DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
The amendment filed on November 14, 2025 cancelled claims 11 and 13. Claims 1-10, 16, and 18 were amended and new claims 21-22 were added. Thus, the currently pending claims addressed below are claims 1-10, 12, and 14-22.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 1-10, 12, and 14-22 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Independent claims 1, 8 and 16 have been amended to include “a confidence score”, “the confidence score”, “associated confidence”, and “associated confidence scores”. The examiner has been unable to find support for these terms in the applicant’s disclosure. The term “confident” and/or confidence” does not appear in the applicant’s disclosure. The terms “score”, “scoring”, and/or “scored” are only disclosed in paragraph 47 in regards to an “F1 score” and paragraph 52 in regards to a similarity score. An F1 score is a machine learning metric, useful for imbalanced datasets, which balances precision (how accurate positive predictions are) and recall (how many actual positives are found) into a single score, representing the harmonic mean. Thus, the F1 score cannot be interpreted as the claimed “confidence score”. A cosign similarity score indicates semantic alignment by measuring the angle between text/data vectors, wherein the score is a value between 0 and 1, with 1 indicating the highest possible semantic similarity between the two text/data vectors. A confidence score is a derived metric that represents a probability-like certainty/trustworthiness of correctness or relevance which could be obtained by combining a cosine similarity with other factors such as softmax probabilities or GPT evaluation. The key differences between a cosign similarity score and a confidence score are that: a cosine similarity score measures similarity/closeness, whereas a confidence score measures certainty/trustworthiness; a cosine similarity score is a direct geometric calculation, whereas a confidence score is often a final output or a combination/transformation of metrics; and a high cosine similarity means semantically alike, whereas a high confidence means likely correct/relevant. Thus, it is clear that a cosign similarity score and confidence score are two different things, so the examiner cannot interpret the claimed “confidence score” as a similarity score. The only other possible support for a confidence score might be the “percentage match” and/or “percentage career fit to the user profile” as disclosed in paragraphs 8, 40 and 49, as well as, figure 6. However, this matching is an indication of how similar the output career is to the user’s profile, not an indication of how confident the machine learning model is with regards to the output career. As such, the disclosed “percentage match” and/or “percentage career fit to the user profile” are similarity scores and not confidence scores. Given that these are the only disclosures the examiner can find of a value that is generated and output to a user with regards to career options, it is clear that the applicant’s disclosure does not support the claimed “a confidence score”, “the confidence score”, “associated confidence”, and “associated confidence scores”. Thus, independent claims 1, 8 and 16 have been amended to include limitations that fail to comply with the written description requirement.
Additionally, the examiner has been unable to find support in the applicant’s specification for a “relevance to the user” with regards to a career that is different from a “similarity score” as recited in independent claims 8 and 16. While paragraph 49 and figure 6 indicate that along with the percentage fit of each career to the user profile, it may also output may additional information such as matching skills the user has, relevant projects the user has worked on, industries the user has worked in, and trainings the user has taken that are relevant to the career. Thus, these are not an indication of a careers “relevance to the user”, but instead relevant skills, experience, etc., that the user has that have a “relevance to the career”. The claims appear to indicate that both a confidence score in the career and the careers relevance to the user”, such that these are two different things. If the examiner were to interpret the “confidence score” to be the percentage fit of the career to the user profile disclosed in the specification, then there is no support for obtaining any other relevance of the career to the user. If the examiner were to interpret the relevance to the user as the percentage fit of the career to the user, then there is no support for an additional determination of a value that might be considered a confidence score. Given that the disclosure only supports a single value that indicates a similarity score, which can be considered a relevance of the career to the user and does not disclose two such values of relevance to the user, independent claims 8 and 16 have been amended to include subject matter that fails to comply with the written description requirement.
Dependent claims 2-7, 8-10, 12, 14-15 and 17-22 fail to correct the deficiencies of the claims from which they depend and, as such, are rejected by virtue of dependency.
For the purpose of prosecuting the claims the examiner is going to interpret the term “confidence” to be similarity, which means the term “confidence score(s)” will be interpreted as similarity score(s). Furthermore, since a similarity score is an indication of a relevance to the user, the phrase “confidence score(s) and relevance to the user” will be interpreted as “similarity score(s)”.
Claims 1-7 and 16-22 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Independent claims 1 and 16, as well as, dependent claim 3 have been amended to include the term normalizing in regards to data or data normalization. The examiner has been unable to find support for normalizing data or data normalization in the applicant’s disclosure. The terms “normalize”, “normalizing”, “normalized” and/or “normalization” are not present in the applicant’s specification. While the applicant’s disclosure supports a data preparation module that cleans, pre-processes, and standardizes data in at least paragraphs 34 and 57 of the specification, there is no disclosure of normalizing the data. In the context of processing data, data standardization and data normalization are two different things. Data standardization (Z-score scaling) centers data around a mean of 0 with a standard deviation of 1, making it ideal for algorithms sensitive to feature scale, while data normalization (Min-Max scaling) rescales data to a fixed range (like 0 to 1), useful when features have vastly different scales or distributions aren't normal, with the key difference being standardization's lack of fixed bounds versus normalization's bounded output. Thus, the disclosure of standardizing data cannot be said to support normalizing data or data normalization. Furthermore, the disclosure in paragraph 57 of sorting extracted key phrases from maximum to minimum cannot be said to support normalizing data or data normalization because it merely supports sorting the number of occurrences of the phrases but does not require rescaling the data to a fixed range. As such, it is clear that the applicant’s disclosure does not support the amended limitations of: “preprocess the heterogeneous raw user data by normalizing, cleaning, and structuring the data to generate unified user input data” in claim 1; “preprocessing comprises applying data normalization, deduplication, and outlier detection algorithms to the raw user data” in claim 3; and “preprocessing, normalizing, and integrating the heterogeneous raw user data to generate unified user input data” in claim 16. Thus, it is clear that claims 1, 3, and 16 as currently amended fail to comply with the written description requirement.
Furthermore, the above limitation of claim 3 includes the terms “deduplication” and “outlier detection”. The examiner has been unable to find support for these terms in the applicant’s disclosure. The terms “deduplication” and “outlier detection” are not used in the applicant’s specification. While paragraphs 34 and 57 support cleaning dataset and/or a data cleaning process, there is no support for the data cleaning of data using “deduplication” and/or “outlier detection”. It is true that “deduplication” and/or “outlier detection” are two common data cleaning tasks. Other common data cleaning tasks include like handling missing value, standardizing formats, correcting errors, validating data, and fixing structural issues. However, cleaning data does not inherently require that every one of these common data cleaning tasks be performed, as data can be said to be cleaned is only a single cleaning task was performed. Thus, the recitation of cleaning in the applicant’s disclosure cannot be said to support performing “deduplication” and/or “outlier detection” because cleaning data does not inherently require that either of these steps be performed. Instead, it would be obvious for cleaning to include “deduplication” and “outlier detection”. However, obviousness is insufficient proof of support with regards to 35 USC 112(a), instead the standard for support in a disclosure when the term itself, an equivalent term, and/or equivalent process are not recited is inherency (see MPEP 2163). As such, it appears that the only specific data cleaning process supported by the applicant’s disclosure is found in paragraph 45 where punctuations are removed and ensuring that all word are in the same casing. Thus, it is clear that the applicant’s disclosure does not support the amended limitation of “preprocessing comprises applying data normalization, deduplication, and outlier detection algorithms to the raw user data” in claim 3. Hence, it is clear that claim 3, as currently amended, fails to comply with the written description requirement.
Finally, dependent claims 2-7 and 17-22 fail to correct the deficiencies of the claims from which they depend and, as such, are rejected by virtue of dependency.
For the purpose of prosecuting the claims the examiner is going to interpret the term normalizing as if it means standardizing. The examiner is going to interpret the terms deduplication and outlier detection as if they merely mean cleaning the data in some manner such as extracting certain data from the raw data.
Claims 1-7 and 21 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Independent claim 1 has been amended to recite “feature vectors representing distinct user attributes”. The examiner has been unable to find support for feature vectors representing distinct user attributes in the applicant’s disclosure. While the applicant disclosure supports generating feature vectors from extracted content in paragraphs 34-35 and 46, there is no support in the applicant’s specification for the feature vectors which are generated to “represent distinct user attributes”. The applicant’s disclosure makes no mention of the term “attributes” much less, “user attributes” or “distinct user attributes”. However, it in no way can paragraphs 34-35 and 46 be said to support “distinct user attributes” because there is no disclosed requirement that the generated feature vectors be “distinct”, or be “user attributes”. As such, it is clear that independent claim 1 has been amended to recite subject matter that fails to comply with the written description requirement.
Dependent claims 2-7 and 21 fail to correct the deficiencies of the claims from which they depend and, as such, are rejected by virtue of dependency.
For the purpose of prosecuting the claims the examiner is going to interpret the limitation as if it recited “and generate a predetermined number of feature vectors”.
Claims 8-10, 12, and 14-20 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Independent claims 8 and 16 have been amended to include the term “integrating” with respect to the heterogeneous raw user data in addition to the preprocessing of the heterogenous raw user data to generate unified user input data. The examiner has been unable to find support for integrating the heterogeneous raw user data in addition to the preprocessing of said data to generate unified user input data. While the applicant’s disclosure supports preprocessing the raw user data to generate unified user input data, there is no disclosure of how the preprocessed data is also integrated to generate user input data. While the applicant’s disclosure supports asking a user questions and integrating this data into the platform in paragraphs 11-12 and 51, this would appear to be specific to data obtained from user responses to questions and said integrating is adding the responses to a resume to make the resume better, not something done on every type of received data, and not for the purpose of using raw data to generate a plurality of potential careers as required by the claims. Figure 6 and paragraphs 20 and 49-50 disclose integrating the machine learning algorithms with career predictions based on user data. However, this would cannot be considered integrating user data to generate unified input data because it is integration of the machine leaning algorithm not integration of the user data algorithms. Paragraph 57 disclose integrating data collection with various machine learning models and APIs. However, this is specific to the collections of data being able to be received from machine learning models and APIs, and not a step of integrating user data that has already been received to generate data. Given that the claims require the receipt of the raw user data before the claimed integrating step, this cannot be considered support for preprocessing and integrating the heterogeneous raw user data to generate unified user input data. As such, it is clear that claims 8 and 16 have been amended to include subject matter that fails to comply with the written description requirement.
Dependent claims 9-10, 12, 14-15 and 17-20 fail to correct the deficiencies of the claims from which they depend and, as such, are rejected by virtue of dependency.
For the purpose of prosecuting the claims the examiner is going to interpret the term integrating to mean storing. The applicant’s disclosure does not mention the term “unified user input data”, but does disclose in paragraph 38, that a user profile contains gathered data, and a user profile would be a unified user input data. However, there is no disclosure of an integration step occurring either prior to or after the preprocessing of data, and no disclosure of an integration step occurring before such data is part of the user profile. Thus, the only possible step that might be performed after the preprocessing of data and before the data is in the user profile is an inherent storing step. Thus, it would appear that the applicant intends the term integrating to mean storing.
Claims 8-10, 12, and 14-15 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Independent claim 8 has been amended to recite: “and generating a report comprising the at least one career, wherein the report is dynamically updated in response to changes in the user's profile and newly received data”. The examiner has been unable to find support for dynamically updating a report in response to changes in the user's profile and newly received data. While the applicant’s disclosure supports receiving new user data in real-time in paragraph 9, there is no support in the applicant’s specification for updating the report based on this newly received user data, much less, doing so dynamically. Instead, the applicant’s specification supports generating, by the machine learning model and based on the new user data and the unified user input data, a new plurality of potential careers for the user and associating each of the new plurality of potential careers with a similarity score; ranking the new plurality of potential careers based on the similarity score; predicting at least one career of the new plurality of careers for the user; and generating a new report comprising the at least one career of the new plurality of careers. As such, it is clear that claims 8 has been amended to include subject matter that fails to comply with the written description requirement.
Dependent claims 9-10, 12, and 14-15 fail to correct the deficiencies of the claims from which they depend and, as such, are rejected by virtue of dependency.
For the purpose of prosecuting the claims the examiner is going to interpret “and generating a report comprising the at least one career, wherein the report is dynamically updated in response to changes in the user's profile and newly received data” as if it recited “and generating a report comprising the at least one career” because claim 8 is directed to a method, and since the claim neither requires changes to the user profile to occur nor requires the receipt of new data, the phrase “wherein the report is dynamically updated in response to changes in the user's profile and newly received data” does not limit the scope of the claim as per MPEP 2111.04.
Claim 4 is rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Claim 4 has been amended to recite “The method of claim 1 wherein the natural language processing algorithms include tokenization, stemming, lemmatization, and named entity recognition”. The examiner has been unable to find support for the natural language processing algorithm of the applicant’s invention including tokenization, stemming, lemmatization, and named entity recognition. While the applicant’s disclosure supports extracting content using a natural language processing algorithm in at least paragraph 52 where it is disclosed that the applicant’s invention extracting keywords from a job description using spaCy, a free, open-source library for advanced Natural Language Processing (NLP) in Python, there is no support for the natural langue processing algorithm including tokenization, stemming, lemmatization, and named entity recognition. First, there is no support in the applicant’s disclosure for a natural langue processing algorithm using named entity recognition. Second, based on paragraph 57 of the applicant’s specification, it appears that the natural language processing algorithm of the applicant’s invention does not include tokenization, stemming, and lemmatization. Paragraph 57 specifically states that “NLP algorithms which utilize a basic process of data collection, stemming/lemmatization, generating a bag of words and corpus and finally feeding this into a model are not used”. There is no other disclosure of the terms stemming or lemmatization in the applicant’s disclosure. Thus, it is clear that claim 4, as currently amended, fails to comply with the written description requirement.
For the purpose of prosecuting the claims the examiner is going to interpret claim 4 as if it recited “The method of claim 1 wherein the natural language processing algorithms is a known natural language processing algorithm”.
Claim 5 is rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Claim 5 has been amended to recite “The method of claim 1, wherein the feature vectors are generated using term frequency-inverse document frequency (TF-IDF) and principal component analysis (PCA).”. The examiner has been unable to find support in the applicant’s disclosure for ranking generating feature vectors using both “frequency-inverse document frequency (TF-IDF) and principal component analysis (PCA)”. Paragraphs 45 and 57 of the applicant’s specification supports “wherein the feature vectors are generated using term frequency-inverse document frequency (TF-IDF)”. However, nowhere in the applicant’s disclosure is the phrase ”principal component analysis” or the term “PCA” mentioned. Furthermore, there is no disclosure of using principal component analysis (PCA) when generating feature vectors, much less, using both term frequency-inverse document frequency (TF-IDF) and principal component analysis (PCA) when generating feature vectors. Principal component analysis (PCA) is defined as an orthogonal linear transformation on a real inner product space that transforms the data to a new coordinate system such that the greatest variance by some scalar projection of the data comes to lie on the first coordinate (called the first principal component), the second greatest variance on the second coordinate, and so on. There is no support for any type of a transformation on a real inner product space in the applicant’s disclosure. Thus, it is clear that claim 5, as currently amended, fails to comply with the written description requirement.
For the purpose of prosecuting the claims the examiner is going to interpret claim 5 as if it recited ““The method of claim 1, wherein the feature vectors are generated using term frequency-inverse document frequency (TF-IDF)”.
Claim 7 is rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Claim 7 has been amended to recite “The method of claim 1, wherein the report further comprises a graphical visualization of the user's career fit scores across multiple industries.”. The examiner has been unable to find support in the applicant’s disclosure for a report that comprises a graphical visualization of the user's career fit scores across multiple industries. First, there is no support in the applicant’s disclosure for a report that further comprises a graphical visualization. Instead, paragraphs 53 and 81-82, as well as figure 10c-f can be said to support displaying the report in a graphical user interface, which is much different than a report further comprising a graphical visualization, much less a graphical visualization of the user's career fit scores across multiple industries. Thus, the applicant’s disclosure does not support a report further comprising a graphical visualization. Second, while it appears that relevant industries in the user’s profile can be included in the report, this appears to be industries in the user’s profile that are relevant to the recommended career. Thus, it does not support the user’s career fit scores across multiple industries. Finally, based on the applicant’s disclosure, the percentage career fit is a fit between the user profile and a recommended career. only the recommended career indicates a percentage fit to the user profile. Thus, the percentage career fit to the user profile is only associated with the recommended career. As such, the applicant’s disclosure does not support “the user's career fit scores across multiple industries” as claimed. Therefore, it is clear that claim 7, as currently amended, fails to comply with the written description requirement.
For the purpose of prosecuting the claims the examiner is going to interpret claim 7 as if it recited “The method of claim 1, wherein the report comprising the at least one predicted career and the associated similarity score is displayable in a graphical user interface”.
Claim 9 is rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Claim 9 has been amended to recite “The method of claim 8, wherein the ranking of the plurality of potential careers is performed using a weighted scoring algorithm based on user preferences and historical success rates”. The examiner has been unable to find support in the applicant’s disclosure for ranking careers using “a weighted scoring algorithm based on user preferences and historical success rates”. Paragraph 7 of the applicant’s specification supports “ranking the plurality of potential careers”. Paragraph 49 discloses outputting the percentage fit of the career to the user profile. Thus, the applicant’s specification could arguably be said to support ranking the careers based on the percentage fit of the career to the user profile (i.e., similarity score). However, the other sections of the applicant’s disclosure which discuss ranking resumes with regards to job descriptions such as the disclosure in paragraph 52 of the applicant’s specification. Based on paragraph 50 of the applicant’s specification we know that the original model had weights, because the weights are retrained. However, there is no disclosure in the applicant’s specification of a weighted scoring algorithm based on user preferences and historical success rates, much less ranking the careers based on such a weighted scoring algorithm. Thus, it is clear that claim 9, as currently amended, fails to comply with the written description requirement.
For the purpose of prosecuting the claims the examiner is going to interpret claim 9 as if it recited “The method of claim 8, wherein the ranking of the plurality of potential careers is performed using a cosine similarity”. While, even this interpretation is questionable with regard to support in the applicant’s specification, at least it appears that the percentage fit of a career to a user profile, disclosed in paragraph 49 is a similarity score, and the applicant’s ranking module (albeit for resume ranking not career ranking), disclosed in paragraph 52, appears to be able to rank based on cosine similarity score which inherently must be calculated using cosine similarity.
Claims 21-22 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Newly added claims 21-22 recite: “The method of claim 1, wherein the system automatically updates the career recommendations in response to changes in the user's online activity or profile data”; and “the system of claim 16, wherein hardware processor is configured to receive real-time updates from user devices and the plurality of data sources and automatically update the predicted at least one career recommendations in response”, respectively. The examiner has been unable to find support for automatically updating the career recommendations in response to changes in the user’s online activity or automatically updating the predicted at least one career recommendation in response to receiving updates from user devices and the plurality of data sources. The only mention of the terms “automatic” or “automatically” in the applicant’s disclosure are found in: paragraph 13 where a user selects a career and the system automatically generates likely career paths for the chosen career; paragraph 47 where the applicant admits that an XGBoost algorithm is known to automatically handle missing data values; and paragraph 55 where the invention automatically generates user summaries that are formatted into a resume. Thus, there is no support in the applicant’s disclosure for automatically updating the career recommendations as recited in claim 21, or automatically updating the predicted at least one career recommendation as recited in claim 22. Second, the only mention of the term “update”, “updating”, or “updated” in the applicant’s disclosure are found in paragraph 51 where it discloses that the invention updates the user’s resume. As such, there is no support in the applicant’s specification of updating the career recommendations or updating the predicted at least one career. Instead, the applicant’s specification supports generating, by the machine learning model and based on newly received user data and the unified user input data, a new plurality of potential careers for the user and associating each of the new plurality of potential careers with a similarity score; ranking the new plurality of potential careers based on the similarity score; predicting at least one career of the new plurality of careers for the user; and generating a new report comprising the at least one career of the new plurality of careers. As such, it is clear that claims 21-22 have been amended to include subject matter that fails to comply with the written description requirement.
Additionally, claim 22, as amended, recites “receive real-time updates from user devices and the plurality of data sources”. The examiner has been unable to find support for this limitation in the applicant’s disclosure. The only support in the applicant’s disclosure of the term “real-time” is found in: paragraph 9 of the applicant’s specification where changes to predictions from the machine learning model can occur in real-time, as the user profile and interest change; and original claim 2 where “the raw user data is gathered in real-time from user activities with mobile applications and internet activities and user profile”. Given the disclosure in original claim 2 the only information that the disclosure supports gathering in real-time is the raw user data “received from a mobile application” (applicant’s specification, paragraphs 9, 56-57); “scraped from the Internet” (applicant’s specification, paragraphs 32 and 34); and “changes to the user profile and interests made by the invention” (applicant’s specification, paragraphs 9, 34, and 38). Therefore, it is clear that the applicant’s disclosure does not support receiving in real-time updates from user devices, as the only received real-time information would be information from an application executing on a single device of the user, nor does it support receiving in real-time information from the plurality of data sources, as the information from the plurality of data sources would need to be scraped from the sources and not received from the sources. Thus, it is clear that claim 22 has been amended to include subject matter that fails to comply with the written description requirement.
For the purpose of prosecuting the claims the examiner is going to interpret claim 21 as if it recited: “The method of claim 1, wherein the machine learning model can generate a new at least one predicted career, if additional updated subject user data is received”; and claim 22 as if it recited “the system of claim 16, wherein hardware processor is configured to receive additional updated subject user data and generate a new at least one predicted career based on the updated subject user data and the additional updated subject user data”.
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 1-10, 12, and 14-22 are directed to methods and a system which would be classified under one of the listed statutory classifications (i.e., 2019 Revised Patent Subject Matter Eligibility Guidance (hereinafter “PEG”) “PEG” Step 1=Yes).
However, claims 1-10, 12, and 14-22 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. Independent claim(s) 1, 8, and 16 recite(s) the following abstract idea:
receiving heterogeneous raw user data from a plurality of sources, wherein the heterogenous raw user data comprises at least one of professional activities, non-professional activities, education, hobbies, skills, interests, and contemporaneous online user activities from multiple platforms and devices;
preprocessing, normalizing, and integrating the heterogeneous raw user data by normalizing, cleaning, and structuring the data to generate unified user input data;
extracting predefined content from the unified user input data using natural language processing algorithms;
transforming the extracted predefined content into frequency distribution;
generating a predetermined number of feature vectors representing distinct user attributes;
label encoding each feature vector to facilitate algorithmic model training;
generating a training data set and a test data set from the encoded feature vectors, wherein the data sets comprise integrated data from the plurality of sources;
training at least one algorithmic model using the training data set;
evaluating a performance of the at least one algorithmic model using the test data set;
inputting the user input data into the at least one algorithmic model;
generating, by the at least one algorithmic model as output, a ranked list of a plurality of potential careers for the user, wherein each of the a plurality of potential careers is associated with a confidence score and relevance to the user, and wherein the ranking is based on the confidence scores and relevance to the user;
predicting at least one career for the user;
generating a report comprising the at least one career, wherein the report is dynamically updated in response to changes in the user’s profile and newly received data;
receiving updated subject user data in real time;
predicting another at least one career for the user, using the trained algorithmic model on the updated subject user data, wherein the at least one predicted career is associated with a confidence score and relevance to the user; and
generating an updated report comprising the another at least one predicted career, wherein the report includes a ranked list of career options and associated confidence score.
The limitations as detailed above, as drafted, falls within the “Certain Method of Organizing Human Activity” grouping of abstract ideas namely managing personal behavior or relationships or interactions between people. Accordingly, the claim recites an abstract idea (i.e., “PEG” Revised Step 2A Prong One=Yes).
This judicial exception is not integrated into a practical application because the claim only recites the additional elements of:
a hardware processor (processing circuitry) and a memory device storing instructions (e.g., general purpose computer with generic computer component), and a machine learning model which is either a support vector machines, neural networks, or gradient boosting algorithms (generic machine learning models which are generic computer components as per the Recentive Analytics decision).
The following limitations, if removed from the abstract idea and considered additional elements, merely perform generic computer function of processing, storing, communicating (e.g., transmitting and receiving), and displaying data and, as such, are insignificant extra-solution activities (see MPEP 2016.05(d)(II) and MPEP 2106.05(g)):
receiving heterogeneous raw user data from a plurality of sources, wherein the heterogenous raw user data comprises at least one of professional activities, non-professional activities, education, hobbies, skills, interests, and contemporaneous online user activities from multiple platforms and devices (receiving data); and
receiving updated subject user data in real time (receiving data).
The additional technical elements above are recited at a high-level of generality (i.e., as one or more generic processors performing generic computer functions of processing, communicating (e.g., transmitting and receiving), and displaying) such that it amounts to no more than mere instructions to apply the exception using one or more general-purpose computers and/or one or more generic computer components. The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional technical elements above do not integrate the abstract idea/judicial exception into a practical application because it does not impose any meaningful limits on practicing the abstract idea. More specifically, the additional elements fail to include (1) improvements to the functioning of a computer or to any other technology or technical field (see MPEP 2106.05(a)), (2) applying or using a judicial exception to effect a particular treatment or prophylaxis for a disease or medical condition (see Vanda memo), (3) applying the judicial exception with, or by use of, a particular machine (see MPEP 2106.05(b)), (4) effecting a transformation or reduction of a particular article to a different state or thing (see MPEP 2106.05(c)), or (5) applying or using the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is more than a drafting effort designed to monopolize the exception (see MPEP 2106.05(e) and Vanda memo).
Rather, the limitations merely add the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea (see MPEP 2106.05(f)), or generally link the use of the judicial exception to a particular technological environment or field of use (see MPEP 2106.05(h)).
Thus, the claim is “directed to” an abstract idea (i.e., “PEG” Revised Step 2A Prong Two=Yes)
When considering Step 2B of the Alice/Mayo test, the claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the claims do not amount to significantly more than the abstract idea.
More specifically, as discussed above with respect to integration of the abstract idea into a practical application, the additional elements of using a hardware processor (processing circuitry) and a memory device storing instructions (e.g., general purpose computer with generic computer component) and executing software (e.g., a machine learning model which is either a support vector machine, neural network, or gradient boosting algorithm) to perform the claimed functions amounts to no more than mere instructions to apply the exception using one or more general-purpose computers and/or one or more generic computer components.
“Generic computer implementation” is insufficient to transform a patent-ineligible abstract idea into a patent-eligible invention (See Affinity Labs, _F.3d_, 120 U.S.P.Q.2d 1201 (Fed. Cir. 2016), citing Alice, 134 S. Ct. at 2352, 2357) and more generally, “simply appending conventional steps specified at a high level of generality” to an abstract idea does not make that idea patentable (See Affinity Labs, _F.3d_, 120 U.S.P.Q.2d 1201 (Fed. Cir. 2016), citing Mayo, 132 S. Ct. at 1300). Moreover, “the use of generic computer elements like a microprocessor or user interface do not alone transform an otherwise abstract idea into patent-eligible subject matter (See FairWarning, 120 U.S.P.Q.2d. 1293, citing DDR Holdings, LLC v. Hotels.com, L.P., 773 F.3d 1245, 1256 (Fed. Cir. 2014)). As such, the additional elements of the claim do not add a meaningful limitation to the abstract idea because they would be generic computer functions in any computer implementation. Thus, taken alone, the additional elements do not amount to significantly more than the above-identified judicial exception (the abstract idea). Looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements taken individually. There is no indication that the combination of elements improves the functioning of the computer or improves any other technology. Their collective functions merely provide generic computer implementation.
The Examiner notes simply implementing an abstract concept on a computer, without meaningful limitations to that concept, does not transform a patent-ineligible claim into a patent-eligible one (See Accenture, 728 F.3d 1336, 108 U.S.P.Q.2d 1173 (Fed. Cir. 2013), citing Bancorp, 687 F.3d at 1280), limiting the application of an abstract idea to one field of use does not necessarily guard against preempting all uses of the abstract idea (See Accenture, 728 F.3d 1336, 108 U.S.P.Q.2d 1173 (Fed. Cir. 2013), citing Bilski, 130 S. Ct. at 3231), and further the prohibition against patenting an abstract principle “cannot be circumvented by attempting to limit the use of the [principle] to a particular technological environment” (See Accenture, 728 F.3d 1336, 108 U.S.P.Q.2d 1173 (Fed. Cir. 2013), citing Flook, 437 U.S. at 584), and finally merely limiting the field of use of the abstract idea to a particular existing technological environment does not render the claims any less abstract (See Affinity Labs, _F.3d_, 120 U.S.P.Q.2d 1201 (Fed. Cir. 2016), citing Alice, 134 S. Ct. at 2358; Mayo, 132 S. Ct. at 1294; Bilski v. Kappos, 561 U.S. 593, 612 (2010); Content Extraction & Transmission LLC v. Wells Fargo Bank, Nat’l Ass’n, 776 F.3d 1343, 1348 (Fed. Cir. 2014); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355 (Fed. Cir. 2014).
Applicant herein only requires one or more general-purpose computers and/or one or more generic computer components (as evidenced from Figures 1 and 11, as well as, paragraphs 30 and 66-71 of the applicant’s specification; Tavares (PGPUB: 2019/0266501 which discloses in at least paragraph 39 that gradient boosting algorithms, support vector machines, and neural networks were conventional machine learning algorithms by at least 2018; and the Recentive Analytics v. Fox Corp decision which states generic machine learning models are just software executing on a computer and, as such, are incapable of transforming an abstract idea into a practical application under Step 2a, Prong 2 and incapable of being considered significantly more under Step 2b); therefore, there does not appear to be any alteration or modification to the generic activities indicated, and they are also therefore recognized as insignificant activity with respect to eligibility. Finally, the following limitations, if removed from the abstract idea and considered additional elements, would be considered insignificant extra solution activity as they are directed to merely receiving, displaying, storing, and/or transmitting data (see MPEP 2016.05(d)(II) and MPEP 2106.05(g)):
receiving heterogeneous raw user data from a plurality of sources, wherein the heterogenous raw user data comprises at least one of professional activities, non-professional activities, education, hobbies, skills, interests, and contemporaneous online user activities from multiple platforms and devices (receiving data); and
receiving updated subject user data in real time (receiving data).
Thus, taken individually and in combination, the additional elements do not amount to significantly more than the above-identified judicial exception (the abstract idea) (i.e., “PEG” Step 2B=No).
The dependent claims 2-7, 9-10, 12, 14-15, and 17-22 appear to merely further limit the abstract idea by further limiting the heterogeneous raw user data which is considered part of the abstract idea (Claim 2); further limiting the preprocessing which is considered part of the abstract ides (Claim 3); further limiting the natural language processing algorithms which is considered part of the abstract idea (Claim 4); further limiting the feature vectors and how they were generated (Claim 5) further limiting the additional element of the machine learning to indicate the type of well-known machine learning models it might be which has been already been addressed in the rejection above (Claim 6); further limiting the report which is considered part of the abstract idea (Claim 7); further limiting the ranking of the plurality of potential careers which is considered part of the abstract idea (Claim 9); further limiting the algorithmic model to include parameters and adding a step of evaluating for accuracy which are both considered part of the abstract idea (Claims 10 and 19-20); adding additional steps of receiving a user resume and suggesting content for inclusion on the resume which are both considered part of the abstract idea (Claim 12); adding an additional step of matching the user to available job opportunities which is considered part of the abstract idea (Claim 14); and adding an additional step of generating a dialogue area and transmitting it for display on a user interface which are both considered part of the abstract idea (Claim 15); adding an additional step of autogenerating at least a portion of a resume which is considered part of the abstract idea (Claim 17); further limiting the predicting which is considered part of the abstract idea (Claim 18); adding an additional step of automatically updating the career recommendations which is considered part of the abstract idea (Claim 21) and adding the additional steps of receiving real-time updates and automatically updating the predicted at least one career recommendations which are both considered part of the abstract idea (Claim 22); , and therefore only further limit the abstract idea (i.e. “PEG” Revised Step 2A Prong One=Yes), does/do not include any new additional elements, which have not been addressed above, that are sufficient to amount to significantly more than the judicial exception, and as such are “directed to” said abstract idea (i.e. “PEG” Step 2A Prong Two=Yes); and do not add significantly more than the idea (i.e. “PEG” Step 2B=No)..
Thus, based on the detailed analysis above, claims 1-10, 12, and 14-22 are not patent eligible.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1-4, 6-8, 15-16, 18 and 21-22 is/are rejected under 35 U.S.C. 103 as being unpatentable over Xie et al. (PGPUB: 2019/0303798) in view of Shashkina et al. (Data preparation for machine learning: a step-by-step guide, April 10, 2023, https://itrexgroup.com/blog/data-preparation-for-machine-learning/, pages 1-13; in further view of Liu et al. (PGPUB: 2014/0279263). (Examiner note: Crossed out claim text are limitations not supported by the applicant’s disclosure. Underlined claim text is the examiner’s interpretation of the claim for the purpose of prosecuting the claims. Please see 35 USC 112(a) rejections above.)
Claims 1, 8, and 16: Xie discloses a method for recommending a career to a user, a method of mapping a career of a user, and a system for recommending a career to a user, comprising:
a hardware processor (processing circuitry) and a memory device on which instructions are encoded to cause the hardware processor (Paragraphs 74-75) to perform the operations of:
receiving heterogeneous raw user data from a plurality of sources, wherein the heterogenous raw user data comprises at least one of professional activities, non-professional activities, education, hobbies, skills, interests, and contemporaneous online user activities from multiple platforms and devices (Paragraph 46: during a runtime stage, an ingestion platform obtains candidate data about a particular user from the profile database, the social graph database, and/or the member activity and behavior database; Paragraph 38: profile database includes member profile data and profile data for various organizations including name, age (e.g., birthdate), gender, interests, contact information, home town, address, spouse's and/or family members' names, educational background (e.g., schools, majors, matriculation and/or graduation dates, etc.), employment history, skills, professional organizations, and so on; Paragraph 39: social graph database includes various associations and relationships that the members establish with other members, or with other entities and objects; Paragraph 40: member activity and behavior database includes the members' interactions and behavior (e.g., content viewed, links or buttons selected, messages responded to, etc.); Paragraph 18: profile and/or usage data includes job title, seniority, industry, major, degree, job start date, job send date, years after graduation, etc.; Paragraphs 33-34, 43-44 and 72: the user data may come third-party applications and/or third-party websites executing on third-party servers);
preprocessing, standardizing, and storing the heterogeneous raw user data by , standardizing and structuring the data to generate unified user input data (e.g., a user profile); extracting predefined content from the unified user input data using natural language processing algorithms; transforming the extracted predefined content into frequency distribution and generating a predetermined number of feature vectors
While Xie discloses:
processing, standardizing, structuring and storing the heterogeneous raw user data to generate unified user input data in at least paragraphs 18, 38, 51, and 54. (Paragraph 18: profile and/or usage data is examined to locate features relevant to career path determinations; examples of such features may be, for example, job title, seniority, industry, major, degree, job start date, job send date, years after graduation, etc.; conditional probability may be applied to estimate the expected time spent at a particular position and the probability of moving to the next position; Paragraph 38: the profile data may be processed (e.g., in the background or offline) to generate various derived profile data. For example, if a member has provided information about various job titles that the member has held with the same organization or different organizations, and for how long, this information can be used to infer or derive a member profile attribute indicating the member's overall seniority level, or seniority level within a particular organization;
extracting predefined content from the unified user input data in at least paragraphs 51 and 54 (Xie - Paragraphs 51 and 54: a feature extractor acts to extract curated features from the candidate user profiles and sample member and activity and behavior information; an extracted feature may be a portion of the member profile, member usage and/or activity data, or social graph; the feature may also be a variable calculated from a portion of the data, such as an average, sum, difference, measurement, and the like);
transforming the extracted predefined content into frequency distributions and generating a predetermined number of feature vectors Xie - Paragraph 20: a set of tables is then created, with each table corresponding to a different current position, and the entries in each table identifying future positions for the current position based on the profile information in the training data; in some example embodiments, only future positions that occurred in the training data more than a threshold number of times will be included in the table; Paragraph 21: each of the subsequent positions listed in a position table may include a count indicating the number of profiles in the training data that listed the subsequent position and also listed the position that corresponds to the position table; Paragraph 22: the position tables may then be used to construct a career path graph; the career path graph contains positions as nodes, with edges between nodes being created if there exists an entry for the position corresponding to the node on one side of the edge in the position table corresponding to a position corresponding to the node on the other side of the edge; Paragraph 47: the training data may then be passed to a career path graph formation component that creates a career path graph with positions as nodes and edges between nodes indicating transitions between positions in one or more user profiles in the training data; in some cases the edges may only exist if the transition count is greater than a predetermined threshold; the career path graph is then sent to a clustering component, which then clusters the nodes in the graph to identify a plurality of potential career paths; Paragraph 50: the career path model component creates feature vectors from the raw information);
labeling the training data (Xie – Paragraph 19: the training data having been labeled to identify particular career paths (e.g., academic, corporate, computer-related professional, etc.; Paragraph 24: The machine learning algorithm is able to learn these different weights based on the training data and the labels; and Paragraph 57: publicly available messaging information (or information that is otherwise not private, such as information a user has voluntarily opted-out of privacy detection for), can be used as training data, which can then be labeled and submitted to the second machine learning algorithm to train the model. With the labeled data, the model may be built via a supervised learning approach that will help the user intention detector detect what topic a user wants to talk about); and
training a machine learning model in at least paragraphs 50, 24, (Xie - Paragraph 50: the career path model component creates feature vectors from the raw information; Paragraph 24: each of these career paths, no matter the granularity at which they are defined, may be input into the first machine learning algorithm along with the training data in order to learn the weights for the particular career paths; in some respects, therefore, the career path model may actually be thought of as separate career path models, one for each of the identified possible career paths; the machine learning algorithm is able to learn these different weights based on the training data and the labels)
Xie does not disclose: cleaning the heterogeneous raw user data; natural language processing is used to extract the content; label encoding each feature vector to facilitate machine learning model training; generating a training data set and a test dataset from the encoded feature vectors, wherein the data sets comprise integrated data from the plurality of sources; training at least one machine learning model using the training data set and evaluating a performance of the at least one machine learning model using the test data set.
However, the analogous art of Shashkina discloses that it is known to:
clean the heterogeneous raw user data (Shashkina - Page 4, lines 10-13: Structured data can be easily organized into tables and cleaned via deduplication, filling in missing values, or standardizing data formats; Page 4, line 17 through Page 8, line 12: when preparing data for machine learning, a large dataset requires sampling the data to select a subset of the data to train the model due to computational limitation; the quality of collected data is crucial as well; once you know what makes good data how to collect it an where to find it is also important; strategies for collecting data include obtaining it from internal sources, obtaining it from external sources, web scraping; once collected should such data be insufficient to satisfy all data points techniques such as data augmentations, active learning, transfer learning, and collaborative data sharing can be performed; the next step to prepare for machine learning is data cleaning which involves: finding and correcting errors, inconsistencies and missing values by using imputations, interpolation and/or deletion; handling outliers by removing, transforming them to reduce their impact, winsorizing, or treating them as a separate class of data; removing duplicates using techniques such as exact matching, fuzzy matching, hashing, or record linkage; handling incorrect data using techniques such as data transformation or removal of data points; handling imbalanced data using techniques such as resampling, synthetic data generation, cost-sensitive learning, and ensemble learning );
use natural language processing algorithms to extract the content (Shashkina - Page 4, lines 14-16: extracting relevant features from unstructured data using natural language processing);
label encode each feature vector to facilitate machine learning model training (Shashkina - Page 9, lines 16-22: converting categorical data into a numerical format using label encoding);
generate a training data set and a test data set from the encoded feature vectors, wherein the data sets comprise integrated data from plurality of sources (Shashkina - Page 11, line 7 through Page 13, line 13: performing data splitting on the data by dividing all the gathered data including training, validation, and testing datasets); and
train at least one machine learning model using the training data set and evaluating a performance of the at least one machine learning model using the test data set (Shashkina - Page 11, line 7 through Page 13, line 13: using the testing dataset to evaluate the performance of the trained model)
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the invention, to modify the invention of Xie to include clean the heterogeneous raw user data; use natural language processing algorithms to extract the content; label encode each feature vector to facilitate machine learning model training; generate a training data set and a test data set from the encoded feature vectors, wherein the data sets comprise integrated data from plurality of sources; and train at least one machine learning model using the training data set and evaluating a performance of the at least one machine learning model using the test data set as disclosed by Shashkina.
The motivation for doing so is to generate more reliable and accurate predictions (Shashkina - Page 2, lines 8-11; Page 8, lines 20-23; and Page 12, lines 3-7).
generating a trained machine learning model (Xie - Paragraph 24: each of these career paths, no matter the granularity at which they are defined, may be input into the first machine learning algorithm along with the training data in order to learn the weights for the particular career paths; in some respects, therefore, the career path model may actually be thought of as separate career path models, one for each of the identified possible career paths; the machine learning algorithm is able to learn these different weights based on the training data and the labels)
using the unified user input data as input to the trained machine learning model (Xie - Paragraph 54: the curated features are then used as input to the career path models, which output scores indicating the likelihood that the candidate user will follow each of the corresponding potential career paths);
generating, using the trained machine learning model a plurality of potential careers for the user and associating each of the plurality of potential careers with a similarity score to the user (Xie – Paragraph 19: the career path model is therefore trained by the first machine learning algorithm to receive an input candidate's profile and/or usage data and output a score for one or more potential career paths for the candidate; the score may reflect the probability that the candidate will progress down a particular career path; Paragraph 26: a score may then be calculated for that career path that may be different than a score calculated for a different career path for the same use; a score between 0 and 1 may be assigned, with 1 reflecting the highest probability and 0 the lowest, and in this example a score of 0.55 may be output for this user with respect to the career path including “project manager” and “vice president” (i.e., a corporate computer-related professional career path), whereas a score of 0.12 may be output for this user with respect to the career path including “assistant professor” and “professor” (i.e., an academic career path); Paragraph 27: the scores calculated for each potential career path may then be used by a recommendation engine to recommend one or more activities for a user to engage in in order to improve chances of a successful transition to another position in the potential career path; Paragraph 54: the curated features are then used as input to the career path models, which output scores indicating the likelihood that the candidate user will follow each of the corresponding potential career paths);
ranking the plurality of potential careers based on the similarity scores;
While Xie and Shashkina disclose assigning similarity scores to each of the plurality of potential careers in at least paragraph 26 of Xie, the no not specifically state that the similarity scores are ranked.
However, the analogous art of Liu discloses that it is known to rank recommendations to a user based on similarity scores in at least paragraph 8.
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the invention to modify Xie and Shashkina to that the recommendations are ranked by the similarity score as disclosed by Liu.
The rationale for doing so is that it would be obvious to try as there are a limited number of ways in which to provide recommendations that include a similarity score and ranking said scores is one such predictable ways to provide such recommendations.
predicting at least one career for the user (Xie - Paragraph 27: the scores calculated for each potential career path may then be used by a recommendation engine to recommend one or more activities for a user to engage in in order to improve chances of a successful transition to another position in the potential career path; Paragraph 48: the one or more career path models can provide predictions at runtime based on candidate data that indicates the likelihood that the corresponding candidate can proceed along the corresponding career path );
generating a report comprising the at least one career,Xie - Paragraph 48: the predictions may also be used directly by a user interface server component to display the predictions; Paragraph 49: predictions and/or recommendations may be passed to the user interface server component to present those predictions and/or recommendations to the user; Paragraph 55: a screen may be presented to a user, via a graphical user interface, displaying potential career paths for the user. Upon selection of one of these potential career paths, the recommendations for the user to increase his or her chances for progressing down the career path may be presented to the user);
receiving updated subject user data in real time (Xie - Paragraph 40: member activity and behavior database includes the members' interactions and behavior (e.g., content viewed, links or buttons selected, messages responded to, etc.; Paragraph 42: member activity with the social networking service occur on mobile applications executing on a mobile device; Paragraphs 26, 46, 49: the ingestion platform can obtain data for use in predicting careers at runtime; thus, each time the user requests recommendation it is based on the current raw user data);
predicting at least one career, using the trained machine learning model on the updated subject user data and the unified user data, wherein the at least one predicted career is associated with a similarity score Paragraph 27: the scores calculated for each potential career path may then be used by a recommendation engine to recommend one or more activities for a user to engage in in order to improve chances of a successful transition to another position in the potential career path; Paragraph 48: the one or more career path models can provide predictions at runtime based on candidate data that indicates the likelihood that the corresponding candidate can proceed along the corresponding career path); and
generating an new report comprising the at least one predicted career, the report includes a ranked list of career options and associated similarity score Xie - Paragraph 48: the predictions may also be used directly by a user interface server component to display the predictions; Paragraph 49: predictions and/or recommendations may be passed to the user interface server component to present those predictions and/or recommendations to the user; Paragraph 55: a screen may be presented to a user, via a graphical user interface, displaying potential career paths for the user. Upon selection of one of these potential career paths, the recommendations for the user to increase his or her chances for progressing down the career path may be presented to the user; and Liu – Paragraph 8: ranking the recommendations based on the similarity score).
Claim 2: Xie, Shashkina, and Liu disclose the method of claim 1, wherein the heterogeneous raw user data is from at least one of a mobile application, a web application, and third-party data sources via secure APIs. (Xie - Paragraph 42: member activity with the social networking service occurs on mobile applications executing on a mobile device)
Claim 3: Xie, Shashkina, and Liu disclose the method of claim 1, wherein preprocessing comprises applying data cleaning Shashkina - Page 4, lines 10-13: Structured data can be easily organized into tables and cleaned via deduplication, filling in missing values, or standardizing data formats; Page 4, line 17 through Page 8, line 12: when preparing data for machine learning, a large dataset requires sampling the data to select a subset of the data to train the model due to computational limitation; the quality of collected data is crucial as well; once you know what makes good data how to collect it an where to find it is also important; strategies for collecting data include obtaining it from internal sources, obtaining it from external sources, web scraping; once collected should such data be insufficient to satisfy all data points techniques such as data augmentations, active learning, transfer learning, and collaborative data sharing can be performed; the next step to prepare for machine learning is data cleaning which involves: finding and correcting errors, inconsistencies and missing values by using imputations, interpolation and/or deletion; handling outliers by removing, transforming them to reduce their impact, winsorizing, or treating them as a separate class of data; removing duplicates using techniques such as exact matching, fuzzy matching, hashing, or record linkage; handling incorrect data using techniques such as data transformation or removal of data points; handling imbalanced data using techniques such as resampling, synthetic data generation, cost-sensitive learning, and ensemble learning )
Claim 4: Xie, Shashkina, and Liu disclose the method of claim 1, wherein the natural language processing algorithms is a known natural language processing algorithm, (Shashkina - Page 4, lines 14-16: extracting relevant features from unstructured data using natural language processing)
Claim 6: Xie, Shashkina, and Liu disclose the method of claim 1, wherein theXie – Paragraphs 23 and 52)
Claim 7: Xie, Shashkina, and Liu disclose the method of claim 1, wherein the report comprising the at least one predicted career and the associated similarity score is displayable in a graphical user interfaceXie - Paragraph 26, 53-54: the score is output to the user; and Paragraph 48: the predictions are displayed to a user on a user interface; Paragraph 49: predictions and/or recommendations may be passed to the user interface server component to present those predictions and/or recommendations to the user; Paragraph 55: a screen may be presented to a user, via a graphical user interface, displaying potential career paths for the user. Upon selection of one of these potential career paths, the recommendations for the user to increase his or her chances for progressing down the career path may be presented to the user)
Claim 15: Xie, Shashkina, and Liu disclose the method of claim 8, wherein the memory device comprises instructions executable by the processing circuitry to generate a dialog area on a user interface for interacting with the user by posing questions to capture the user’s skills, interests and achievements. (Xie – Paragraph 38: when a person initially registers to become a member of the social networking service, the person will be prompted to provide some personal information, such as his or her name, age (e.g., birthdate), gender, interests, contact information, home town, address, spouse's and/or family members' names, educational background (e.g., schools, majors, matriculation and/or graduation dates, etc.), employment history, skills, professional organizations, and so on)
Claim 18: Xie, Shashkina, and Liu disclose the system of claim 16, wherein the trained model predicts at least one career based on at least one of the user data, the at least one skill, the at least one related industry, wherein the at least one suitable career is predicted dynamically in response to changes in the user profile, activities and interest. (Xie – Paragraph 66: a candidate user profile is submitted into a given career path model, outputting a baseline prediction indicating a likelihood of a user corresponding to the candidate user profile progressing down a career path corresponding to the given career path model; then the candidate user profile is altered and the altered candidate user profile is resubmitted to the given career path model to obtain a revised prediction; the revised prediction is compared to the baseline prediction from the given career path model; in response to a determination that the difference between the revised prediction to the baseline prediction exceeds a predetermined threshold, an activity related to the altered portion of the candidate user profile is recommended to the user corresponding to the candidate user profile)
Claim 21: Xie, Shashkina, and Liu disclose the method of claim 1, wherein the machine learning model can generate a new at least one predicted career, if additional updated subject user data is received Xie - Paragraph 40: member activity and behavior database includes the members' interactions and behavior (e.g., content viewed, links or buttons selected, messages responded to, etc.; Paragraph 42: member activity with the social networking service occur on mobile applications executing on a mobile device; Paragraphs 26, 46, 49: the ingestion platform can obtain data for use in predicting careers at runtime; thus, each time the user requests recommendation it is based on the current raw user data; Paragraph 27: the scores calculated for each potential career path may then be used by a recommendation engine to recommend one or more activities for a user to engage in in order to improve chances of a successful transition to another position in the potential career path; Paragraph 48: the one or more career path models can provide predictions at runtime based on candidate data that indicates the likelihood that the corresponding candidate can proceed along the corresponding career path)
Claim 22: Xie, Shashkina, and Liu disclose the system of claim 16, wherein hardware processor is configured to receive additional updated subject user data and generate a new at least one predicted career based on the updated subject user data and the additional updated subject user data Xie - Paragraph 40: member activity and behavior database includes the members' interactions and behavior (e.g., content viewed, links or buttons selected, messages responded to, etc.; Paragraph 42: member activity with the social networking service occur on mobile applications executing on a mobile device; Paragraphs 26, 46, 49: the ingestion platform can obtain data for use in predicting careers at runtime; thus, each time the user requests recommendation it is based on the current raw user data; Paragraph 27: the scores calculated for each potential career path may then be used by a recommendation engine to recommend one or more activities for a user to engage in in order to improve chances of a successful transition to another position in the potential career path; Paragraph 48: the one or more career path models can provide predictions at runtime based on candidate data that indicates the likelihood that the corresponding candidate can proceed along the corresponding career path)
Claim(s) 5 is/are rejected under 35 U.S.C. 103 as being unpatentable over Xie et al. (PGPUB: 2019/0303798) in view of Shashkina et al. (Data preparation for machine learning: a step-by-step guide, April 10, 2023, https://itrexgroup.com/blog/data-preparation-for-machine-learning/, pages 1-13) in view of Liu et al. (PGPUB: 2014/0279263) in further view of Inamdar et al. (PGPUB: 2021/0097471).
Claim 5: Xie, Shashkina, and Liu disclose the method of claim 1, wherein the
Xie, Shashkina, and Liu disclose the method of claim 1, including generating feature vectors using term frequency in the cited sections above.
Xie, Shashkina, and Liu do not disclose that term frequency-inverse document frequency (TF-IDF).
However, the analogous art of Inamdar discloses that it is known to generate feature vectors from heterogenous raw data by weighting features using term frequency-inverse document frequency (TF-IDF) in at least paragraph 30.
It would have been obvious to one of ordinary skill in the art to modify the invention of Xie, Shashkina, and Liu to use term frequency-inverse document frequency (TF-IDF) when generating the feature vectors.
The motivation for doing so is to accord more weight per occurrence to words that occur relatively infrequently. (Inamdar – Paragraph 30)
Claim(s) 9 is/are rejected under 35 U.S.C. 103 as being unpatentable over Xie et al. (PGPUB: 2019/0303798) in view of Shashkina et al. (Data preparation for machine learning: a step-by-step guide, April 10, 2023, https://itrexgroup.com/blog/data-preparation-for-machine-learning/, pages 1-13) in view of Liu et al. (PGPUB: 2014/0279263) in further view of Al Jadda et al. (PGPUB: 2021/0073891).
Claim 9: Xie, Shashkina, and Liu disclose the method of claim 8, wherein the ranking of the plurality of potential careers is performed using a cosine similarity.
Xie, Shashkina, and Liu disclose the method of claim 8, including ranking of the plurality of potential careers based on the similarity score. The similarity scores of Xie, Shashkina, and Liu are between 0 and 1 and, as such, are not cosine similarity scores.
However, the analogous art of Al Jadda discloses that it is known to use cosine similarity scores when providing recommendations in at least paragraph 9.
It would have been obvious to one of ordinary skill in the art to modify the invention of Xie, Shashkina, and Liu to use cosine similarity scores as disclosed by Al Jadda.
The rationale for doing so is that it would be obvious to try. There are a limited number of predictable ways in which a similarity score may be calculated and a cosine similarity score is one such predictable type of similarity score.
Claim(s) 12 and 17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Xie et al. (PGPUB: 2019/0303798) in view of Shashkina et al. (Data preparation for machine learning: a step-by-step guide, April 10, 2023, https://itrexgroup.com/blog/data-preparation-for-machine-learning/, pages 1-13) in view of Liu et al. (PGPUB: 2014/0279263) in further view of Ochs et al. (PGPUB: 2025/0005294).
Claims 12 and 17: The method of claim 8 and the system of claim 16, wherein the memory device comprises instructions executable by the processing circuitry to receive a user resume, and wherein the trained machine learning model autogenerates suggestions for content for inclusion in at least one portion of the resume based on the predicted at least one career.
Xie, Shashkina, and Liu disclose the method of claim 4, the method of claim 8, and the system of claim 16.
Xie, Shashkina, and Liu do not disclose receiving a user resume and wherein the trained machine learning model autogenerates suggestions for content for inclusion in at least one portion of the resume based on the predicted at least one career.
However, the analogous art of Ochs discloses that it is known to receive a user resume, and wherein the trained machine learning model autogenerates suggestions for content for inclusion in at least one portion of the resume based on the predicted at least one career in at least paragraphs 18-19, 36, 40, 74, 90-92 and 131.
It would have been obvious to one of ordinary skill in the art to modify the invention of Xie, Shashkina, and Liu to include receiving a user resume and wherein the trained machine learning model autogenerates suggestions for content for inclusion in at least one portion of the resume based on the predicted at least one career as disclosed by Ochs.
The rationale for doing so is that it merely requires combining prior art elements according to known methods to yield predictable results. It can be seen that each element claimed is taught by either the combination of Xie, Shashkina, and Liu or Ochs. Receiving a user resume and wherein the trained machine learning model autogenerates suggestions for content for inclusion in at least one portion of the resume based on the predicted at least one career (taught by Ochs) does not change nor effect the normal functions of predicting a career as taught by Xie, Shashkina, and Liu. Since the functionalities of the elements in the combination of Xie, Shashkina, and Liu and Ochs do not interfere with each other the results of the combination would be predictable.
Claim(s) 14 is/are rejected under 35 U.S.C. 103 as being unpatentable over Xie et al. (PGPUB: 2019/0303798) in view of Shashkina et al. (Data preparation for machine learning: a step-by-step guide, April 10, 2023, https://itrexgroup.com/blog/data-preparation-for-machine-learning/, pages 1-13) in view of Liu et al. (PGPUB: 2014/0279263) in further view of Gibbs (PGPUB: 2017/0323270).
Claim 14: The method of claim 8, wherein the memory device comprises instructions executable by the processing circuitry to match the user to available job opportunities based on the predicted at least one career.
Xie, Shashkina, and Liu disclose the method of claim 8.
Xie, Shashkina, and Liu do not disclose matching the user to available job opportunities based on the predicted at least one career.
However, the analogous art of Gibbs discloses that it is known to match a user to available job opportunities based on the at least one predicted career in at least paragraphs 21 and 55.
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the invention, to modify the invention of Xie, Shashkina, and Liu to include matching the user to available job opportunities based on the predicted at least one career.
The rationale for doing so is that it merely requires combining prior art elements according to known methods to yield predictable results. It can be seen that each element claimed is taught by either the combination of Xie, Shashkina, and Liu or Gibbs. matching the user to available job opportunities based on the predicted at least one career (taught by Gibbs) does not change nor effect the normal functions of predicting a career as taught by Xie, Shashkina, and Liu. Since the functionalities of the elements in the combination of Xie, Shashkina, and Liu and Gibbs do not interfere with each other the results of the combination would be predictable.
Claim(s) 10 and 19-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Xie et al. (PGPUB: 2019/0303798) in view of Shashkina et al. (Data preparation for machine learning: a step-by-step guide, April 10, 2023, https://itrexgroup.com/blog/data-preparation-for-machine-learning/, pages 1-13) in further view of Liu et al. (PGPUB: 2014/0279263) in further view of Otten (F1 Score The Ultimate Guide: Formulas, Explanations, Examples, Advantages, Disadvantages, Alternatives, and Python code, May 8, 2023, https:// spotintelligence.com/2023/05/08/f1-score/#:~:text=However%2C%20as%20a%20 general%20rule,false%20positives%20and%20false%20negatives, pages 1-22).
Claims 10 and 19-20: The method of claim 8 and the system of claim 18, wherein the trained machine learning model comprises parameters including precision, recall and an F1 score functionality of about 80%, wherein the trained machine learning model is evaluated for accuracy based on the F1 score;
Xie, Shashkina, and Liu disclose the method of claim 8 and the system of claim 18, wherein the trained machine learning model uses a test dataset to assess the accuracy of the model.
Xie, Shashkina, and Liu do not disclose that the trained model comprises parameters including precision, recall and an F1 score functionality of about 80%, wherein the trained machine learning model is evaluated for accuracy based on the F1 score.
However, the analogous art of Otten disclose that it is known to evaluate the accuracy of a trained model based on precision, recall and F1 score functionality, wherein the trained model is evaluated for accuracy based on the F1 score; and wherein the F1 score is about 80% on at least page 1, lines 9-10; page 2, line 3 through page 3, line 5; page 5, lines 18-26; page 6, line 20 though page 9, line 6 (Examiner note: 70% or higher is about 80%).
It would have been obvious to one of ordinary skill in the art to modify the invention of Xie, Shashkina, and Liu to include parameters including precision, recall and an F1 score functionality of about 80%, wherein the trained machine learning model is evaluated for accuracy based on the F1 score as disclosed by Otten.
The motivation for doing so is that it would be obvious to try, an F1 score is one of a limited number of predictable measures commonly used to evaluate the performance of a classification model and more likely than not would predictably result in a high overall performance of the model when the F1 score is about 80%.
Response to Arguments
Applicant's arguments filed November 14, 2025 have been fully considered but they are not persuasive.
The applicant argues with regards to the 35 USC 101 rejection that the claims overcome the 35 USC 101 rejections because the steps of the claim recite a concrete technological process for data integrations, feature engineering, and predictive modeling by preprocessing, normalizing and structuring data followed by feature extraction using NLP algorithms. The examiner disagrees. Each and every step associated with the concrete technological process for data integrations, feature engineering, and predictive modeling is part of the abstract idea itself including the preprocessing, normalizing and structuring data followed by feature extraction. As such, any purported improvement obtained by performing such a concrete technological process is an improvement to an abstract idea which is an improvement in ineligible subject matter (see SAP v. Investpic: Page 2, line 22 through Page 3, line 13 - Even assuming that the algorithms claimed are groundbreaking, innovative or even brilliant, the claims are ineligible because their innovation is an innovation in ineligible subject matter because they are nothing but a series of mathematical algorithms based on selected information and the presentation of the results of those algorithms. Thus, the advance lies entirely in the realm of abstract ideas, with no plausible alleged innovation in the non-abstract application realm. An advance of this nature is ineligible for patenting; and Page 10, lines 18-24 - Even if a process of collecting and analyzing information is limited to particular content, or a particular source, that limitations does not make the collection and analysis other than abstract.). The only limitation mention in the applicant’s argument that might be considered an “additional element” of the abstract idea would be NLP, and even then it would only be an additional element if it were a NLP machine learning model, as algorithms are by definition part of the abstract idea itself. An improvement must be rooted in the “additional elements” of a claim for it to even be capable of overcoming a 35 USC 101 rejection under Step 2a, Prong 2 and/or be capable of overcome a 35 USC 101 rejection under Step 2b. An “additional element” is defined as an element of the claim outside the identified abstract idea itself. However, based on the applicant’s disclosure, the applicant’s invention uses known NLP machine learning models and, as such, the applicant did not invent a new NLP machine learning model. As such, the NLP machine learning models are generic machine learning models. Based on the Recentive Analytics decision generic machine learning models are merely software executed on a general-purpose computer and, as such, are merely used as a tool upon which an abstract idea is merely being applied which is insufficient to transform an abstract idea into a practical application under Step 2a, Prong 2 and insufficient to be considered significantly more under Step 2b. The applicant appears to misconstrue the term “additional element” with an “element of the claim”. Elements of a claim can be broken down into two categories: elements of the abstract idea itself and “additional elements”. It matters not whether recited elements of the abstract idea itself are considered technical steps as such steps irrespective of how groundbreaking, innovative or even brilliant they might be are improvement to the abstract idea which are improvements in ineligible subject matter (see SAP v. Investpic cited above). Thus, the rejection has been maintained.
The applicant further argues with regards to the 35 USC 101 rejection that the claims overcome the 35 USC 101 rejections because they recite steps that are not conventional or generic computer functions. The examiner disagrees. Under Step 2b it is the “additional elements” when not well-understood, routine, and conventional can be considered significantly more than the abstract idea. Whether the steps of the abstract idea itself, which is merely applied using a general-purpose computer with generic computer components and executing generic machine learning algorithms, are well-understood, routine, and conventional has no impact when considering Step 2b. The well-understood, routine, and conventional analysis is reserved for the “additional elements” of a claim. Thus, the rejections have been maintained.
The applicant further argues with regards to the 35 USC 101 rejection that the claims overcome the 35 USC 101 because the claims are similar to the claims in the Desjardins decision. The examiner strongly disagrees. The claims in the Desjardins decision recited a new kind of machine learning model that the applicant invention which operated in a manner different from known machine learning models. In contrast, the claims of the instant case merely use generic machine learning models to apply an abstract idea. As made clear in the Recentive Analytics decision, merely training and/or retraining a known machine learning model with specific data to obtain a desired output is merely applying a generic machine learning model which is insufficient to transform an abstract idea into a practical application under Step 2a, Prong 2 and insufficient to be considered significantly more under Step 2b. The claims of the instant invention are much more similar to the claims of the Recentive Analytics decision and, as such recite merely applying an abstract idea using a generic machine learning model. Thus, the rejection has been maintained.
The applicant further argues with regards to the 35 USC 101 rejection that the claims overcome the 35 USC 101 because the claims are similar to the claims in the Enfish decision. The examiner disagrees. The claims of the Enfish decision recited a new type of database, called a self-referential database, that the applicant invented which operated in a manner different from the way in which traditional databases worked and this distinction was supported by the applicant's specification. In contrast, the instant claim merely apply an abstract idea using known machine learning models. There is no indication in the applicant’s specification that it invented any of the machine learning models, or that the inner workings of the claimed machine learning model operate in a manner different from any other machine learning model trained in the manner disclosed by the applicant. Thus, the applicant’s invention merely applies an abstract idea using a generic machine learning model which is insufficient to transform an abstract idea into a practical application under Step 2a, Prong 2 and insufficient to be considered significantly more under Step 2b. Thus, the rejections have been maintained.
In regards to the 35 USC 103 rejections, the applicant assert that the Xie does not disclose the transforming of text into frequency distributions. The examiner disagrees. Xie specifically discloses at least paragraphs 20-22, 47, and 50 that the frequency of occurances in the text are counted, and based on said count are included in the table. Xie even discloses representing these frequency distributions in a career path graph. Thus, Xie does recite transforming text into frequency distributions and the rejections have been maintained.
The remainer of the applicant’s arguments with regards to the 35 USC 103 rejections appear to be a piecemeal analysis of the prior art where the applicant indicates a specific prior art reference does not teach something that the examiner has cited a different reference to teach. In response to applicant's arguments against the references individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Song Hee Seok (KR20220008645A) which discloses which discloses a deep learning-based job matching system that utilizes a user profile with gender, location, university, final education, major, years of experience, working company, skill, desired salary, professional experience, and detailed career information to recommend a job suitable for the user, wherein the performance of the deep learning-based job matching system is evaluated using precision (Precision), recall (Recall) and f1.
Cheng et al. (PGPUB: 2014/0245184) which discloses providing career recommendations to a member of a social network by receiving input associated with a professional or aspirational goal from a member of a social network, determining a recommendation based on information stored by the social network, and provide the recommendation to the member of the social network .
Terhark et al. (PGPUB: 2018/0232751) which discloses a predictive modeling system having a marketability service module and a career path module. The marketability service module of the system server employs a marketability algorithm to determine how marketable a user is based on self-reported work experience and skills for a job. The career path module employs a career path algorithm to provide to the user a prediction of a most successful path from a users current position to a destination position on the basis of a career paths database of transitions provided by a career path analyzer in communication with the career paths database.
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 JOHN W VAN BRAMER whose telephone number is (571)272-8198. The examiner can normally be reached Monday-Thursday 5:30 am - 4 pm EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Spar Ilana can be reached on 571-270-7537. 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.
/John Van Bramer/Primary Examiner, Art Unit 3622