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 .
DETAILED ACTION
This office action is in response to correspondence 02/26/26 regarding application 18/789,839, in which claims 1, 6, 9, 12, and 17 were amended. Claims 1-20 are pending in the application and have been considered.
Response to Arguments
The examiner agrees with Applicant on page 7 that no new matter was added via the amendments to claims 1, 6, 9, 12, and 17.
Applicant’s arguments on pages 7-8 regarding the 35 U.S.C. 103 rejections based on Wong, Messing, Zhang, Gdak, and Kumar have been considered but are moot in view of the new grounds for rejection based in part on the newly discovered reference to Butvinik et al. (US 20240013223), which describes generating synthetic labeled tabular financial transaction data. The new grounds for rejection based in part on Butvinik are necessitated by Applicant’s amendments to the claims.
Still Missing Oath
Applicant’s attention is directed to the notice 08/16/24 informing Applicant that no properly executed inventor’s oaths or declarations have been received. Applicant is respectfully requested to submit the oath in a timely manner in order to prevent potential delays, should the application otherwise be found in condition for allowance.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102 of this title, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 9, 11, 12, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Wong et al. (US 20230118240) in view of Messing et al. (US 20030126134), in further view of Butvinik et al. (US 20240013223).
Consider claim 1, Wong discloses a method for training a machine learning model to automatically identify transactions (method of training machine learning system to detect anomalies within transaction, i.e. identify transactions that are anomalies, [0112]), the method comprising:
obtaining sample data (obtain training set 1202, Fig. 12, [0122]), the sample data including:
(i) a plurality of live transactions (training set of transaction data samples, [0122], which are live, timestamped transactions, Fig 11A, Fig 11B, [0119]-[0121]); and
(ii) a first set of labels, each label in the first set of labels corresponding to a respective attribute of a plurality of different attributes of each of the plurality of live transactions (samples are assigned label “0”, so the set of “0” labels respectively indicates that each sample is not an anomaly, the transactions having a plurality of different attributes such as presence or absence of anomaly, amount, merchant_id, etc,. [0123], Fig. 11B);
generating training data based on the sample data (synthetic data is generated based on the labeled and partitioned training set, Fig 12 steps 1202-1208, [0122-0125]), the training data comprising:
(i) a plurality of synthetic transactions (generating synthetic data samples by swapping the context features, representing history transaction data, with the observable features, [0124-0125]; this is considered to create a representation of new, synthetic transactions, Fig 12 step 1208); and
(ii) a second set of labels including one or more labels that differ from each label included in the first set of labels (the synthetic transactions are each labeled with “1”, so the set of 1s differ from the set of 0s with which the original samples were labeled, [0126], Fig 12 step 1210); and
training the machine learning model to automatically identify transactions using the training data (the supervised ML system is trained with the training set to automatically identify anomalous transactions, [0127], Fig 12 step 1214).
Wong does not specifically mention obtaining data from a webpage including a first set of labels, each label in the first set of labels comprising natural language text; and extracting transactions from webpages, and data from a webpage.
Messing discloses obtaining data from a webpage including a first set of labels, each label in the first set of labels comprising natural language text (screen scraping from e.g. the webpage shown in Fig. 5 to obtain labels “Account Number”, “Account Value”, etc., [0024]-[0026], [0033-0034]), and extracting transactions from webpages, and data from a webpage (financial analysis system captures transaction descriptions and amounts from webpages, [0024]-[0026]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong by obtaining data from a webpage including a first set of labels, each label in the first set of labels comprising natural language text; and extracting transactions from webpages, and data from a web page in order to provide users with account information via the internet without having to login to each institution at which accounts are held, as suggested by Messing ([0002]-[0004]). Doing so would have led to predictable results of facilitating easier access and aggregation of information across various account types, as suggested by Messing ([0002]). The references cited are analogous art in the same field of financial transaction processing.
Wong and Messing do not specifically mention generating a second set of labels comprising natural language text corresponding to a respective attribute of a plurality of different attributes of each of the plurality of synthetic transactions.
Butvinik discloses generating a second set of labels comprising natural language text corresponding to a respective attribute of a plurality of different attributes of each of the plurality of synthetic transactions (synthetic fraud 530 having generated column values f1, f2, f3, … fk, corresponding to labels “payer name”, “amount”, “address”, “bank branch”, etc., [0084], as well as label “synthetic fraud”, Fig 5, [0098]; these are labels considered to be generated by the CTGAN as part of generating synthetic data, which is combined with cleaned tabular data to training dataset, [0074]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong and Messing by generating a second set of labels comprising natural language text corresponding to a respective attribute of a plurality of different attributes of each of the plurality of synthetic transactions in order to improve class imbalances for machine learning involving financial transactions, as suggested by Butvinik ([0003], [0005]). Doing so would have led to predictable results of helping financial institutions with fraud detection and prevention, as suggested by Butvinik ([0004], [0007]). The references cited are analogous art in the same field of financial transaction processing.
Consider claim 9, Wong discloses a method for automatically identifying transactions (method of training machine learning system to detect anomalies within transaction, i.e. identify transactions that are anomalies, [0112]), comprising:
providing input data to a machine learning model trained to automatically identify transactions using training data including a plurality of synthetic transactions generated from sample data, the input data comprising text (transaction data including alphanumeric text is input to machine learning system, [0051], [0052], machine learning system trained to predict validity of transactions from sample transaction training set and synthetic data samples, Fig 12 steps 1202, [0122]-[0127]); and
receiving output data from the machine learning model based on the input data (machine learning model outputs a prediction for proposed transaction 1102, [0120]);
wherein the sample data includes sample data includes: (i) a plurality of live transactions (training set of transaction data samples, [0122], which are live, timestamped transactions, Fig 11A, Fig 11B, [0119]-[0121]); and (ii) a first set of labels, each label in the first set of labels corresponding to a respective attribute of a plurality of different attributes of each of the plurality of live transactions (samples are assigned label “0”, so the set of “0” labels respectively indicates that each sample is not an anomaly, the transactions having a plurality of different attributes such as presence or absence of anomaly, amount, merchant_id, etc,. [0123], Fig. 11B); and
wherein the training data further includes a second set of labels including one or more labels that differ from each label included in the first set of labels (the synthetic transactions are each labeled with “1”, so the set of 1s differ from the set of 0s with which the original samples were labeled, [0126], Fig 12 step 1210).
Wong does not specifically mention extracting transactions from webpages, text displayed on a webpage, and the output data comprising one or more transactions included in the text; each label in the first set of labels comprising natural language text correspond to a respective attribute of a plurality of different attributes of each of the plurality of live transactions.
Messing discloses extracting transactions from webpages, text displayed on a webpage, and the output data comprising one or more transactions included in the text (financial analysis system captures transaction descriptions and amounts from webpages, [0024]-[0026], including textual descriptions such as “Account”, Fig. 5, [0034], which are normalized and output to database for storage, [0029]); each label in the first set of labels comprising natural language text correspond to a respective attribute of a plurality of different attributes of each of the plurality of live transactions (screen scraping from e.g. the webpage shown in Fig. 5 to obtain labels “Account Number”, “Account Value”, etc., [0024]-[0026], [0033-0034]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong by extracting transactions from webpages, providing input data comprising text displayed on a webpage, the outputting data comprising one or more transactions included in the text, and each label in the first set of labels comprising natural language text correspond to a respective attribute of a plurality of different attributes of each of the plurality of live transactions for reasons similar to those for claim 1.
Wong and Messing do not specifically mention wherein the training data further includes a second set of labels comprising natural language text corresponding to a respective attribute of a plurality of different attributes of each of the plurality of synthetic transactions.
Butvinik discloses generating a second set of labels comprising natural language text corresponding to a respective attribute of a plurality of different attributes of each of the plurality of synthetic transactions (synthetic fraud 530 having generated column values f1, f2, f3, … fk, corresponding to labels “payer name”, “amount”, “address”, “bank branch”, etc., [0084], as well as label “synthetic fraud”, Fig 5, [0098]; these are labels considered to be generated by the CTGAN as part of generating synthetic data, which is combined with cleaned tabular data to training dataset, [0074]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong and Messing by generating a second set of labels comprising natural language text corresponding to a respective attribute of a plurality of different attributes of each of the plurality of synthetic transactions for reasons similar to those for claim 1.
Consider claim 12, Wong discloses system for training a machine learning model to automatically identify transactions (method of training machine learning system to detect anomalies within transaction, i.e. identify transactions that are anomalies, [0112]), the system comprising:
a memory including computer executable instructions (instructions stored on memory, [0154]); and
a processor configured to execute the computer executable instructions (processor executes the instructions, [0154]) and cause the system to:
obtain sample data (obtain training set 1202, Fig. 12, [0122]), the sample data including: (i) a plurality of live transactions (training set of transaction data samples, [0122], which are live, timestamped transactions, Fig 11A, Fig 11B, [0119]-[0121]); and (ii) a first set of labels, each label in the first set of labels corresponding to a respective attribute of a plurality of different attributes of each of the plurality of live transactions (samples are assigned label “0”, so the set of “0” labels respectively indicates that each sample is not an anomaly, the transactions having a plurality of different attributes such as presence or absence of anomaly, amount, merchant_id, etc,. [0123], Fig. 11B);
generate training data based on the sample data (synthetic data is generated based on the labeled and partitioned training set, Fig 12 steps 1202-1208, [0122-0125]), the training data comprising: (i) a plurality of synthetic transactions (generating synthetic data samples by swapping the context features, representing history transaction data, with the observable features, [0124-0125]; this is considered to create a representation of new, synthetic transactions, Fig 12 step 1208); and (ii) a second set of labels including one or more labels that differ from each label included in the first set of labels (the synthetic transactions are each labeled with “1”, so the set of 1s differ from the set of 0s with which the original samples were labeled, [0126], Fig 12 step 1210); and
train the machine learning model to automatically identify transactions using the training data (the supervised ML system is trained with the training set to automatically identify anomalous transactions, [0127], Fig 12 step 1214).
Wong does not specifically mention obtaining data from a webpage including a first set of labels, each label in the first set of labels comprising natural language text; and extracting transactions from webpages, and data from a webpage.
Messing discloses obtaining data from a webpage including a first set of labels, each label in the first set of labels comprising natural language text (screen scraping from e.g. the webpage shown in Fig. 5 to obtain labels “Account Number”, “Account Value”, etc., [0024]-[0026], [0033-0034]), and extracting transactions from webpages, and data from a webpage (financial analysis system captures transaction descriptions and amounts from webpages, [0024]-[0026]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong by obtaining data from a webpage including a first set of labels, each label in the first set of labels comprising natural language text; and extracting transactions from webpages, and data from a web page for reasons similar to those for claim 1.
Wong and Messing do not specifically mention generating a second set of labels comprising natural language text corresponding to a respective attribute of a plurality of different attributes of each of the plurality of synthetic transactions.
Butvinik discloses generating a second set of labels comprising natural language text corresponding to a respective attribute of a plurality of different attributes of each of the plurality of synthetic transactions (synthetic fraud 530 having generated column values f1, f2, f3, … fk, corresponding to labels “payer name”, “amount”, “address”, “bank branch”, etc., [0084], as well as label “synthetic fraud”, Fig 5, [0098]; these are labels considered to be generated by the CTGAN as part of generating synthetic data, which is combined with cleaned tabular data to training dataset, [0074]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong and Messing by generating a second set of labels comprising natural language text corresponding to a respective attribute of a plurality of different attributes of each of the plurality of synthetic transactions for reasons similar to those for claim 1.
Consider claim 11, Wong discloses the input data comprises a transactions table including the text, the transactions table including multiple rows and multiple columns, each of the rows including a different transaction of a plurality of transactions and each of the rows corresponding to a respective attribute of a plurality of different attributes of the plurality of transactions (transaction data including alphanumeric text is input to machine learning system, [0051], [0052], with multiple rows and columns, Fig 3A, including timestamps, amounts, etc., Fig 3B).
Consider claim 20, Wong does not, but Messing discloses the webpage comprises a hypertext markup language (HTML) webpage (retrieving data from an HTML screen, [0029], Fig 3A).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong such that the webpage comprises a hypertext markup language for reasons similar to those for claim 1.
Claims 2-6 and 13-17 are rejected under 35 U.S.C. 103 as being unpatentable over Wong et al. (US 20230118240) in view of Messing et al. (US 20030126134), in further view of Butvinik et al. (US 20240013223), in further view of Zhang et al. (US 20240020312).
Consider claim 2, Wong discloses generating the training data comprises: generating the plurality of synthetic transactions (synthetic data is generated based on the labeled and partitioned training set, Fig 12 steps 1202-1208, [0122-0125]).
Wong, Messing, and Butvinik do not specifically mention generating a dictionary including one or more alternate labels for each respective label in the first set of labels; and replacing one or more respective labels included in the first set of labels with an alternate label for the one or respective labels in the dictionary.
Zhang discloses generating a dictionary including one or more alternate labels for each respective label in the first set of labels (a dictionary storing mapping relationships between each normalized label and its related unnormalized labels, [0073]); and replacing one or more respective labels included in the first set of labels with an alternate label for the one or respective labels in the dictionary (the records are normalized by replacing unnormalized labels with normalized ones using dictionary storing mappings, [0072]-[0073]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik by generating a dictionary including one or more alternate labels for each respective label in the first set of labels; and replacing one or more respective labels included in the first set of labels with an alternate label for the one or respective labels in the dictionary in order to address records from sampling sources having different names, as suggested by Zhang ([0067]), predictably enabling unification and efficient management and usage of samples, as suggested by Zhang ([0068]). The references cited are analogous art in the same field of financial transaction processing (“The TSDBs have applications in various fields such as financial transaction systems”, Zhang, [0002]).
Consider claim 3, Wong, Messing, and Butvinik do not, but Zhang discloses the one or more alternate labels for each respective label in the first set of labels comprises one or more synonyms of the respective label (e.g. “http_requests_total” and “httprequest_total” are synonyms for the normalized “http_request_total”, [0073]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik such that the one or more alternate labels for each respective label in the first set of labels comprises one or more synonyms of the respective label for reasons similar to those for claim 2.
Consider claim 4, Wong does not, but Messing discloses obtaining a label from a different webpage than the webpage from which the sample data is obtained (data collected from a variety of different web pages is arranged or normalized into the same format, [0029]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong by obtaining a label from a different webpage than the webpage from which the sample data is obtained in order to provide users with account information via the internet without having to login to each institution at which accounts are held, as suggested by Messing ([0002]-[0004]). Doing so would have led to predictable results of facilitating easier access and aggregation of information across various account types, as suggested by Messing ([0002]).
Wong, Messing, and Butvinik do not specifically mention one or more alternate labels for each respective label in the first set of labels comprise a label for each respective label.
Zhang discloses one or more alternate labels for each respective label in the first set of labels comprise a label for each respective label (dictionary of alternative labels for each normalized label, [0073]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik such that the one or more alternate labels for each respective label in the first set of labels comprise a label for each respective label as in Zhang on a different webpage than the webpage from which the sample data is obtained as in Messing for reasons similar to those for claim 2.
Consider claim 5, Wong discloses randomly replacing (synthetic samples are generated by separating observable and context features, then randomly replacing the context features in a sample with another context feature for which the entity identifier does not match, [0115], [0116]).
Wong, Messing, and Butvinik do not specifically mention the replacing comprises replacing the one or more respective labels included in the first set of labels with the alternate label for the one or more respective labels in the dictionary.
Zhang discloses the replacing comprises replacing the one or more respective labels included in the first set of labels with the alternate label for the one or more respective labels in the dictionary (the records are normalized by replacing unnormalized labels with normalized ones using dictionary storing mappings, [0072]-[0073]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik such that the replacing comprises randomly replacing, as in Wong, the one or more respective labels included in the first set of labels with the alternate label for the one or more respective labels in the dictionary as in Zhang for reasons similar to those for claim 2.
Consider claim 6, Wong discloses: a first attribute of one or more of the plurality of live transactions has a monetary format (amount “45.75”, Fig. 3B, [0052]); and generating the training data of one or more of the synthetic transactions (synthetic data is generated based on the labeled and partitioned training set, Fig 12 steps 1202-1208, [0122-0125]).
Wong, Messing, and Butvinik do not specifically mention adding an operand for the first attribute.
Zhang discloses adding an operand to a first attribute (normalizing temperature samples by adding “C” or “Celsius” as an operand to the numeric temperature after converting it, [0073-0074]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik by adding an operand for the first attribute as in Zhang for reasons similar to those for claim 2.
Consider claim 13, Wong discloses wherein to generate the training data, the computer executable instructions cause the system to: generate the plurality of synthetic transactions (synthetic data is generated based on the labeled and partitioned training set, Fig 12 steps 1202-1208, [0122-0125]).
Wong, Messing, and Butvinik do not specifically mention generating a dictionary including one or more alternate labels for each respective label in the first set of labels; and replacing one or more respective labels included in the first set of labels with an alternate label for the one or respective labels in the dictionary.
Zhang discloses generating a dictionary including one or more alternate labels for each respective label in the first set of labels (a dictionary storing mapping relationships between each normalized label and its related unnormalized labels, [0073]); and replacing one or more respective labels included in the first set of labels with an alternate label for the one or respective labels in the dictionary (the records are normalized by replacing unnormalized labels with normalized ones using dictionary storing mappings, [0072]-[0073]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik by generating a dictionary including one or more alternate labels for each respective label in the first set of labels; and replacing one or more respective labels included in the first set of labels with an alternate label for the one or respective labels in the dictionary for reasons similar to those for claim 2.
Consider claim 14, Wong, Messing, and Butvinik do not, but Zhang discloses the one or more alternate labels for each respective label in the first set of labels comprises one or more synonyms of the respective label (e.g. “http_requests_total” and “httprequest_total” are synonyms for the normalized “http_request_total”, [0073]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik such that the one or more alternate labels for each respective label in the first set of labels comprises one or more synonyms of the respective label for reasons similar to those for claim 2.
Consider claim 15, Wong does not, but Messing discloses obtaining a label from a different webpage than the webpage from which the sample data is obtained (data collected from a variety of different web pages is arranged or normalized into the same format, [0029]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong by obtaining a label from a different webpage than the webpage from which the sample data is obtained in order to provide users with account information via the internet without having to login to each institution at which accounts are held, as suggested by Messing ([0002]-[0004]). Doing so would have led to predictable results of facilitating easier access and aggregation of information across various account types, as suggested by Messing ([0002]).
Wong, Messing, and Butvinik do not specifically mention one or more alternate labels for each respective label in the first set of labels comprise a label for each respective label.
Zhang discloses one or more alternate labels for each respective label in the first set of labels comprise a label for each respective label (dictionary of alternative labels for each normalized label, [0073]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik such that the one or more alternate labels for each respective label in the first set of labels comprise a label for each respective label as in Zhang on a different webpage than the webpage from which the sample data is obtained as in Messing for reasons similar to those for claim 2.
Consider claim 16, Wong discloses the computer executable instructions cause the system to: randomly replace (synthetic samples are generated by separating observable and context features, then randomly replacing the context features in a sample with another context feature for which the entity identifier does not match, [0115], [0116]).
Wong, Messing, and Butvinik do not specifically mention replacing the one or more respective labels included in the first set of labels with the alternate label for the one or more respective labels in the dictionary.
Zhang discloses replacing the one or more respective labels included in the first set of labels with the alternate label for the one or more respective labels in the dictionary (the records are normalized by replacing unnormalized labels with normalized ones using dictionary storing mappings, [0072]-[0073]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik such by randomly replacing, as in Wong, the one or more respective labels included in the first set of labels with the alternate label for the one or more respective labels in the dictionary as in Zhang for reasons similar to those for claim 2.
Consider claim 17, Wong discloses: a first attribute of one or more of the plurality of live transactions has a monetary format (amount “45.75”, Fig. 3B, [0052]); and generating the training data of one or more of the synthetic transactions (synthetic data is generated based on the labeled and partitioned training set, Fig 12 steps 1202-1208, [0122-0125]).
Wong, Messing, and Butvinik do not specifically mention adding an operand for the first attribute.
Zhang discloses adding an operand to a first attribute (normalizing temperature samples by adding “C” or “Celsius” as an operand to the numeric temperature after converting it, [0073-0074]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik by adding an operand for the first attribute as in Zhang for reasons similar to those for claim 2.
Claims 7 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Wong et al. (US 20230118240) in view of Messing et al. (US 20030126134), in further view of Butvinik et al. (US 20240013223), in further view of Gdak et al. (US 11922328).
Consider claim 7, Wong discloses using the machine learning model (machine learning model to detect anomalies within transactions, [0112]).
Wong does not specifically mention using the machine learning model to automatically extract transactions from a webpage other than the webpage from which the sample data is obtained.
Messing discloses automatically extracting transactions from a webpage other than the webpage from which the sample data is obtained (data is automatically collected from a variety of different web pages is arranged or normalized into the same format, [0029]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong by using the machine learning model of Wong to automatically extract transactions from a webpage other than the webpage from which the sample data is obtained as in Messing in order to provide users with account information via the internet without having to login to each institution at which accounts are held, as suggested by Messing ([0002]-[0004]). Doing so would have led to predictable results of facilitating easier access and aggregation of information across various account types, as suggested by Messing ([0002]).
Wong, Messing, and Butvinik do not specifically mention subsequent to training the machine learning model using the training data, fine-tuning the trained machine learning model.
Gdak discloses training the machine learning model using the training data, fine-tuning the trained machine learning model (fine-tuning a base ML model on banking statements, Col 2 lines 28-35).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik by, subsequent to training the machine learning model using the training data, fine-tuning the trained machine learning model as in Gdak based on using the machine learning model to automatically extract transactions from a webpage other than the webpage from which the sample data is obtained as in Messing in order to improve accuracy, as suggested by Gdak (Col 2 lines 28-35), predictably improving ability to extract data points of interest to the user, as suggested by Gdak (Col 1 lines 34-39). The references cited are analogous art in the same field of financial transaction processing.
Consider claim 18, Wong discloses using the machine learning model (machine learning model to detect anomalies within transactions, [0112]).
Wong does not specifically mention using the machine learning model to automatically extract transactions from a webpage other than the webpage from which the sample data is obtained.
Messing discloses automatically extracting transactions from a webpage other than the webpage from which the sample data is obtained (data is automatically collected from a variety of different web pages is arranged or normalized into the same format, [0029]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong by using the machine learning model of Wong to automatically extract transactions from a webpage other than the webpage from which the sample data is obtained as in Messing in order to provide users with account information via the internet without having to login to each institution at which accounts are held, as suggested by Messing ([0002]-[0004]). Doing so would have led to predictable results of facilitating easier access and aggregation of information across various account types, as suggested by Messing ([0002]).
Wong, Messing, and Butvinik do not specifically mention subsequent to training the machine learning model using the training data, fine-tuning the trained machine learning model.
Gdak discloses training the machine learning model using the training data, fine-tuning the trained machine learning model (fine-tuning a base ML model on banking statements, Col 2 lines 28-35).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik by, subsequent to training the machine learning model using the training data, fine-tuning the trained machine learning model as in Gdak based on using the machine learning model to automatically extract transactions from a webpage other than the webpage from which the sample data is obtained as in Messing for reasons similar to those for claim 7.
Claims 8, 10, and 19 rejected under 35 U.S.C. 103 as being unpatentable over Wong et al. (US 20230118240) in view of Messing et al. (US 20030126134), in further view of Butvinik et al. (US 20240013223), in further view of Kumar et al. (US 20230113578).
Consider claim 8, Wong discloses the machine learning model (machine learning model to detect anomalies within transactions, [0112]).
Wong, Messing, and Butvinik do not specifically mention the machine learning model comprises a named entity recognition model.
Kumar discloses a named entity recognition model (unlabeled text is analyzed with Named Entity Recognition, [0016]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik such that the machine learning model of Wong comprises a named entity recognition model as in Kumar in order to reduce time consumed and errors made in analyzing bank statements with multiple layouts and formats, as suggested by Kumar ([0002]), predictably enabling more wide scale adoption of bank statement analysis, as suggested by Kumar ([0002]). The references cited are analogous art in the same field of financial transaction processing.
Consider claim 10, Wong discloses the machine learning model (machine learning model to detect anomalies within transactions, [0112]).
Wong, Messing, and Butvinik do not specifically mention the machine learning model comprises a named entity recognition model.
Kumar discloses a named entity recognition model (unlabeled text is analyzed with Named Entity Recognition, [0016]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik such that the machine learning model of Wong comprises a named entity recognition model as in Kumar for reasons similar to those for claim 8.
Consider claim 19, Wong discloses the machine learning model (machine learning model to detect anomalies within transactions, [0112]).
Wong, Messing, and Butvinik do not specifically mention the machine learning model comprises a named entity recognition model.
Kumar discloses a named entity recognition model (unlabeled text is analyzed with Named Entity Recognition, [0016]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the invention of Wong, Messing, and Butvinik such that the machine learning model of Wong comprises a named entity recognition model as in Kumar for reasons similar to those for claim 8.
Conclusion
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 Jesse Pullias whose telephone number is 571/270-5135. The examiner can normally be reached on M-F 8:00 AM - 4:30 PM. The examiner’s fax number is 571/270-6135.
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, Andrew Flanders can be reached on 571/272-7516.
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.
/Jesse S Pullias/
Primary Examiner, Art Unit 2655 03/23/26