system for detecting and tracking items/ classes of documents including attribute data associated with a time parameter in a database. Kobayashi further discloses a heaviside step function for tracking data, the associated weighting factor and confirmation that an occurrence has occurred. Heath, Xue and Kobayashi are directed to the same field of endeavor since they are related to matching/tracking data in a computing environment. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the method/system for online job matching/placement of Heath with the techniques/search system for classifying job search queries of Xue and the method/system for detecting and tracking items/classes of documents as taught by Kobayashi since it allows for combining weighted time parameters into a document-attribute/vector matrix to form a combined matrix for improved detection and tracking of documents (Fig 4, Fig 6, Fig 13C, ¶22, ¶28, ¶32, ¶35, ¶37, ¶40, ¶42, ¶61, ¶66).
With respect to claims 6, 13 and 20,
Heath, Xue and Kobayashi disclose all of the above limitations, Kobayashi further discloses,
wherein the value of "k" is set to two (¶17: “method for detecting and tracking new items and / or classes of documents in very large databases by essentially capturing time stamp parameters such as date and time …includes attribute data associated with a time parameter, the method comprising: obtaining a vector of the document using the attribute data including the time parameter included in the document;”; ¶18: “The attribute data includes at least one keyword, and the keyword is weighted. In addition, the detecting and tracking step further includes a step of providing a time window, and the detecting and tracking step is performed using the time window. The time window is configured by a user interactively on a delta function, or a symmetric Gaussian function, or a step function, or a display window for a particular date”; ¶19: “The time parameter is further weighted with respect to the passage of time around a particular date, the weight of the time parameter being less than the sum of the weights of the keywords… If a plurality of different time windows are provided and the number of keywords in each document varies, the relative weight between the keywords and the time parameters is made substantially constant between the documents”; ¶37: “FIG. 6 shows an embodiment of a combined matrix obtained by adding a date / time stamp to a document matrix. This associative matrix includes elements used as date / time stamps after the elements originally included in the document matrix. These elements are given the number 1 as shown in FIG. 6; however, other positive real numbers may be present as elements if the weighting factors are used to form a combined matrix”; ¶42: weight factors and / or weight functions can be used for both keywords and time parameters when forming a connection matrix”; ¶54: “Construct / form a combined matrix containing date / time stamps corresponding to the temporal attributes of each document”; ¶61: FIG. 13 (c) shows a time window of a step function with a weighting factor of zero before a particular time and / or date and a constant weighting factor for things after that particular time and / or date. This time window is useful for keeping track of certain things, i.e. topics that have occurred after a certain time and / or date. Also, a time window, which is zero after a particular time and / or date, and a constant weighting factor before this particular date, i.e. a "heaviside step function", is for example the searcher, whose This can be used when confirming that the occurrence has occurred on a specific day and / or a specific time”; ¶66: “a Heaviside "step function" that takes the value 1 before the particular date and has a value of zero after the particular date is effective”)
Kobayashi discloses a method/system for detecting and tracking items/ classes of documents including attribute data associated with a time parameter in a database. Heath, Xue and Kobayashi are directed to the same field of endeavor since they are related to matching/tracking data in a computing environment. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the method/system for online job matching/placement of Heath with the techniques/search system for classifying job search queries of Xue and the techniques for detecting and tracking items/classes of documents as taught by Kobayashi since it allows for utilizing a step function with a weighting factor, matching and tracking data; and ranking documents, hence, improving the accuracy of a search and better relate to time/date dependent topics (Fig 6, Fig 13C, ¶17-¶19, ¶37, ¶42, ¶43, ¶47, ¶54, ¶61, ¶66)
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Halabi et al., US Patent No US11526512 B1, “Rewriting Queries”, relating to systems/methods for mitigating errors during processing of user inputs whereby a query may be derived from processed user input. A performance predictor analyzes the query and uses historical data to predict whether the query will return relevant results if executed. If the query's predicted performance is below a threshold, a query rewriter may identify potential alternatives to the query from a library of “known good” queries.
Wu et al., US Patent Application Publication No US 2017/0364596 A1, “Smart Suggestions for Query Refinements”, relating to a system is provided whereby refinements to a search query are automatically suggested to a searcher based on multiple facets (e.g., title, company, industry, school, location, etc.) of the search results simultaneously.
Conclusion
Any inquiry of a general nature or relating to the status of this application or concerning this communication or earlier communications from the Examiner should be directed to Kimberly L. Evans whose telephone number is 571.270.3929. The Examiner can normally be reached on Monday-Friday, 9:30am-5:00pm. If attempts to reach the examiner by telephone are unsuccessful, the Examiner’s supervisor, Lynda Jasmin can be reached at 571.272.6782.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://portal.uspto.gov/external/portal/pair <http://pair-direct.uspto.gov >. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866.217.9197 (toll-free). Any response to this action should be mailed to: Commissioner of Patents and Trademarks, P.O. Box 1450, Alexandria, VA 22313-1450 or faxed to 571-273-8300. Hand delivered responses should be brought to the United States Patent and Trademark Office Customer Service Window: Randolph Building 401 Dulany Street, Alexandria, VA 22314.
/KIMBERLY L EVANS/Examiner, Art Unit 3629
/NATHAN C UBER/Supervisory Patent Examiner, Art Unit 3626