DETAILED ACTION
This Office action is in reply to correspondence filed 31 March 2026 in regard to application no. 18/351,667. Claim 2 has been cancelled. Claims 1 and 3-20 are pending and are considered below.
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 .
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(s) 1, 3, 5, 6, 10, 12, 13, 17 and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Pruksachatkun et al. (U.S. Publication No. 2021/0312128) in view of Bai et al. (U.S. Publication No. 2022/0067534).
In-line citations are to Pruksachatkun.
With regard to Claim 1:
Pruksachatkun teaches: A system comprising:
a memory that stores computer executable components; [0062; "memory that stores methods, codes, instructions and programs"] and
a processor that executes computer executable components stored in the memory, [0062; a "processor may execute" the instructions] wherein the computer executable components comprise:
a linking component that associates one or more unmasked elements of a logical form of a natural language text segment with one or more corresponding structured knowledge elements of a knowledge base; [0025; structured data is "associated with the text" of "medical records"; Sheet 1, Fig. 1 showing the data is in plaintext (i.e. not masked)]
a prediction component that employs a prediction model to predict one or more masked elements of the logical form based on extended context of the one or more corresponding structured knowledge elements of the knowledge base to generate one or more predicted elements; [0059; "predicting the masked words" based on contextual data; abstract; “trained models” are used] and
a rules component that determines one or more rules that describe the natural language text segment in terms of structured knowledge elements… [0040; "training the word embedding model, the sentence embedding model, and the multi-label classifier together such that the parameters of the models are updated together"; such parameters read on rules]
Pruksachatkun does not explicitly teach the rules component iteratively updates the prediction model of the prediction component based on a combined loss for each of the one or more masked elements, but it is known in the art. Bai teaches a supervised learning system [title] in which it is disclosed that “masked inputs” may be predicted “based on their left and right contexts”. [0013] A combined “loss function” based on a “shifted masked reconstruction lost” may be “combined as the loss function” to “update [a] neural model”. [abstract] Bai and Pruksachatkun are analogous art as each is directed to electronic means for managing masked data.
It would have been obvious to one of ordinary skill in the art just prior to the filing of the claimed invention to combine the teaching of Pruksachatkun with that of Bai in order to improve efficiency, as taught by Bai; [0013] further, it is simply a substitution of one known part for another with predictable results, simply managing a model in the manner of Bai rather than, or in addition to, that of Pruksachatkun; the substitution produces no new and unexpected result.
In this and the subsequent claims, referring to software components by name, such as "linking component", "prediction component" and “rules component”, is considered mere labeling and given no patentable weight.
With regard to Claim 3:
The system of claim 1, wherein the computer executable components further comprise:
a conversion component that converts the natural language text segment to the logical form. [0050; the use of the "word embedding model" and "sentence embedding model" read on this]
With regard to Claim 5:
The system of claim 1, wherein the extended context comprises one or more candidate paths between the one or more corresponding structured knowledge elements. [Sheet 4, Fig. 4]
This claim is not patentably distinct from claim 1. First, it consists entirely of nonfunctional, descriptive language, disclosing at most human interpretation of data but which imparts neither structure nor functionality to the claimed system and so is considered but given no patentable weight. Second, .as the context only "comprises" this, it can include other information, and any further processing can be based entirely on the other information. The reference is provided for the purpose of compact prosecution.
With regard to Claim 6:
The system of claim 1, wherein the extended context comprises one or more candidate structured knowledge elements associated with the one or more corresponding structured knowledge elements via known paths. [Sheet 4, Fig. 4]
This claim is not patentably distinct from claim 1. First, it consists entirely of nonfunctional, descriptive language, disclosing at most human interpretation of data but which imparts neither structure nor functionality to the claimed system and so is considered but given no patentable weight. Second, as the context only "comprises" this, it can include other information, and any further processing can be based entirely on the other information. The reference is provided for the purpose of compact prosecution.
With regard to Claim 10:
Pruksachatkun teaches: A computer-implemented method, comprising:
associating, by a system operably coupled to a processor, [0062; a "processor may execute" instructions] one or more unmasked elements of a logical form of a natural language text segment with one or more corresponding structured knowledge elements of a knowledge base; [0025; structured data is "associated with the text" of "medical records"; Sheet 1, Fig. 1 showing the data is in plaintext (i.e. not masked)]
predicting, by the system, via a prediction model, one or more masked elements of the logical form based on extended context of the one or more corresponding structured knowledge elements of the knowledge base to generate one or more predicted elements… [0059; "predicting the masked words" based on contextual data; abstract; trained models are used]
Pruksachatkun does not explicitly teach iteratively updating, by the system, the prediction model of the prediction component based on a combined loss for each of the one or more masked elements, but it is known in the art. Bai teaches a supervised learning system [title] in which it is disclosed that “masked inputs” may be predicted “based on their left and right contexts”. [0013] A combined “loss function” based on a “shifted masked reconstruction lost” may be “combined as the loss function” to “update [a] neural model”. [abstract] Bai and Pruksachatkun are analogous art as each is directed to electronic means for managing masked data.
It would have been obvious to one of ordinary skill in the art just prior to the filing of the claimed invention to combine the teaching of Pruksachatkun with that of Bai in order to improve efficiency, as taught by Bai; [0013] further, it is simply a substitution of one known part for another with predictable results, simply managing a model in the manner of Bai rather than, or in addition to, that of Pruksachatkun; the substitution produces no new and unexpected result.
With regard to Claim 12:
The computer-implemented method of claim 10, wherein the extended context comprises one or more candidate paths between the one or more corresponding structured knowledge elements. [Sheet 4, Fig. 4]
This claim is not patentably distinct from claim 10. First, it consists entirely of nonfunctional, descriptive language, disclosing at most human interpretation of data but which imparts neither structure nor functionality to the claimed method and so is considered but given no patentable weight. Second, as the context only "comprises" this, it can include other information, and any further processing can be based entirely on the other information. The reference is provided for the purpose of compact prosecution.
With regard to Claim 13:
The computer-implemented method of claim 10, wherein the extended context comprises one or more candidate structured knowledge elements associated with the one or more corresponding structured knowledge elements via known paths. [Sheet 4, Fig. 4]
This claim is not patentably distinct from claim 10. First, it consists entirely of nonfunctional, descriptive language, disclosing at most human interpretation of data but which imparts neither structure nor functionality to the claimed method and so is considered but given no patentable weight. Second, as the context only "comprises" this, it can include other information, and any further processing can be based entirely on the other information. The reference is provided for the purpose of compact prosecution.
With regard to Claim 17:
Pruksachatkun teaches: A computer program product facilitating knowledge acquisition for a knowledge base, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor [0062; "memory that stores methods, codes, instructions and programs" for execution by a "processor"] to cause the processor to:
associate one or more unmasked elements of a logical form of a natural language text segment with one or more corresponding structured knowledge elements of the knowledge base; [0025; structured data is "associated with the text" of "medical records"; Sheet 1, Fig. 1 showing the data is in plaintext (i.e. not masked)]
predict, via a prediction model, one or more masked elements of the logical form based on extended context of the one or more corresponding structured knowledge elements of the knowledge base to generate one or more predicted elements… [0059; "predicting the masked words" based on contextual data; abstract; trained models are used]
Pruksachatkun does not explicitly teach to iteratively update the prediction model of the prediction component based on a combined loss for each of the one or more masked elements, but it is known in the art. Bai teaches a supervised learning system [title] in which it is disclosed that “masked inputs” may be predicted “based on their left and right contexts”. [0013] A combined “loss function” based on a “shifted masked reconstruction lost” may be “combined as the loss function” to “update [a] neural model”. [abstract] Bai and Pruksachatkun are analogous art as each is directed to electronic means for managing masked data.
It would have been obvious to one of ordinary skill in the art just prior to the filing of the claimed invention to combine the teaching of Pruksachatkun with that of Bai in order to improve efficiency, as taught by Bai; [0013] further, it is simply a substitution of one known part for another with predictable results, simply managing a model in the manner of Bai rather than, or in addition to, that of Pruksachatkun; the substitution produces no new and unexpected result.
With regard to Claim 19:
The computer program product of claim 17, wherein the program instructions are further executable by the processor to cause the processor to:
determine one or more rules that describe the natural language text segment in terms of structured knowledge elements. [0040; "training the word embedding model, the sentence embedding model, and the multi-label classifier together such that the parameters of the models are updated together"; such parameters read on rules]
Claim(s) 4, 11 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Pruksachatkun et al. in view of Bai et al. further in view of Kulkarni et al. (U.S. Patent No. 11,748,634, filed 19 October 2020).
These claims are similar so are analyzed together.
With regard to Claim 4:
The system of claim 1, wherein computer executable components further comprise:
a masking component that masks one or more elements of the logical form of a natural language text segment resulting in the one or more masked elements and the one or more unmasked elements.
With regard to Claim 11:
The computer-implemented method of claim 10, further comprising:
masking, by the system, one or more elements of the logical form of a natural language text segment resulting in the one or more masked elements and the one or more unmasked elements; and
determining, by the system, one or more rules that describe the natural language text segment in terms of structured knowledge elements. [0040; "training the word embedding model, the sentence embedding model, and the multi-label classifier together such that the parameters of the models are updated together"; such parameters read on rules]
With regard to Claim 18:
The computer program product of claim 17, wherein the program instructions are further executable by the processor to cause the processor to: mask one or more elements of the logical form of a natural language text segment resulting in the one or more masked elements and the one or more unmasked elements.
Pruksachatkun and Bai teach the system of claim 1, method of claim 10, and computer program product of claim 17, including that data may be masked and the determination of rules as cited above, but do not explicitly teach masking the data, but it is known in the art. Kulkarni teaches a search system using machine learning. [title] It performs "masking a portion of event data" which may be "text" such as a "credit card number". [Col. 20, lines 4-5] Kulkarni and Pruksachatkun are analogous art as each is directed to electronic means for processing text in which some of the text may be masked.
It would have been obvious to one of ordinary skill in the art just prior to the filing of the claimed invention to combine the teaching of Kulkarni with that of Pruksachatkun and Bai in order to obfuscate sensitive data, as taught by Kulkarni; further, it is simply a substitution of one known part for another with predictable results, simply masking text as in Kulkarni rather than, or in addition to, processing already-masked text as in Pruksachatkun; the substitution produces no new and unexpected result.
Claims 7 and 20 are similar so are analyzed together.
With regard to Claim 7:
The system of claim 2, wherein the computer executable components further comprise: a scoring component that calculates estimated scores corresponding to candidate predicted elements.
With regard to Claim 20:
The computer program product of claim 19, wherein the program instructions are further executable by the processor to cause the processor to:
calculate estimated scores corresponding to candidate predicted elements.
Pruksachatkun and Bai teach the system of claim 2 and computer program product of claim 19, including the use of predicted elements as cited above, but do not explicitly teach this scoring, but it is known in the art. Cohen teaches a system for dynamically updating a user interface. [title] It can determine a "quality rating" and later determine a "modified quality rating". [0018] This may include using "a particular rule for modifying the initial quality rating" and may be based on or result in a "decrease" to the initial rating based on "structured data" and "unstructured data". [0027] The rule set may be modified during processing. [0023] Cohen and Pruksachatkun are analogous art as each is directed to electronic means of processing structured and unstructured data.
It would have been obvious to one of ordinary skill in the art just prior to the filing of the claimed invention to combine the teaching of Cohen with that of Pruksachatkun and Bai in order to improve interfaces, as taught by Cohen; [0005] further, it is simply a substitution of one known part for another with predictable results, simply computing a score (i.e. rating) as in Cohen rather than, or in addition to, the computations of Pruksachatkun; the substitution produces no new and unexpected result.
With regard to Claim 8:
The system of claim 7, wherein the computer executable components further comprise:
a loss component that determines a loss based on the estimated scores and target scores associated with the one or more masked elements. [Cohen, 0027 as cited above; a decrease reads on a loss]
With regard to Claim 9:
The system of claim 8, wherein the rules component iteratively updates the one or more rules based on the loss. [Cohen, 0023 as cited above]
Claim(s) 14-16 are rejected under 35 U.S.C. 103 as being unpatentable over Pruksachatkun et al. in view of Bai et al. further in view of Kulkarni et al. further in view of Cohen et al.
With regard to Claim 14:
The computer-implemented method of claim 11, further comprising: calculating, by the system, estimated scores corresponding to candidate predicted elements.
Pruksachatkun, Bai and Kulkarni teach the method of claim 11, including the use of predicted elements as cited above, but does not explicitly teach this scoring, but it is known in the art. Cohen teaches a system for dynamically updating a user interface. [title] It can determine a "quality rating" and later determine a "modified quality rating". [0018] This may include using "a particular rule for modifying the initial quality rating" and may be based on or result in a "decrease" to the initial rating based on "structured data" and "unstructured data". [0027] The rule set may be modified during processing. [0023] Cohen and Pruksachatkun are analogous art as each is directed to electronic means of processing structured and unstructured data.
It would have been obvious to one of ordinary skill in the art just prior to the filing of the claimed invention to combine the teaching of Cohen with that of Pruksachatkun, Bai and Kulkarni in order to improve interfaces, as taught by Cohen; [0005] further, it is simply a substitution of one known part for another with predictable results, simply computing a score (i.e. rating) as in Cohen rather than, or in addition to, the computations of Pruksachatkun; the substitution produces no new and unexpected result.
With regard to Claim 15:
The computer-implemented method of claim 14, further comprising:
determining, by the system, a loss based on the estimated scores and target scores associated with the one or more masked elements. [Cohen, 0027 as cited above; a decrease reads on a loss]
With regard to Claim 16:
The computer-implemented method of claim 15, further comprising:
iteratively updating, by the system, the one or more rules based on the loss. [Cohen, 0023 as cited above]
Response to Arguments
Applicant’s arguments with respect to claim(s) 1 and 3-20 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument. To the extent the arguments focus on rejections made under 35 U.S.C. § 102, these arguments are moot because no such rejection is made herein. To the extent they focus on language added by amendment, the Examiner has herein incorporated the teaching of Bai to meet the additional limitations.
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 SCOTT C ANDERSON whose telephone number is (571)270-7442. The examiner can normally be reached M-F 9:00 to 5:30.
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, Bennett Sigmond can be reached at (303) 297-4411. 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.
/SCOTT C ANDERSON/Primary Examiner, Art Unit 3694
Prosecution Insights