DETAILED ACTION
Response to Amendment
1. This office action is in response to applicant’s communication filed on 12/26/2025 in response to the non-final office action mailed on 10/01/2025. The Applicant’s remarks and amendments to the claims and/or the specification were considered with the results as follows.
2. In response to the last Office Action, claims 17, 26 and 27 are amended. Claim 10 is canceled. Claim 21 is added. As a result, claims 1-9 and 11-21 are pending in this office action.
Response to Arguments
3. Applicant's arguments with respect to 35 USC 101 have been fully considered but are not persuasive and details are as follow:
Applicant’s argument stated as “The amended independent claims require that candidate training data set is to be evaluated prior to updating AI models using the candidate training data set. This approach prevents the worthless expenditures of computation resources and thereby improves the operation of the computing systems through increased availability of computing resources in the system…”.
In response to Applicant’s argument, the Examiner disagrees, because adding the feature “evaluating the candidate training data set prior to update AI models” is still directed to a judicial exception (i.e., an abstract idea) without significantly more since observing a candidate training data set prior to updating AI models can be performed in the human mind. The additional feature merely uses a computer/device as a tool to evaluate the candidate training data using a series of data gathering steps. The data gathering/retrieval steps are insignificant extra-solution activity; thus, the judicial exception is not integrated into a practical application. Evaluating a training data set prior to updating AI models does not appear to be improvements to the functioning of a computer or to any other technology or technical field. The additional elements do not amount to significantly more than the above-identified judicial exception (the abstract idea).
4. Applicant’s arguments with respect to 35 USC 102 have been fully considered but are moot in view of new ground(s) of rejection.
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.
5. Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter.
Claim 1 is rejected under 35 U.S.C 101 because the claimed invention is directed to a judicial exception (i.e., an abstract idea) without significantly more. Claim 1 is directed to the abstract idea of managing an artificial intelligence AI model, as explained in detail below. The claim does not include elements that are sufficient to amount to significantly more than the judicial exception because the elements can be concepts performed in the human mind which do not add meaningful limits to practicing the abstract idea.
Claim 1 recites a method comprising at least in part:
obtaining a candidate training data set usable to update an instance of the AI model (e.g., observing a candidate training data set can be performed in the human mind);
prior to updating the instance of the AI model using the candidate training data set;
identifying a historical training data set, the historical training data set being obtained prior to the candidate training data set and the historical training data set already having been used to train the instance of the AI model (e.g., identifying a historical training data set can be performed in the human mind including observation and evaluation);
performing an analysis of the candidate training data set and the historical training data set to obtain a score reflecting a likelihood that the candidate training data set comprises poisoned training data (e.g., identifying poisoned training data based on evaluating the historical training data set and the candidate training data set to obtain a score value can be performed in the human mind including observation, evaluation and judgment using pen and paper);
making a first determination regarding whether the score exceeds a score threshold; in a first instance of the first determination in which the score exceeds the score threshold, treating the candidate training data set as comprising poisoned training data (e.g., identifying poisoned training data in the candidate training data set when the score value is greater than a threshold value can be performed in the human mind including observation, evaluation, judgment and opinion);
and in a second instance of the first determination in which the score does not exceed the score threshold, treating the candidate training data set as not comprising poisoned training data (e.g., identifying good candidate training data set when the score value is not greater than the threshold value can be performed in the human mind including observation, evaluation and judgment and opinion).
Claim 1 as it is recited falls within one of the groupings of abstract ideas [e.g., mental process] enumerated in the 2019 PEG. The recited concept can be performed in human mind including an observation, evaluation, judgement, opinion. That is, other than reciting an AI model to receive good candidate training data set (e.g., not comprising poisoned training data), nothing in the claim elements preclude the step from practically being performed in the mind. The AI model is recited at a high level of generality and add no more to the claimed invention than the computer components that perform an abstract idea. The additional feature merely uses a computer/device as a tool to retrieve data result after a series of data gathering steps. The data gathering/retrieval steps are insignificant extra-solution activity; thus, the judicial exception is not integrated into a practical application. The AI model does not appear to be improvements to the functioning of a computer or to any other technology or technical field. Thus, taken alone, the additional elements do not amount to significantly more than the above-identified judicial exception (the abstract idea). Looking at the limitation 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 a computer or improves any other technology. Their collective functions merely provide conventional computer implementation. Therefore, claim 1 is not patent eligible.
Claim 2 recite similar features as claim 1, is also fall within the mental processes abstract ideas enumerated in the 2019 PEG. The recited concept can be performed in human mind including an observation, evaluation, judgement, opinion. Claim 2 further defines the score of the candidate data comprises poisoned training data. There is no additional feature amount to significantly more than the above-identified judicial exception (the abstract idea). Looking at the limitation 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 a computer or improves any other technology. Their collective functions merely provide conventional computer implementation. Therefore, claim 2 is not patent eligible.
Claims 3-7 recite similar features as claim 1, are also fall within the mental processes abstract ideas enumerated in the 2019 PEG. The recited concept can be performed in human mind including an observation, evaluation, judgement, opinion. Claims 3-7 further recite performing a cluster analysis of training data set to obtain a set of clusters and making determination regarding whether data value falls within the set of clusters and identifying a similarity measure using a Euclidean distance, a cosine similarity; and a Manhattan distance. There is no additional feature amount to significantly more than the above-identified judicial exception (the abstract idea). Looking at the limitation 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 a computer or improves any other technology. Their collective functions merely provide conventional computer implementation. Therefore, claims 3-7 are not patent eligible.
Claims 8 and 9 recite similar features as claim 1, are also fall within the mental processes abstract ideas enumerated in the 2019 PEG. The recited concept can be performed in human mind including an observation, evaluation, judgement, opinion. Claims 8 and 9 further recite the action would need to take when the candidate training containing or not containing poisoned training data. There is no additional feature amount to significantly more than the above-identified judicial exception (the abstract idea). Looking at the limitation 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 a computer or improves any other technology. Their collective functions merely provide conventional computer implementation. Therefore, claims 8 and 9 are not patent eligible.
Claim 11 is rejected under 35 U.S.C 101 because the claimed invention is directed to a judicial exception (i.e., an abstract idea) without significantly more. Claim 11 is directed to the abstract idea of managing an artificial intelligence AI model, as explained in detail below. The claim does not include elements that are sufficient to amount to significantly more than the judicial exception because the elements can be concepts performed in the human mind which do not add meaningful limits to practicing the abstract idea.
Claim 11 recites an article of manufacture (e.g., a machine readable medium) comprising at least in part:
obtaining a candidate training data set usable to update an instance of the AI model (e.g., observing a candidate training data set can be performed in the human mind); prior to updating the instance of the AI model using the candidate training data set; identifying a historical training data set, the historical training data set being obtained prior to the candidate training data set and the historical training data set already having been used to train the instance of the AI model (e.g., identifying a historical training data set can be performed in the human mind including observation and evaluation);
performing an analysis of the candidate training data set and the historical training data set to obtain a score reflecting a likelihood that the candidate training data set comprises poisoned training data (e.g., identifying poisoned training data based on evaluating the historical training data set and the candidate training data set to obtain a score value can be performed in the human mind including observation, evaluation and judgment using pen and paper);
making a first determination regarding whether the score exceeds a score threshold; in a first instance of the first determination in which the score exceeds the score threshold, treating the candidate training data set as comprising poisoned training data (e.g., identifying poisoned training data in the candidate training data set when the score value is greater than a threshold value can be performed in the human mind including observation, evaluation, judgment and opinion);
and in a second instance of the first determination in which the score does not exceed the score threshold, treating the candidate training data set as not comprising poisoned training data (e.g., identifying good candidate training data set when the score value is not greater than the threshold value can be performed in the human mind including observation, evaluation and judgment and opinion).
Claim 11 as it is recited falls within one of the groupings of abstract ideas [e.g., mental process] enumerated in the 2019 PEG. The recited concept can be performed in human mind including an observation, evaluation, judgement, opinion. That is, other than reciting a processor performs operations for managing an artificial intelligence AI model to receive good candidate training data set (e.g., not comprising poisoned training data), nothing in the claim elements preclude the step from practically being performed in the mind. The processor and the AI model are recited at a high level of generality and add no more to the claimed invention than the computer components that perform an abstract idea. The additional feature merely uses a computer/device as a tool to retrieve data result after a series of data gathering steps. The data gathering/retrieval steps are insignificant extra-solution activity; thus, the judicial exception is not integrated into a practical application. The processor and the AI model do not appear to be improvements to the functioning of a computer or to any other technology or technical field. Thus, taken alone, the additional elements do not amount to significantly more than the above-identified judicial exception (the abstract idea). Looking at the limitation 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 a computer or improves any other technology. Their collective functions merely provide conventional computer implementation. Therefore, claim 11 is not patent eligible.
Claim 12 recite similar features as claim 1, is also fall within the mental processes abstract ideas enumerated in the 2019 PEG. The recited concept can be performed in human mind including an observation, evaluation, judgement, opinion. Claim 12 further defines the score of the candidate data comprises poisoned training data. There is no additional feature amount to significantly more than the above-identified judicial exception (the abstract idea). Looking at the limitation 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 a computer or improves any other technology. Their collective functions merely provide conventional computer implementation. Therefore, claim 12 is not patent eligible.
Claims 13-15 recite similar features as claim 12, are also fall within the mental processes abstract ideas enumerated in the 2019 PEG. The recited concept can be performed in human mind including an observation, evaluation, judgement, opinion. Claims 3-6 further recite performing a cluster analysis of training data set to obtain a set of clusters and making determination regarding whether data value falls within the set of clusters. There is no additional feature amount to significantly more than the above-identified judicial exception (the abstract idea). Looking at the limitation 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 a computer or improves any other technology. Their collective functions merely provide conventional computer implementation. Therefore, claims 13-15 are not patent eligible.
Claim 16 recites a machine (e.g., a data processing system) comprising at least in part:
obtaining a candidate training data set usable to update an instance of the AI model (e.g., observing a candidate training data set can be performed in the human mind);
prior to updating the instance of the AI model using the candidate training data set;
identifying a historical training data set, the historical training data set being obtained prior to the candidate training data set and the historical training data set already having been used to train the instance of the AI model (e.g., identifying a historical training data set can be performed in the human mind including observation and evaluation);
performing an analysis of the candidate training data set and the historical training data set to obtain a score reflecting a likelihood that the candidate training data set comprises poisoned training data (e.g., identifying poisoned training data based on evaluating the historical training data set and the candidate training data set to obtain a score value can be performed in the human mind including observation, evaluation and judgment using pen and paper);
making a first determination regarding whether the score exceeds a score threshold; in a first instance of the first determination in which the score exceeds the score threshold, treating the candidate training data set as comprising poisoned training data (e.g., identifying poisoned training data in the candidate training data set when the score value is greater than a threshold value can be performed in the human mind including observation, evaluation, judgment and opinion);
and in a second instance of the first determination in which the score does not exceed the score threshold, treating the candidate training data set as not comprising poisoned training data (e.g., identifying good candidate training data set when the score value is not greater than the threshold value can be performed in the human mind including observation, evaluation and judgment and opinion).
Claim 16 as it is recited falls within one of the groupings of abstract ideas [e.g., mental process] enumerated in the 2019 PEG. The recited concept can be performed in human mind including an observation, evaluation, judgement, opinion. That is, other than reciting a processor and a memory perform operations for managing an artificial intelligence AI model to receive good candidate training data set (e.g., not comprising poisoned training data), nothing in the claim elements preclude the step from practically being performed in the mind. The processor, memory and the AI model are recited at a high level of generality and add no more to the claimed invention than the computer components that perform an abstract idea. The additional feature merely uses a computer/device as a tool to retrieve data result after a series of data gathering steps. The data gathering/retrieval steps are insignificant extra-solution activity; thus, the judicial exception is not integrated into a practical application. The processor, the memory and the AI model do not appear to be improvements to the functioning of a computer or to any other technology or technical field. Thus, taken alone, the additional elements do not amount to significantly more than the above-identified judicial exception (the abstract idea). Looking at the limitation 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 a computer or improves any other technology. Their collective functions merely provide conventional computer implementation. Therefore, claim 16 is not patent eligible.
Claim 17 recite similar features as claim 1, is also fall within the mental processes abstract ideas enumerated in the 2019 PEG. The recited concept can be performed in human mind including an observation, evaluation, judgement, opinion. Claim 17 further defines the score of the candidate data comprises poisoned training data. There is no additional feature amount to significantly more than the above-identified judicial exception (the abstract idea). Looking at the limitation 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 a computer or improves any other technology. Their collective functions merely provide conventional computer implementation. Therefore, claim 17 is not patent eligible.
Claims 18-20 recite similar features as claim 17, are also fall within the mental processes abstract ideas enumerated in the 2019 PEG. The recited concept can be performed in human mind including an observation, evaluation, judgement, opinion. Claims 18-20 further recite performing a cluster analysis of training data set to obtain a set of clusters and making determination regarding whether data value falls within the set of clusters. There is no additional feature amount to significantly more than the above-identified judicial exception (the abstract idea). Looking at the limitation 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 a computer or improves any other technology. Their collective functions merely provide conventional computer implementation. Therefore, claims 18-20 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.
invention.
Claims 1, 2, 8, 9, 11-14, 16, 17 and 21 are rejected under 35 U.S.C. 103 as being unpatentable by Baracaldo et al. (US 2020/0019821 A1), hereinafter Baracaldo and in view of Liu et al. (US 2023/0274003 A1), hereinafter Liu.
As to claims 1, 11 and 16, Baracaldo discloses a method of managing an artificial intelligence (AI) model (See para. [0047], managing a ML model by identifying poisoned data and eliminating poisoned data from the training set to prevent or mitigate an impact of a poison attack), the method comprising:
obtaining a candidate training data set usable to update an instance of the AI model (See para. [0081] and para. [0083] and Figure 3, obtaining an untrusted training dataset 302, the untrusted dataset 302 can be applied to an ML model after a performance metric analysis); […]
identifying a historical training data set (See para. [0079], para. [0082] and Figure 3, identifying a provenance training dataset 304), the historical training data set being obtained prior to the candidate training data set and the historical training data set already having been used to train the instance of the AI model (See para. [0078]-para. [0082] and Figure 3, the provenance features of the provenance training dataset 304 are selected for evaluation of corresponding data points to determine whether the untrusted training dataset 302 contains poisoned data);
performing an analysis of the candidate training data set and the historical training data set to obtain a score reflecting a likelihood that the candidate training data set comprises poisoned training data (See para. [0082] and para. [0084] and para. [0093] and Figure 10, performing an analysis between the untrusted dataset 302 with a corresponding provenance record 304a…304n of the provenance 304, note in para. [0086], the user can choose a threshold depending on his/her needs, an untrusted segment was deemed poisonous if the performance metric was greater than the mean, plus one standard deviation of the change in performance during a calibration trials)
making a first determination regarding whether the score exceeds a score threshold; in a first instance of the first determination in which the score exceeds the score threshold, treating the candidate training data set as comprising poisoned training data (See para. [0086] and para. [0091]-para. [0094] and Figure 10, the user can choose a threshold depending on his/her needs, an untrusted segment was deemed poisonous if the performance metric was greater than the mean, plus one standard deviation of the change in performance during a calibration trials, note this threshold can be adjusted to increase precision at the expense of recall or vice versa);
and in a second instance of the first determination in which the score does not exceed the score threshold, treating the candidate training data set as not comprising poisoned training data (See para. [0086] and para. [0091]-para. [0094] and Figure 10, an untrusted segment was not deemed poisonous if the performance metric was not greater than the mean, plus one standard deviation of the change in performance during a calibration trials).
Baracaldo does not explicitly disclose evaluate the candidate training data set prior to updating AI models using the candidate training data set.
Liu discloses evaluate the candidate training data set prior to updating AI models using the candidate training data set (See para. [0026], the data veracity assessment component performs data veracity assessment of the training data before the machine learning model using the training data, the data veracity assessment component determines whether the training data has been poisoned).
Therefore, it would have been obvious to a person of ordinary skill in the computer art before the effective filing date of the claimed invention to modify the system of Baracaldo to evaluate the candidate training data set prior to updating AI models, taught by Liu. Skilled artisan would have been motivated to perform risk assessment based on the training data for machine learning model to implement a corrected machine learning model (See Liu, para. [0014]). In addition, all of the references (Liu, Bequet and Baracaldo) teach features that are directed to same field of endeavor, such as training AI models. This close relation between all of the references highly suggests an expectation of success.
As to claims 2, 12 and 17, Baracaldo discloses wherein a higher score indicates a higher likelihood that the candidate training data set comprises poisoned training data (See para. [0089] and para. [0142[, different performance metrics can be used for this purpose, including but not limited to F1-measure and accuracy. In line 8, E serves as a tunable parameter to determine how large the performance decrease should be to conclude a segment of data points is poisonous. Methodologies for computing E are discussed above regarding the calibration procedure and thresholds generated thereby. Algorithm 1 returns a set of tuples containing data points that are suspected of being poisonous, associated provenance signatures and corresponding expected loss in performance if the suspect data points are not filtered (i.e. removed from the untrusted dataset).
As to claim 8, Baracaldo discloses wherein treating the candidate training data as comprising poisoned training data comprises one selected from a list consisting of: removing the candidate training data set from consideration as training data for the AI model; treating the candidate training data set as being part of a malicious attack; discarding the candidate training data set; identifying a data source of the candidate training data set; and treating the data source of the candidate training data set as a potentially malicious data source (See para. [0130] and para. [0131], removing the untrusted data set that have data points are designated as poisonous, to train a final prediction model, note in para. [0047] and para. [0050], eliminating the poised training set to prevent or mitigate the impact of a poison attack includes one or more observations from one or more data sources).
As to claim 9, Baracaldo discloses treating the candidate training data set as not comprising poisoned training data comprises one selected from a list consisting of: updating the instance of the AI model using the candidate training data to obtain a new instance of the AI model; and adding the candidate training data set to the historical training data set to obtain an updated historical training data set (See para. [0130] and para. [0131], adding the untrusted data set to a final filtered training set [e.g., the training set have no data points are designated as poisonous] to train a final prediction model).
As to claim 13, Baraaldo does not explicitly disclose performing a clustering analysis of the historical training data set to obtain a set of clustering.
Liu discloses performing a clustering analysis of the historical training data set to obtain a set of clustering; identifying a first data value of the candidate training data set; making a second determining regarding whether the first data value falls within the set of clusters (See Liu, para. [0026], with an activation clustering model, to identify poisoned data in the image, audio, or sensor training data. Activation clustering includes analyzing neural network activations of training data to determine whether the training data has been poisoned, and, if so, which datapoints are poisoned. Activation weights for poisoned training data may break up into distinguishable clusters, while activation weights for clean training data may not break up into clusters, note the activation weights are pointers to identify low or dirty training data); in a first instance of the second determination in which the first data value falls within the set of clusters; increase the score; and in a second instance of the second determination in which the first data value does not fall within the set of clusters; decreasing the score; and approving the first data value for AI model training purposes (See para. Liu, [0027], If the data veracity assessment component identifies poisoned data in the training data, the data sanitization component of the assessment system may remove the identified poisoned data from the training data. In some implementations, the assessment system may provide, to the user device and/or to users associated with the machine learning model, notifications (e.g., email messages, instant messages, and/or the like) indicating that the training data includes the poisoned data, may validate the training data without the poisoned data, may backup the training data without the poisoned data, may audit the training data, and/or the like).
Therefore, it would have been obvious to a person of ordinary skill in the computer art before the effective filing date of the claimed invention to modify the system of Baracaldo to evaluate the candidate training data set prior to updating AI models and obtain a revised candidate training data set when the candidate data set has a low information value portion, taught by Liu. Skilled artisan would have been motivated to perform risk assessment based on the training data for machine learning model to implement a corrected machine learning model (See Liu, para. [0014]). In addition, all of the references (Liu, Bequet and Baracaldo) teach features that are directed to same field of endeavor, such as training AI models. This close relation between all of the references highly suggests an expectation of success.
As to claim 14, Baracaldo does not explicitly for each of the clusters of a set of clusters making a comparison between the first data value and a bounding area of a respective cluster.
Liu discloses for each of the clusters of the set of clusters: making a comparison between the first data value and a bounding area of a respective cluster to determine whether the first data value falls within the respective cluster; and in an instance of the comparison in which the first data value falls within the respective cluster, concluding that the first data value falls within the set of clusters (See Liu, para. [0026], with an activation clustering model, to identify poisoned data in the image, audio, or sensor training data. Activation clustering includes analyzing neural network activations of training data to determine whether the training data has been poisoned, and, if so, which datapoints are poisoned. Activation weights for poisoned training data may break up into distinguishable clusters, while activation weights for clean training data may not break up into clusters, note the activation weights are pointers to identify low or dirty training data).
Therefore, it would have been obvious to a person of ordinary skill in the computer art before the effective filing date of the claimed invention to modify the system of Baracaldo to evaluate the candidate training data set prior to updating AI models and obtain a revised candidate training data set when the candidate data set has a low information value portion, taught by Liu. Skilled artisan would have been motivated to perform risk assessment based on the training data for machine learning model to implement a corrected machine learning model (See Liu, para. [0014]). In addition, all of the references (Liu, Bequet and Baracaldo) teach features that are directed to same field of endeavor, such as training AI models. This close relation between all of the references highly suggests an expectation of success.
As to claim 21, Baracaldo in view of Liu discloses treating the candidate training data set as not comprising poisoned training data comprises: for each portion of the candidate training data set: determining whether the portion of the candidate training data set falls within a
cluster of the historical training data set obtained during the performing of the analysis to identify a low information value portion of the candidate training data set (See Liu, para. [0026], with an activation clustering model, to identify poisoned data in the image, audio, or sensor training data. Activation clustering includes analyzing neural network activations of training data to determine whether the training data has been poisoned, and, if so, which datapoints are poisoned. Activation weights for poisoned training data may break up into distinguishable clusters, while activation weights for clean training data may not break up into clusters, note the activation weights are pointers to identify low or dirty training data) removing the low information value portion of the candidate training data set from the candidate training data set prior to obtain a revised candidate training data set; and obtaining a new instance of the AI model using the revised candidate training data set (See para. Liu, [0027], If the data veracity assessment component identifies poisoned data in the training data, the data sanitization component of the assessment system may remove the identified poisoned data from the training data. In some implementations, the assessment system may provide, to the user device and/or to users associated with the machine learning model, notifications (e.g., email messages, instant messages, and/or the like) indicating that the training data includes the poisoned data, may validate the training data without the poisoned data, may backup the training data without the poisoned data, may audit the training data, and/or the like).
Therefore, it would have been obvious to a person of ordinary skill in the computer art before the effective filing date of the claimed invention to modify the system of Baracaldo to evaluate the candidate training data set prior to updating AI models and obtain a revised candidate training data set when the candidate data set has a low information value portion, taught by Liu. Skilled artisan would have been motivated to perform risk assessment based on the training data for machine learning model to implement a corrected machine learning model (See Liu, para. [0014]). In addition, all of the references (Liu, Bequet and Baracaldo) teach features that are directed to same field of endeavor, such as training AI models. This close relation between all of the references highly suggests an expectation of success.
Claim 10 is rejected under 35 U.S.C. 103 as being unpatentable over Baracaldo (US 2020/0019821 A1) and in view of Liu (US 2023/0274003 A1) and further in view of Bequet (US 2019/0384790 A1).
As to claim 10, Baracaldo does not explicitly disclose obtaining the candidate training data set: making an identification that a re-training condition is met for the AI model, wherein the candidate training data set is obtained in response to the identification.
Bequet discloses obtaining the candidate training data set: making an identification that a re-training condition is met for the AI model, wherein the candidate training data set is obtained in response to the identification (See para. [0019], para. [0536], para. [0543] and Figure 26F, obtaining a new data object within the training set in response to a condition has been met for retraining).
Therefore, it would have been obvious to a person of ordinary skill in the computer art before the effective filing date of the claimed invention to modify the system of Baracaldo to configure a re-training condition, taught by Bequet. Skilled artisan would have been motivated to improve accuracy of the AI model (See Bequet, para. [0536]). In addition, all of the references (Liu, Bequet and Baracaldo) teach features that are directed to same field of endeavor, such as training AI models. This close relation between all of the references highly suggests an expectation of success.
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 YUK TING CHOI whose telephone number is (571)270-1637. The examiner can normally be reached Monday-Friday 9am-6pm.
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, AMY NG can be reached at 5712701698. 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.
/YUK TING CHOI/Primary Examiner, Art Unit 2164