DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
This Office Action is in response to the amendment filed June 26, 2025. Claims 1, 5, and 9 have been amended. Claims 1, 2-3, 5-7, 9-11, and 13-15 are pending.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1, 2-3, 5-7, 9-11, and 13-15 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. Claims 1, 5, and 9 are directed to an information processing non-transitory computer-readable medium, method and apparatus for performing steps for acquiring a target sentence; acquiring word strings that indicates a context relating to the target sentence using a corpus; generating a plurality of combined sentences by combining each of the acquired word strings with the target sentence; outputting a plurality of probability distributions concerning the word strings by inputting each of the plurality of combined sentences generated at generating and the target sentence into a language model; calculating, based on a difference between each of the plurality of the probability distributions output at outputting, confidence in output when the target sentence is input into the language model; and outputting, based on the calculated confidence, an output result when the target sentence is input into the language model, wherein the program, method or apparatus is summarizing news articles, responding in interactive systems or translating in translation systems. The limitation for acquiring a target sentence is a data gathering step that can be achieved by a person hearing a sentence or reading the sentence from a document source. The limitation for acquiring, word strings that indicated a context relating to a target sentence using a corpus is a data gathering step that can be achieved by a person, accessing a list of word stings or, using pen and paper, generating a list of word strings that meet a similarity criteria based on context to sentences within an available corpus. The feature for “generating a plurality of combined sentences…”, can be achieved by the person, using pen and paper, rewriting/reorganizing the acquired word strings with the target sentence to create a new set of desired sentences. The feature for “outputting a plurality of probability distributions concerning the word stings by inputting each of a plurality of combined sentences…..”can be achieved by the person, using pen and paper, pairing the strings with the target sentence, and using principles and rules of natural language processing representing learning techniques, and determining an output of representing a relationship of the strings and the target sentence by using mathematical calculations based on probability distributions. The step for “calculating.....confidence..” is a mathematical processing step that can be achieved by the person, using pen and paper, calculating probability distributions of the determined outputs. The step for “outputting…an output result…” can be achieved by the person speaking an output, or using pen and paper, writing the output. The step for “…summarizing news articles….” can be achieved by the person gathering and observing articles from a generic interactive/translation system (data gathering step), reading the article, and using mental processing or pen and paper, generate or write a brief summary of the article that was read. The recited limitations are directed a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of the generic computer, apparatus, computer readable medium, processor, control unit, and/or generic computer components. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claims recite an abstract idea.
This judicial exception is not integrated into a practical application because the recited generic computer, apparatus, computer readable medium, processor, control unit, and/or generic computer components amounts to no more than mere instructions to apply the exception using generic computer components. Accordingly, the elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claims are directed to an abstract idea. The claims are not patent eligible.
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because, as indicated with respect to integration of the abstract idea into a practical application, the additional elements of the generic computer, apparatus, computer readable medium, processor, control unit, and/or generic computer components to perform the various steps amounts to no more than mere instructions to apply the exception using generic computer components. Mere instructions to apply an exception using generic computer components cannot provide an inventive concept. The claims are not patent eligible.
Dependent claims 2-3, 6-7,10-11, and 13-15 do not integrate the judicial exception into a practical application and do not include additional elements that are sufficient to amount to significantly more than the judicial exception. The limitations of the dependent claims are directed to steps of organizing or manipulating sentences/text/strings and natural language data similarities via mathematical calculations, threshold comparisons, and/or natural language processing principles.
Claim Rejections - 35 USC § 103
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Claims 1, 2-3, 5-7, and 9-11 are rejected under 35 U.S.C. 103 as being unpatentable over Nishida et al (US Patent Application Publication No. 2022/0138601), hereinafter Nishida, in view of Feigenblat et al (US Patent No. 10,902,191), hereinafter Feigenblat.
Nishida teaches a question-answering apparatus, method and program. Regarding claims 1, 5, and 9, Nishida teaches a non-transitory computer-readable recording medium having stored thereon a program, method, and apparatus [Figs 1, 4, 7A, 7B, 8, 11A, 11B, para 0061] to perform a process [para 0117-0170] comprising: acquiring a target sentence [Fig 7A -S301; para 0080 – acquire one item in the data]; acquiring a plurality of word strings that indicate a context relating to the target sentence using a corpus [para 0036-0038 –data related to the Olympics indicates a context; 0124; 0164-0167 – inputs include a question, a document set, an answer style, and a right answer sentence according to the answer style; para 0117 -- document set is adequate for generating answer sentences are closely related to accuracy and the like of the generated answer sentences – where the document set is a corpus and closely related accuracy provides a form of similarity]; generating a plurality of combined sentences by combining each of the acquired word strings with the target sentence [para 0036-0038 –data related to the Olympics indicates a context; 0094 – generated context vectors; 0124; 0164-0167 – inputs include a question, a document set, an answer style, and a right answer sentence according to the answer style – combined data; para 0117 -- document set is adequate for generating answer sentences are closely related to accuracy and the like of the generated answer sentences]; outputting a plurality of probability distributions concerning the word strings [Fig 7B – S313-S317; para 0093-0105 – calculate probability of generating the word …..where the process is performed for each item in the training set] by inputting each of a plurality of combined sentences for which each of the acquired word strings is combined with the target sentence, and the target sentence into a language model [para 0117-0170; 0126-0127 – input data processed with neural network]; calculating, based on a difference between each of the plurality of the probability distributions output ,confidence in output when the target sentence is input into the language model [para 0040-0042; 0126-0127; 0158-0161 – document fitness, right document fitness, the answerableness a, and right answer ability]; and outputting, based on the calculated confidence, an output result when the target sentence is input into the language model [para 0113; 0170 – outputs]. Nishida suggests that generation of an answer sentence by reading comprehension is a summary of a question and document set [para 0003] and specifically teaches the document set can be sets of news articles [para 0036-0037]. Nishida fails to specifically teach the invention is summarizing news articles responding in interactive systems. Feigenblat teaches natural language processing techniques for generating a document summary, which generates a summary and then edits the summary with abstracted sentences, so as to improve the summary [abstract; col. 2, line 24 to col. 3, line 23] and teaches the system can be used in a cloud environment for virtual education classroom delivery [col. 14, lines 1-2 – “interactive system”]. One having ordinary skill in the art at the time of the invention would have recognized the advantages of implementing the summarizing techniques suggested by Feigenblat, in the system of Nishida, so as to provide improved overall readability, accuracy and minimal redundancy in the generated output, as taught by Feigenblat, and thereby provide an enhanced and improve user experience.
Regarding claims 2, 6, and 10, the combination of Nishida and Feigenblat teaches calculating calculates variance based on each distribution and assumes the calculated variance to be an index value of the confidence [para 0040-0042; 0126-0127; 0158-0161 – document fitness, right document fitness, the answerableness a, and right answer ability].
Regarding claims 3, 7, and 11, combination of Nishida and Feigenblat teaches calculating calculates a distance based on each distribution and assumes the calculated distance to be an index value of the confidence [para 0040-0042; 0126-0127; 0158-0161 – document fitness, right document fitness, the answerableness a, and right answer ability].
.
Claims 13-15 are rejected under 35 U.S.C. 103 as being unpatentable over Nishida in view of Feigenblat, and further in view of Block et al (US Patent NO 10,282,678), hereinafter Block.
Regarding claims 13-15, Nishida fails to specifically teach comparing the calculated confidence to a predetermined threshold; and in response to the calculated confidence exceeds the predetermined threshold, outputting the output result, and otherwise not outputting the output result. In a similar field of endeavor, Block teaches a QA system that generates answers for input questions that outputs the most probable answers according to a confidence measure [col. 6, lines 52-62] and utilizes predetermined threshold comparisons to output the most likely answer [col. 13, lines 25-40]. One having ordinary skill in the art at the time of the invention would have recognized the advantages of implementing the confidence comparison threshold processing, suggested by Block, in the system of Nishida/Feigenblat, for the purpose of determining the answer result is the right, as taught by Block [Abstract; col.2, lines 5-8] and thereby provide an improved question-answering system, as suggested by Block [col. 1, lines 7-10].
Response to Arguments
Applicant's arguments filed June 26, 2025 have been fully considered but they are not persuasive.
Applicant argues the pending independent claims are not directed to an abstract idea but to a practical application that improves the functioning of a computer system. Applicant argues the claims recite a specific technical process that is inextricably tied to computer technology and cannot be practically performed by a human mind or with pen and paper. The Examiner respectfully disagrees. The Examiner notes, as indicated in the rejection above, the identified judicial exception is not integrated into a practical application because the recited generic computer, apparatus, computer readable medium, processor, control unit, and/or generic computer components amounts to no more than mere instructions to apply the exception using generic computer components. Applicant’s limitation for the “summarizing news articles, responding in interactive systems or translating in translation systems” is merely another limitation that recites an abstract idea. As indicated in the rejection above, this step can be achieved by the person gathering and observing articles from a generic interactive/translation system (data gathering step), reading the article, and using mental processing or pen and paper, generate or write a brief summary of the article that was read. The limitation does not provide or describe how the summary is obtained, or how the recited limitations are specifically implemented or performed to be tied to the practical application to achieve any process that is applied in the practical application, or what specific (non-generic) device/additional elements are used generating the summary. Accordingly, the elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claims are directed to an abstract idea. The claims are not patent eligible.
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because, as indicated with respect to integration of the abstract idea into a practical application, the additional elements of the generic computer, apparatus, computer readable medium, processor, control unit, and/or generic computer components to perform the various steps amounts to no more than mere instructions to apply the exception using generic computer components. Additionally, as indicated in the rejection above, the recitation of the generic interactive/translation system is merely a data gathering step used to obtain the article that is then summarized via human mental processing or pen and paper. Mere instructions to apply an exception using generic computer components cannot provide an inventive concept. The recited judicial exception is not integrated or include any specific recitation or claim language directed to a specific device or additional elements to perform the summarizing news article processing.
Applicant argues "inputting each of the plurality of combined sentences... into a language model, generated by using a machine learning" and "calculating, based on a difference between of the plurality of the probability distributions... confidence" are not tasks a human can perform. The Examiner respectfully disagrees. As indicated in the rejections above, feature for “outputting a plurality of probability distributions concerning the word stings by inputting each of a plurality of combined sentences…..”can be achieved by the person, using pen and paper, pairing the strings with the target sentence, and using principles and rules of natural language processing representing learning techniques, and determining an output of representing a relationship of the strings and the target sentence by using mathematical calculations based on probability distributions. The step for “calculating.....confidence..” is a mathematical processing step that can be achieved by the person, using pen and paper, calculating probability distributions of the determined outputs. Applicant argues the language model is generated by using a machine learning and argues the steps require a processor to execute complex mathematical operations on a machine-learned model. However, the recited claims fail to recite the feature for the machine-learned model. Additionally, the language model can be represented, via pen and paper, using the rules and principles of natural language processing and the machine learning via the generic recited processor is merely a generic computer element that can be used to perform the various steps amounts to no more than mere instructions to apply the exception using generic computer components. Mere instructions to apply an exception using generic computer components cannot provide an inventive concept. The claims are not patent eligible.
Applicant argues the process as a whole is a 7-step specific implementation that improves the computer's functionality by making the output of a language model more reliable. The Examiner respectfully disagrees. As indicated in the rejection above, the identified judicial exception is not integrated into a practical application because the recited generic computer, apparatus, computer readable medium, processor, control unit, and/or generic computer components amounts to no more than mere instructions to apply the exception using generic computer components. The claims fail to recite any additional elements to reflect improvement of a computer or a technological environment. The recitation of “summarizing news articles, responding in interactive systems or translating in translation systems” is merely another limitation that recites an abstract idea and amounts to linking the use of the judicial exception to a field of use.
Applicant argues the subject matter of the independent claims is similar to, for example, "Example 39- Method for Training a Neural Network." The Examiner respectfully disagrees. The claims fail to recite limitations for training a network or NLP model. The claims may provide utilizing a generic model to achieve an output, but does not recite limitations for how the model is trained. Absent any specifically recited limitation for training the model, the model functionality can be achieved and represented, via pen and paper, using the rules and principles of natural language processing.
Applicant argues Nishida does not disclose or suggest a method for calculating the confidence of an output. Applicant argues neither Nishida nor Feigenblat disclose or suggest how to improve accuracy of output confidence in language models, much less the particular limitations recited in the independent claims. The Examiner respectfully disagrees. Nishida teaches outputting a plurality of probability distributions concerning the word strings [Fig 7B – S313-S317; para 0093-0105] via the teachings to calculate probability of generating the word such that the process is performed for each item in the training set; inputting each of a plurality of combined sentences for which each of the acquired word strings is combined with the target sentence, and the target sentence into a language model via the neural network processing [para 0117-0170; 0126-0127] and further provides for calculating, based on a difference between each of the plurality of the probability distributions output ,confidence in output when the target sentence is input into the language model based on the teachings of document fitness, right document fitness, the answerableness a, and right answer ability [para 0040-0042; 0126-0127; 0158-0161]. The teachings of Nishida and Feigenblat provide adequate support for the broadly claimed limitations.
In response to applicant’s argument that there is no teaching, suggestion, or motivation to combine the references, the examiner recognizes that obviousness may be established by combining or modifying the teachings of the prior art to produce the claimed invention where there is some teaching, suggestion, or motivation to do so found either in the references themselves or in the knowledge generally available to one of ordinary skill in the art. See In re Fine, 837 F.2d 1071, 5 USPQ2d 1596 (Fed. Cir. 1988), In re Jones, 958 F.2d 347, 21 USPQ2d 1941 (Fed. Cir. 1992), and KSR International Co. v. Teleflex, Inc., 550 U.S. 398, 82 USPQ2d 1385 (2007). In this case, Feigenblat teaches natural language processing techniques for generating a document summary, which generates a summary and then edits the summary with abstracted sentences, so as to improve the summary [abstract; col. 2, line 24 to col. 3, line 23] and teaches the system can be used in a cloud environment for virtual education classroom delivery [col. 14, lines 1-2 – “interactive system”] and specifically teaches the system provides improved overall readability, accuracy and minimal redundancy in the generated output [col. 2, lines 62-64].
In response to applicant's argument that the examiner's conclusion of obviousness is based upon improper hindsight reasoning, it must be recognized that any judgment on obviousness is in a sense necessarily a reconstruction based upon hindsight reasoning. But so long as it takes into account only knowledge which was within the level of ordinary skill at the time the claimed invention was made, and does not include knowledge gleaned only from the applicant's disclosure, such a reconstruction is proper. See In re McLaughlin, 443 F.2d 1392, 170 USPQ 209 (CCPA 1971). Since Feigenblat specifically teaches the system provides improved overall readability, accuracy and minimal redundancy in the generated output [col. 2, lines 62-64], one having ordinary skill in the art at the time of the invention would have recognized the advantages of implementing the summarizing techniques suggested by Feigenblat, in the system of Nishida, so as to provide improved overall readability, accuracy and minimal redundancy in the generated output, as taught by Feigenblat, and thereby provide an enhanced and improve user experience.
The combination of the teachings of Nishida and Feigenblat provide adequate support for the broadly claimed limitations as recited in claims 1, 2-3, 5-7, and 9-11. Further the combination of the teachings of Nishida, Feigenblat and Block provide adequate support for the broadly claimed limitations as recited in claims 13-15. Accordingly, the rejections under 35 USC 103 are maintained.
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 ANGELA A ARMSTRONG whose telephone number is (571)272-7598. The examiner can normally be reached M,T,TH,F 11:30-8:00.
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, Pierre Desir can be reached at 571-272-7799. 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.
ANGELA A. ARMSTRONG
Primary Examiner
Art Unit 2659
/ANGELA A ARMSTRONG/Primary Examiner, Art Unit 2659