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 Arguments
Applicant’s arguments with respect to the 112 rejection overcomes the rejection.
Applicant's arguments filed 7/17/2026 have been fully considered but they are not persuasive.
Applicant argues,
Here, the Office Action essentially acknowledges that the claimed "training" limitations in the independent claims 1, 15, and 18 are not mental concepts. In this regard, under Prong One of Step 2A, the independent claims 1, 15, and 18 as a whole cannot be construed to fall within the judicial exception of mental processes.”
Remarks 9.
Training is not a mental concept, agreed. MPEP 2106.04(a)(3) sates, “If the claim as a whole integrates the tentative abstract idea into a practical application, the claim is not directed to a judicial exception…” MPEP 2106.04(a)(2) states, “Claims do not recite a mental process when they do not contain limitations that can practically be performed in the human mind…” These are two separate considerations. Here, the claim does contain claim limitations that cannot be performed in the mind – training. Training is an additional limitation, not the mental concept. However, when considering the claim as a whole, including the additional non-mental-concept limitation of training, the claim as a whole does not integrate the mental concept into a practical application. This is because the training step is well understood, routine and conventional in the field of computing.1 “If… the additional element (or combination of elements) is no more than well-understood, routine, conventional activities previously known to the industry, which is recited at a high level of generality, then this consideration does not favor eligibility.” MPEP 2106.05(d).
Applicant argues,
recites additional elements that integrate the supposed judicial exception into a practical application. In particular, the independent claims 1, 15, and 18 recite additional elements that reflect an improvement to other technology or technical field with regard to the identification of personally identifiable information (PII).…The embodiments can be especially useful in edge locations, where the quick identification of PII may be required….
Remarks 10-11.
Identification of a PII is not a technology or technical field. An edge location is not claimed. Therefore, the claim does not improve the function of a computer or a technological field.
Applicant argues,
receiving, by the processing platform, event-based data from a given information source of the one or more information sources, the event-based data comprising one or more attributes of a given application or database of the given information source, wherein the event-based data comprises schema level information which represents the event-based data in a given schema format associated with the given application or database;
converting, by the processing platform, the event-based data in the given schema format into a common format for analysis by the processing platform;
extracting, by the processing platform, one or more attributes from the event-based data in the common format.
These limitations provide a practical technical solution whereby a machine learning model can be trained to classify PII data from disparate data sources (e.g., different databases and applications) with different information schemas.
Remarks 11-12.
Receiving data is a “Mere Data Gathering” and it is insignificant extra-solution activity and does not integrate the abstract idea into a practical application and does not amount to significantly more than the abstract idea. MPEP 2106.05(g).
Applicant argues, “Medalion does not disclose that the flagged data 124 is stored in a data repository to provide secure access to the attributes which are classified as comprising personally identifiable information, within the context of claims 1, 15, and 18.” Remarks 14. Applicant claims, “storing, by the processing platform, attributes of the one or more extracted attributes which are classified by the at least one trained machine learning neural network model as comprising personally identifiable information, in a data repository to provide secured access to the attributes which are classified as comprising personally identifiable information.” Claim 1. Medalion paragraph 118 teaches this idea “Processing system 500 further includes storage 530, which in this example includes production data 532, which may be like production data collection 122 described above with respect to FIG. 1. Storage 510 also includes flagged data 534, which may be flagged data collection 124 in FIG. 1.” Medalion also teaches a separate flagged data storage unit in fig. 1 as flagged data 124.
Applicant argues,
Medalion does not disclose or suggest, e.g.:
receiving, by the processing platform, event-based data from a given information source of the one or more information sources, the event-based data comprising one or more attributes of a given application or database of the given information source, wherein the event-based data comprises schema level information which represents the event-based data in a given schema format associated with the given application or database;
converting, by the processing platform, the event-based data in the given schema format into a common format for analysis by the processing platform;
extracting, by the processing platform, one or more attributes from the event-based data in the common format, as recited in claim 1, as well as clams 15 and 18, as currently amended.
Remarks 14.
Medalion paragraph 70 teaches receiving converting and extracting, “a text string, such as “Sure. My SSN is 1234 oh kids please be quiet 56780 I mean 9” may be filtered (e.g., by filters 104 in FIG. 1) to remove punctuation and then sent to an embedding layer as, for example, a vector or tensor input.” Text string is the received data. filtering is extracting and embedding the filtered data is extracting.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to abstract idea mental concept without significantly more. The claims recite extracting attributes, analyzing attributes and classifying data as PII. This judicial exception is not integrated into a practical application because elements directed to receiving and collecting data are mere data gathering and do not improve the functioning of a computer or technical field; further the training step is well understood, routine and conventional in the field of computing.2 The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because computer parts are generic computer parts; and the training step is well-understood, routine and conventional. Id.
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.
Claims 1-20 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. In claims 1, 6, 15 and 18, Applicant claims a “common format” and this is never described in the specification. The same is true for “schema format” in claims 1, 15 and 18.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claims 1-8, 10 and 15-20 are rejected under 35 U.S.C. 102(a)(1) as being described by US20210125615A1 to Medalion et al.
Medalion teaches claims 1, 15 and 18. A method, comprising:
training, by a processing platform, at least one machine learning neural network model which is trained to identify personally identifiable information that is extracted from attributes which are included in one or more applications and databases of one or more information sources that are coupled to the processing platform over a communications network; (Medalion para 118 “Storage 510 also includes training data 536, which may be like training data collection 126 in FIG. 1.” Medalion para 92 “After human review, the text strings with verified PII data (as well as with verified no PII data) may be provided to training data collection 126, and the training data may subsequently be used to train or refine machine learning model(s) 108.” The independent variable is the text string. The dependent variable is the flag for PII. Medalion fig. 2 and abstract “one or more bidirectional long short-term memory (BiLSTM) neural network models…”)
receiving by the processing platform, event-based data from a given information source of the one or more information sources, the event-based data comprising one or more attributes of a given application or database of the given information source, wherein the event-based data comprises schema level information which represents the event-based data in a given schema format associated with the given application or database; (Medalion para 97 “step 402 with receiving a plurality of text strings, each of the plurality of text strings associated with a user support session.” The schema is the session and associated strings.)
converting, by the processing platform, the event-based data in the given schema format into a common format for analysis by the processing platform; (The specification paragraph 32 says that the conversion and extracting are the same operation, “The schema information is parsed by the event collection and conversion component 121 to extract relevant information and put the relevant information into a format for further analysis.” This conversion step is taught by Medalion para 98, see below.)
extracting, by the processing platform, one or more attributes from the event-based data in the common format; and (Medalion para 98 “step 404 with providing the plurality of text strings to a bidirectional long short-term memory (BiLSTM) neural network model.” Medalion para 70 “a text string, such as “Sure. My SSN is 1234 oh kids please be quiet 56780 I mean 9” may be filtered (e.g., by filters 104 in FIG. 1) to remove punctuation and then sent to an embedding layer as, for example, a vector or tensor input.” The embeddings are the extracted data.)
utilizing, by the processing platform, the at least one trained machine learning neural network model to analyze the one or more extracted attributes to classify whether the one or more extracted attributes comprise personally identifiable information wherein the one or more extracted attributes are applied to an input layer of the at least one trained machine learning neural network model, (Medalion fig. 2 the extracted embedding is the input layer.) analyzed by at least one hidden layer of the at least one trained machine learning neural network model, (Medalion fig. 2 210 shows the embedding being analyzed by hidden layers of a neural network.) and classified by an output layer of the at least one trained machine learning neural network model; and (Medalion para 100 “step 406 with receiving output from the BiLSTM neural network model, the output indicating one or more text data elements in the plurality of text strings comprising predicted personally identifiable information.” Medalion fig. 2 show the output layer with classifications y.)
storing, by the processing platform, attributes of the one or more extracted attributes which are classified by the at least one trained machine learning neural network model as comprising personally identifiable information, in a data repository to provide secured access to the attributes which are classified as comprising personally identifiable information. (Medalion fig. 5 storage 530, flagged data 534. Medalion para 61 “When the predicted PII text is flagged, it may be indicated as such in the text string by specific formatting so that a reviewer can easily review and determine whether the flagged text is indeed PII.” Medalion fig. 1 flagged data storage 124 and fig. 5 flagged data storage 534.)
Medalion teaches claim 2. The method of claim 1, wherein the event-based data corresponds to one or more events in which one or more attributes are added to at least one of a database and an application. (Medalion fig. 5 storage 530)
Medalion teaches claim 3. The method of claim 2, wherein the one or more attributes are added to at least one of a table of the database and an object model of the application. (Medalion storage 530 the object model is the flagged data and production data. Medalion para 42 “The example transcript in Table 1 may be used to generate training data for a machine learning-based model,”)
Medalion teaches claim 4. The method of claim 2, wherein utilizing the at least one trained machine learning neural network model to analyze the one or more extracted attributes the analyzing is performed in real-time responsive to the one or more events. (Medalion para 97 “Method 400 begins at step 402 with receiving a plurality of text strings, each of the plurality of text strings associated with a user support session.”)
Medalion teaches claim 5. The method of claim 1, wherein the schema level information comprises information which describes at least one of an organization and structure of the data in a given database. (Medalion para 97 “step 402 with receiving a plurality of text strings, each of the plurality of text strings associated with a user support session.” The schema is the session and associated strings. Strings are structure and organization. Associating strings with a session is a structure and organization of the data also.)
Medalion teaches claim 6. The method of claim 1, wherein the extracting of the one or more attributes from the event-based data in the common format is based at least in part on one or more context rules which identify a given context to apply to a given attribute. (Medalion para 70 “Initially, a text string, such as “Sure. My SSN is 1234 oh kids please be quiet 56780 I mean 9” may be filtered (e.g., by filters 104 in FIG. 1) to remove punctuation and then sent to an embedding layer as, for example, a vector or tensor input.” Filtering is a context rule.)
Medalion teaches claims 7, 16 and 19. The method of claim 1, wherein the at least one trained machine learning neural network model implements a neural network-based binary classification algorithm to classify whether the one or more attributes comprise personally identifiable information. (Medalion para 100 “Method 400 then proceeds to step 406 with receiving output from the BiLSTM neural network model, the output indicating one or more text data elements in the plurality of text strings comprising predicted personally identifiable information.”)
Medalion teaches claims 8, 17 and 20. The method of claim 1, wherein the at least one trained machine learning neural network model is trained with training data comprising a plurality of attributes as independent variables, wherein respective ones of the plurality of attributes correspond to respective dependent variables indicating whether the respective ones of the plurality of attributes comprise personally identifiable information. (Medalion para 118 “Storage 510 also includes training data 536, which may be like training data collection 126 in FIG. 1.” Medalion para 92 “After human review, the text strings with verified PII data (as well as with verified no PII data) may be provided to training data collection 126, and the training data may subsequently be used to train or refine machine learning model(s) 108.” The independent variable is the text string. The dependent variable is the flag for PII.)
Medalion teaches claim 10. The method of claim 1, wherein:
the at least one trained machine learning neural network model comprises a neural network which implements a neural network-based binary classification algorithm; and
the neural network comprises a plurality of nodes with connections between nodes, the connections each comprise a weight factor; and (Medalion para 75 “The neural network models 212, 222, and 224 may in some embodiments have the same structure, but different learned weights…”)
each node of the plurality of nodes comprises a bias factor. (Medalion para 80 “during initial model training, weighting between the three independent predictions ŷ1, ŷ2, and ŷ3 via weights W1, W2, and W3 may be biased for better generalization to improve detection of PII from context.”)
Claim Rejections - 35 USC § 103
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 9 is rejected under 35 U.S.C. 103 as being unpatentable over US20210125615A1 to Medalion et al and US 20210271822 A1 to Bui et al.
Claims 11-14 are rejected under 35 U.S.C. 103 as being unpatentable over US20210125615A1 to Medalion et al and US20140025660A1 to Mohammed et al.
Medalion teaches claim 9. The method of claim 1, wherein the at least one hidden layer of the at least one trained machine learning neural network model comprises nodes that implement respective(Medalion layers 204 and 206 in fig. 2 are hidden layers.) Medalion doesn’t have a ReLU.
However, Bui teaches rectified linear unit activation functiona. (Bui para 11 “the control module, in order to regulate the GCN representation vectors, applies an activation function (e.g., a rectified linear unit (ReLU)) to the BiLSTM representation vector of the target word to generate a BiLSTM control vector and a GCN control vector, respectively…” Bui fig. 2 shows two control modules 13, in two different encoders. Each control module has a ReLu according to bui para 105.)
Bui, Medalion and the claims are all processing natural language with neural networks. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to include a ReLU because a ReLU is computationally simpler and faster to compute than a sigmoid or tanh.
Medalion teaches claim 11. The method of claim 1, wherein storing attribtues of the one or more extracted attributes which are classified by the at least one trained machine learning neural network model as comprising personally identifiable information, in a data repository to provide secured access to the attributes which are classified as comprising personally identifiable information, comprises storing, (Medalion para 100 “Method 400 then proceeds to step 406 with receiving output from the BiLSTM neural network model, the output indicating one or more text data elements in the plurality of text strings comprising predicted personally identifiable information.”) Medalion doesn’t teach a social graph.
However, Mohammed teaches, in one or more relationship graphs, the one or more attributes which are classified as comprising personally identifiable information, wherein the one or more relationship graphs comprise a plurality of relationships between a plurality of nodes, wherein the plurality of relationships comprise edges of the one or more relationship graphs. (Mohammed para 82 “The social graph 600 may include one or more connected person nodes associated with a user node 300 (e.g., person-to-person links), and may generally reflect the composition of a user's social, family, and/or professional network.”)
Mohammed, Medalion and the claims all deal with PII data. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to use a social graph in order to provide “users with an opportunity to opt-in and/or opt-out of data collection services, and/or limiting the transmission of collected data to trusted services…” Mohammed para 34.
Mohammed teaches claim 12. The method of claim 11, wherein the plurality of nodes comprise the one or more extracted attributes which are classified as comprising personally identifiable information and one or more other attributes. (Mohammed para 97 “In certain embodiments, anonymizing the personal information may comprise removing and/or filtering certain personally identifiable information from personal information reflected in the personal ontology graph 100…”)
Mohammed teaches claim 13. The method of claim 11, wherein the plurality of relationships comprise interactions between respective pairs of the plurality of nodes. (Mohammed para 82 “The social graph 600 may include one or more connected person nodes associated with a user node 300 (e.g., person-to-person links), and may generally reflect the composition of a user's social, family, and/or professional network.”)
Mohammed teaches claim 14. The method of claim 11, wherein the one or more relationship graphs are in one of a resource description framework (RDF) format and a labeled property graph (LPG) format. (Mohammed para 76 “interconnected relationships between nodes may comprise one or more resource description framework ("RDF") triples.”)
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Austin Hicks whose telephone number is (571)270-3377. The examiner can normally be reached Monday - Thursday 8-4 PST.
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, Mariela Reyes can be reached at (571) 270-1006. 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.
/AUSTIN HICKS/Primary Examiner, Art Unit 2142
1 Medalion para 118 “Storage 510 also includes training data 536, which may be like training data collection 126 in FIG. 1.” Medalion para 92 “After human review, the text strings with verified PII data (as well as with verified no PII data) may be provided to training data collection 126, and the training data may subsequently be used to train or refine machine learning model(s) 108.”
2 Medalion para 118 “Storage 510 also includes training data 536, which may be like training data collection 126 in FIG. 1.” Medalion para 92 “After human review, the text strings with verified PII data (as well as with verified no PII data) may be provided to training data collection 126, and the training data may subsequently be used to train or refine machine learning model(s) 108.”
Prosecution Insights