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 action is responsive to the claims dated 8/29/2024.
Claims 1-10 are presented for examination.
Priority
Acknowledgement is made of applicant’s claim for foreign priority under 35 U.S.C. 119 (a)-(d). The certified copy has been received for the foreign priority application no. JP 2023-144682 filed 6/9/2023.
Information Disclosure Statement
The information disclosure statements (IDS) submitted on 12/2/2024 and
8/29/2024 have been considered by the examiner.
Specification
The title of the invention is not descriptive. A new title is required that is clearly indicative of the invention to which the claims are directed.
The following title is suggested: "Information Processing Apparatus, Survey System, and Method for Adjusting a Language Model to Suit a User Based on Job-Related Data of the User".
The disclosure is objected to because of the following informalities:
(a) [0015] page 10 line 1 recites "an questionnaire", which should read "a questionnaire";
(b) [0026] page 16 line 17 recites "related to the job of the predetermined user,.", which carries an extraneous comma before the period;
(c) [0039] page 24 line 9 recites "or may be accept one or more requests", which should read "or may accept one or more requests";
(d) [0057] page 35 line 20 recites "The hierarchy information may indicates", which should read "The hierarchy information may indicate";
(e) [0058] page 36 line 2 recites "illustrated as example of the usable job-related data", which should read "illustrated as examples of the usable job-related data";
(f) [0070] page 41 line 14 recites "designates a level having predetermined relationship", which should read "designates a level having a predetermined relationship", and recites "the designating section 101A only need to designate", which should read "the designating section 101A only needs to designate";
(g) [0078] page 45 line 23 recites "and do not limited to the example illustrated", which should read "and are not limited to the example illustrated";
(h) [0090] page 52 lines 12-13 recites "the information processing apparatus information processing apparatus 1B", in which the words "information processing apparatus" are repeated;
(i) [0108] page 63 three times refers to the accepting section of the information processing apparatus 3B as "the accepting section 301", whereas FIG. 11 and [0096], [0097] and [0109] designate that accepting section by reference character 301B, and reference character 301 designates the accepting section of the information processing apparatus 3 of FIG. 3;
(j) [0118] page 67 line 22 recites "the sender is judges not rightful", which should read "the sender is judged not rightful", and recites "the processing of the flow F1b end", which should read "the processing of the flow F1b ends";
(k) [0132] page 74 line 22 recites "(YES in S35)", which should read "(YES in S35b)", because the judgment described in that sentence is made in step S35b and FIG. 13 contains no step S35;
(l) [0139] page 78 line 23 recites "the modifying section 210B accept a modification", which should read "the modifying section 210B accepts a modification";
(m) [0154] page 85 line 21 recites "This make it possible", which should read "This makes it possible";
(n) [0156] page 86 recites "the answer sent in S17a of the flow F1b of Fig. 12 in a case of NO judgment in S14a", but the flow F1b of FIG. 12 consists of steps S11b through S17b and contains no step S17a or S14a, and [0120] describes the answer informing that answering is not allowed as being sent in S17b after a NO judgment in S14b, so the passage appears to be intended to refer to S17b and S14b;
(o) [0162] page 90 line 12 recites "making changed to a condition", which should read "making changes to a condition";
(p) [0164] page 91 recites "the reliability judged in S14d of Fig. 15", but step S14d is shown in FIG. 14, not in FIG. 15;
(q) [0165] page 92 recites that "the question adding section 307B notifies the requester of the answer received in S32d", but the answer is received in step S31d of FIG. 16, as [0164] describes, and step S32d is the notifying step itself;
(r) [0166] page 92 lines 11-12 recites "After the ends of the processing", which should read "After the end of the processing";
(s) [0196] page 106 recites "note B1 or B2, in the designating process" and [0203] recites "note C1 or C2, the designating means", each of which omits the words "in which" that the parallel supplementary notes use, for example at [0188] and [0211]; and
(t) [0204] page 109 recites "An response program" and [0218] recites "an response program", each of which should read "a response program". Appropriate correction is required.
The lengthy specification has not been checked to the extent necessary to determine the presence of all possible minor errors. Applicant's cooperation is requested in correcting any errors of which applicant may become aware in the specification.
Claim Objections
Claim 4 is objected to because of the following informalities:
Claim 4 recites "each data related to a job of the predetermined user". The determiner "each" is paired with the mass noun "data", and the indefinite article in "a job" reintroduces the job of the predetermined user that claim 1, from which claim 4 depends, already recites. The limitation should read, for example, "each piece of data related to the job of the predetermined user".
Appropriate correction is required.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitations are:
(1) an intermediary apparatus for accepting a request for a survey in which answers to a predetermined question are collected, and conducting, with the information processing apparatus, negotiations for causing the language model to generate an answer to the predetermined question, recited in claim 9. The term "apparatus" is a generic placeholder, and the modifier "intermediary" identifies its task rather than implementing structure. The remaining recitations, including "with the information processing apparatus," identify the request the apparatus receives and the party with which it negotiates rather than any structure of the intermediary apparatus, which, unlike the information processing apparatus of claim 5, is not recited as comprising a processor or other structure. The claimed function is to accept a request for a survey in which answers to a predetermined question are collected, and to conduct, with the information processing apparatus, negotiations for causing the language model to generate an answer to the predetermined question. The corresponding structure is the information processing apparatus 3 (FIG. 3) or 3B (FIG. 11), implemented by the computer C in which the at least one processor C1 executes the intermediary program P stored in the at least one memory C2 (FIG. 18), as described in [0038], [0043], [0079], [0096] and [0179]-[0183], programmed with the procedure of FIGS. 4 and 12 described in [0039]-[0040], [0044], [0097]-[0099], [0108]-[0109] and [0113]-[0115], namely, for the accepting function, receiving from one or more requesters a request for a survey containing at least one question to be answered and, in the embodiment of FIG. 12, a condition as to an answer to the question (steps S31 and S31a), and, for the negotiating function, notifying the information processing apparatus 1B of the question, of a condition as to its answer, or of both, and receiving its answer to the notification ([0040], [0099]), as performed in FIG. 12 by sending the question and its condition to the information processing apparatus 1B determined as the request receiver (step S34a), receiving from that apparatus either an answer generated by the vicariously answering language model 4B or an answer informing that the model cannot answer the question (step S35a), and judging whether the answers are complete and, when they are not, determining another request receiver and repeating the exchange (steps S36a and S33a); or, alternatively, that procedure as supplemented in the other disclosed embodiments by answering inquiries from the information processing apparatus 1B about the question and the condition (FIG. 13; [0129]-[0132]), by judging whether to approve an alternative condition proposed by the information processing apparatus 1B ([0150]-[0151]), or by generating and sending an alternative condition and requesting the answer upon its approval (FIG. 15; [0156]-[0161]). This limitation is interpreted to cover the foregoing structure programmed with any one of those disclosed procedures, and equivalents thereof.
Because this/these claim limitations are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, they are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 U.S.C. 112(b)
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 6-9 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Regarding claims 6 and 9, claim 6 recites wherein the language model is a vicariously answering language model having been trained by machine learning so as to be capable of generating, as a surrogate for the predetermined user, an answer to a question inputted, and claim 9 recites negotiations for causing the language model to generate an answer to the predetermined question. The recitation the language model in each claim is indefinite because it cannot be determined which of two different language models recited in claim 5, from which both claims depend, is intended. A claim is indefinite when it contains words or phrases whose meaning is unclear, and where two different elements of the same name are recited earlier in the claims, a later reference to that element with a definite article is unclear where it is uncertain which of the two was intended. See MPEP 2173.05(e); In re Packard, 751 F.3d 1307, 1314 (Fed. Cir. 2014). Claim 5 recites a responding process that generates an answer to the query in one of two alternatives. In the first alternative, the answer is generated via a language model having been trained by machine learning so as to output an answer to a query, the language model having been adjusted to suit the predetermined user with use of job-related data related to a job of the predetermined user. In the second alternative, the answer is generated with use of the job-related data and a language model. The language model of the second alternative is introduced by its own indefinite article, and it is not required to have been adjusted to suit the predetermined user. Claim 5 therefore recites two different language models, and neither claim 6 nor claim 9 identifies which of them is the language model. The uncertainty is material to the scope of both claims. If the language model refers to the adjusted language model of the first alternative, claims 6 and 9 require that adjusted model and do not reach an apparatus that generates its answers only by the second alternative. If it refers to the language model of the second alternative, claims 6 and 9 require that alternative and do not reach an apparatus that generates its answers only via an adjusted model. If it refers to whichever language model the responding process uses, claims 6 and 9 reach both alternatives. The specification does not resolve the uncertainty, because it describes the vicariously answering language model 4B both as a model that has been adjusted to suit the predetermined user and as a model that has not been so adjusted and is used together with the job-related data of the predetermined user (specification [0081], [0084], [0089]), and it describes the survey system 7B with either kind of model (specification [0083], [0084]). Accordingly, claims 6 and 9, read in light of the specification, do not inform a person of ordinary skill in the art about the scope of the claimed subject matter with reasonable certainty. See Nautilus, Inc. v. Biosig Instruments, Inc., 572 U.S. 898, 901 (2014).
For purposes of examination, the language model in claims 6 and 9 is interpreted as the language model used in the responding process of claim 5, whichever of the two recited alternatives is performed, that is, either the adjusted language model via which the answer to the query is generated or the language model used together with the job-related data to generate that answer. This interpretation is consistent with the specification, which describes the vicariously answering language model 4B in both roles (specification [0081], [0084], [0089]).
Claims 7 and 8 depend from claim 6; they incorporate and do not cure this defect, and are rejected for the same reason.
Claim Rejections - 35 U.S.C. 101
The following is a quotation of 35 U.S.C. 101: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 1-10 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The analysis of the claims will follow the 2019 Revised Patent Subject Matter Eligibility Guidance, 84 Fed. Reg. 50 (“2019 PEG”).
Claim 1
Step 1: This claim recites “An information processing apparatus, comprising”; therefore, it is directed to the statutory category of a machine.
Step 2A Prong 1: This claim recites, inter alia:
a designating process of designating data related to a job of a predetermined user as job-related data: These limitations recite a mentally performable process, with the aid of pen and paper, of looking over the data associated with a particular person's work, such as the person's daily reports, the documents the person has prepared and the mail the person has sent and received, and judging which of those items relate to the person's job, which is an observation, evaluation and judgment. See MPEP 2106.04(a)(2)(III).
Thus, this claim recites a judicial exception.
Step 2A Prong 2: The judicial exception is not integrated into a practical application. The additional elements of this claim are as follows:
An information processing apparatus, comprising at least one processor, the at least one processor carrying out: These additional elements are recited at a high level of generality and merely represents generic computer machinery performing in its ordinary capacity to implement the underlying judicial exception. See MPEP 2106.05(f).
and an adjusting process of adjusting, with use of the job-related data, a language model having been trained by machine learning so as to output an answer to a query, such that the language model suits the predetermined user: These additional elements are recited at a high level of generality. The language model is identified only by what it has been trained to do, which is to output an answer to a query, and the adjusting is identified only by its input, the job-related data selected by the abstract idea, and by its intended result, a language model that suits the predetermined user. The claim does not recite how that result is achieved, it names no training procedure, no model parameter that is changed, no model structure, and no manner of storing, locating or referring to the job-related data, so it covers every way of adjusting any language model with the selected data. The specification confirms that breadth, stating that "A method for adjusting the language model with use of the job-related data is not particularly limited provided that the method enables the language model which suits the predetermined user to be generated" ([0018]), and giving as examples retraining the model on text contained in the job-related data, retraining it on that text associated with ground truth data, and registering the job-related data as data to be referred to in generating an answer ([0018]-[0020]). A limitation that covers any solution to an identified problem, with no restriction on how the result is accomplished and no description of the mechanism for accomplishing it, is equivalent to the words "apply it." See MPEP 2106.05(f).
Thus, the way in which the additional elements use or interact with the judicial exception
when analyzed with this claim as a whole do not integrate the judicial exception into a
practical application.
Step 2B: The additional elements from Step 2A Prong 2 include invoking computers or
other machinery to apply the underlying judicial exception and adding the words equivalent to “apply it” with the judicial exception. Thus, the additional elements, viewed individually or in combination, do not provide an inventive concept or otherwise amount to significantly mor than the abstract idea itself. See MPEP 2106.05.
Claim 2
Step 1: a machine, as in claim 1.
Step 2A Prong 1: This claim recite the same abstract ideas as in claim 1 as the judicial exception.
Step 2A Prong 2: The judicial exception is not integrated into a practical application.
The additional elements of this claim are as follows:
wherein in the adjusting process, the at least one processor adjusts the language model such that the language model suits the predetermined user, by registering the job-related data as data to be referred to in generating an answer via the language model: These additional elements do not integrate the judicial exception into a practical application, for the reasons given for claim 1. The registering that claim 2 adds is a new additional element, and it does not integrate the exception either. Registering the job-related data as data to be referred to in generating an answer amounts to storing the selected data so that it is available when the language model generates an answer, and the claim does not recite how the registered data is located, retrieved or combined with a query, or how the language model uses it. The specification likewise describes the registration by its result and by an example in which the job-related data related to a query "can be detected by, for example, searching the job-related data with a character string extracted from the query" ([0020]). Because the registering is the manner of adjusting that claim 2 adds, it is evaluated for what it contributes to the claim rather than as incidental activity. It recites only the idea of a solution, making the selected data available to a generic language model, without any detail of how the data is made available or how the model uses it, and it invokes the computer and the language model merely as tools in their ordinary capacities of storing data and generating answers. It is therefore a mere instruction to apply the exception using generic computer and machine learning technology, see MPEP 2106.05(f), and it generally links the exception to the technological environment of machine learning language models, see MPEP 2106.05(h).
Thus, the way in which the additional elements use or interact with the judicial exception
when analyzed with this claim as a whole do not integrate the judicial exception into a
practical application.
Step 2B: The additional elements from Step 2A Prong 2 include adding the words equivalent to “apply it” with the judicial exception and generically linking the exception into a field of use/technological environment. Thus, the additional elements, viewed individually or in combination, do not provide an inventive concept or otherwise amount to significantly mor than the abstract idea itself. See MPEP 2106.05.
Claim 3
Step 1: a machine, as in claim 1.
Step 2A Prong 1: This claim recites, inter alia:
wherein in the designating process, uses hierarchy information indicating a hierarchy of an organization to which the predetermined user belongs, to designate a level having a predetermined relationship with a level to which the predetermined user belongs, and designates the job-related data related to the level designated: These limitations further recite the mentally performable process, with the aid of pen and paper, of consulting an organization chart of the organization to which a person belongs, identifying a level having a chosen relationship to the person's own level, such as the same level, the level directly above it or the level directly below it, and selecting the job data related to the level identified, which is an observation, evaluation and judgment. See MPEP 2106.04(a)(2)(III). The specification's own example is of this kind: for a user in department X2, the data of department X4 directly below, of department X1 directly above, or of department X3 at the same level may be designated ([0059]).
Thus, this claim further recites the judicial exception.
Step 2A Prong 2: The judicial exception is not integrated into a practical application. The additional elements of this claim are as follows:
the at least one processor uses: These additional elements are recited at a high level of generality and merely represents generic computer machinery performing in its ordinary capacity to implement the underlying judicial exception. See MPEP 2106.05(f).
Thus, the way in which the additional elements use or interact with the judicial exception
when analyzed with this claim as a whole do not integrate the judicial exception into a
practical application.
Step 2B: The additional elements from Step 2A Prong 2 include invoking computers or other machinery to apply the underlying judicial exception. Thus, the additional elements, viewed individually or in combination, do not provide an inventive concept or otherwise amount to significantly mor than the abstract idea itself. See MPEP 2106.05.
Claim 4
Step 1: a machine, as in claim 1.
Step 2A Prong 1: This claim recites, inter alia:
wherein in the designating process, designates the job-related data according to a set degree of confidentiality of each piece of data related to [[a]]the job (interpreted per the Claim Objections set forth above) of the predetermined user: These limitations further recite the mentally performable process, with the aid of pen and paper, of checking the confidentiality marking assigned in advance to each item of a person's job data and selecting the items according to that marking, for example leaving out an item marked as data to be kept confidential and selecting an item marked as data that need not be kept confidential, which is an observation, evaluation and judgment. See MPEP 2106.04(a)(2)(III). The specification's example is a degree of confidentiality set to 1 for data that should be kept confidential and to 0 for data that does not need to be ([0061]).
Thus, this claim further recites the judicial exception.
Step 2A Prong 2: The judicial exception is not integrated into a practical application. The additional elements of this claim are as follows:
the at least one processor designates: These additional elements are recited at a high level of generality and merely represents generic computer machinery performing in its ordinary capacity to implement the underlying judicial exception. See MPEP 2106.05(f).
Thus, the way in which the additional elements use or interact with the judicial exception
when analyzed with this claim as a whole do not integrate the judicial exception into a
practical application.
Step 2B: The additional elements from Step 2A Prong 2 include invoking computers or other machinery to apply the underlying judicial exception. Thus, the additional elements, viewed individually or in combination, do not provide an inventive concept or otherwise amount to significantly mor than the abstract idea itself. See MPEP 2106.05.
Claim 5
Step 1: This claim recites “An information processing apparatus, comprising”; therefore, it is directed to the statutory category of a machine.
Step 2A Prong 1: This claim recites, inter alia:
a responding process of generating an answer to the query accepted in the accepting process, or generating an answer to the query accepted in the accepting process, with use of the job-related data: These limitations recite the mentally performable process, with the aid of pen and paper, of answering a question that a particular person has asked, including by consulting documents about that person's work while composing the answer, which is an evaluation, judgment and opinion. See MPEP 2106.04(a)(2)(III). The specification describes the claimed concept as one that people perform, with the language model used as a tool to perform it: in the survey embodiment the model serves as a surrogate or private secretary of the user ( [0080]); in the medical field it generates an answer similar to one provided by medical personnel, for use as a second opinion ([0030]); the user may use the generated answer as a guide in deciding how to answer a question put to the user ([0022]); and in the second alternative the job-related data related to the query is found by searching for it and is added to the query (paragraph [0027]), as a person would look up related documents before answering. That the at least one processor and a language model carry out the responding process does not take it outside the mental process grouping, because a mental process performed using a computer as a tool, or performed in a computer environment, is still a mental process. See MPEP 2106.04(a)(2)(III)(C).
Thus, this claim recites a judicial exception.
Step 2A Prong 2: The judicial exception is not integrated into a practical application. The additional elements of this claim are as follows:
An information processing apparatus, comprising at least one processor, the at least one processor carrying out: These additional elements are recited at a high level of generality and merely represents generic computer machinery performing in its ordinary capacity to implement the underlying judicial exception. See MPEP 2106.05(f).
accepting an input of a query from a predetermined user: These additional elements are recited at a high level of generality, without any particular input device, interface or format, and the specification describes the query only as accepted through a screen displayed on the user's terminal or through an input section ([0074]). Receiving the information to be processed is mere data gathering, which is insignificant extra-solution activity. See MPEP 2106.05(g).
via a language model having been trained by machine learning so as to output an answer to a query, the language model having been adjusted to suit the predetermined user with use of job-related data related to a job of the predetermined user, or with use of the job-related data and a language model: These additional limitations for a language model are identified only by what they have been trained to do and, for the first alternative, by the data with which the model was earlier adjusted. Claim 5 does not require the at least one processor to perform that adjustment or recite how it was done, and it does not recite how an answer is generated via the adjusted model or how the job-related data and the model are used together to generate one. The specification confirms the generality, in the second alternative the answer may be generated by searching for job-related data related to the query, adding that data to the query or rewriting the query on the basis of it, and inputting the result to a language model ([0027]). These limitations use a generic machine learning model as a tool to produce the answer that the abstract idea calls for, which is equivalent to the words "apply it", see MPEP 2106.05(f), and they generally link the abstract idea to the technological environment of machine learning language models, see MPEP 2106.05(h).
Thus, the way in which the additional elements use or interact with the judicial exception
when analyzed with this claim as a whole do not integrate the judicial exception into a
practical application.
Step 2B: The additional elements from Step 2A Prong 2 include invoking computers or other machinery to apply the underlying judicial exception and insignificant extra-solution activity of mere data gathering recited by “accepting an input of a query from a predetermined user”. Data gathering where the query is received from the user's terminal, the accepting is also receiving data over a network, a computer function that the courts have recognized as well understood, routine and conventional when claimed at a high level of generality, see MPEP 2106.05(d)(II). The additional elements further include adding the words equivalent to “apply it” with the judicial exception and generically linking the exception into a field of use/technological environment. Thus, the additional elements, viewed individually or in combination, do not provide an inventive concept or otherwise amount to significantly mor than the abstract idea itself. See MPEP 2106.05.
Claim 6
Step 1: a machine, as in claim 5.
Step 2A Prong 1: This claim recites, inter alia:
capable of generating, as a surrogate for the predetermined user, an answer to a question inputted, in the accepting process, accepts an input of a question asked the predetermined user, and further carries out an answering allowance judging process of judging whether to cause to generate an answer to the question: These limitations recites a question asked the predetermined user by another party and a model that answers as a surrogate for the predetermined user, and the judging decides whether a question put to one person will be answered on that person's behalf, which is among the certain methods of organizing human activity managing interactions between people. See MPEP 2106.04(a)(2)(II)(C). That the surrogate is a language model, which is evaluated below as an additional element, does not take the decision outside this grouping, because the grouping includes activity between a person and a computer. See MPEP 2106.04(a)(2)(II). The claim places no limit on the criterion used for the judgment, and the specification's criteria include whether the user satisfies the conditions received with the question, whether the offered reward reaches a lower limit set by the user, and the user's own selection ([0121]). The judging further recites the mentally performable process, with the aid of pen and paper, of deciding whether a stand-in should answer, on a person's behalf, a question that someone has put to that person, which is an evaluation and judgment. See MPEP 2106.04(a)(2)(III)
Thus, this claim further recites the judicial exception.
Step 2A Prong 2: The judicial exception is not integrated into a practical application. The additional elements of this claim are as follows:
wherein the language model is a vicariously answering language model having been trained by machine learning so as to be capable of generating…the vicariously answering language model to generate: These additional elements regarding a vicariously answering language model is identified only by what it has been trained to do, which is to generate answers to questions as a surrogate for the predetermined user, and the claim does not recite any model structure or training technique that accomplishes that result. The specification confirms that a language model adjusted with the user's job-related data, or a language model with which that data is registered for reference, may serve as the vicariously answering language model ([0081]). Thus, the limitations use a generic machine learning model as a tool, which is equivalent to the words "apply it", see MPEP 2106.05(f).
the at least one processor accepts an input…the at least one processor further carries out: These additional elements recite generic computer components performing in their ordinary capacity to implement the underlying judicial exception. See MPEP 2106.05(f).
Thus, the way in which the additional elements use or interact with the judicial exception
when analyzed with this claim as a whole do not integrate the judicial exception into a
practical application.
Step 2B: The additional elements from Step 2A Prong 2 include adding the words equivalent to “apply it” with the judicial exception and invoking computers or other machinery to apply the underlying judicial exception. Thus, the additional elements, viewed individually or in combination, do not provide an inventive concept or otherwise amount to significantly mor than the abstract idea itself. See MPEP 2106.05.
Claim 7
Step 1: a machine, as in claim 6.
Step 2A Prong 1: This claim recites, inter alia:
wherein in the answering allowance judging process, in a case where the predetermined user satisfies a condition associated with the question, judges that an answer to the question should be generated corresponding to the predetermined user: These limitation further recites a mentally performable process, with the aid of pen and paper, of checking whether the person to whom a question is addressed meets a condition attached to the question, such as a required number of years of practical experience in a department or a reward the person desires ( [0092] and [0094]), and deciding on that basis that the person's surrogate should answer, which is an evaluation and judgment. See MPEP 2106.04(a)(2)(III). It also recites following a rule for managing an interaction between people, namely that a question asked of the person is answered on the person's behalf when the person meets the condition associated with the question; as explained for claim 6, the use of a language model as the surrogate does not take that rule outside the grouping, which includes activity between a person and a computer. See MPEP 2106.04(a)(2)(II)(C).
Thus, this claim further recites the judicial exception.
Step 2A Prong 2: The judicial exception is not integrated into a practical application. The additional elements of this claim are as follows:
the at least one processor…by the vicariously answering language model: These additional elements are recited at a high level of generality and merely represents generic computer machinery performing in its ordinary capacity to implement the underlying judicial exception. See MPEP 2106.05(f).
Thus, the way in which the additional elements use or interact with the judicial exception
when analyzed with this claim as a whole do not integrate the judicial exception into a
practical application.
Step 2B: The additional elements from Step 2A Prong 2 include invoking computers or
other machinery to apply the underlying judicial exception. Thus, the additional elements, viewed individually or in combination, do not provide an inventive concept or otherwise amount to significantly mor than the abstract idea itself. See MPEP 2106.05.
Claim 8
Step 1: a machine, as in claim 6.
Step 2A Prong 1:
an alternative condition generating process of generating an alternative condition which replaces a condition associated with the question; and a negotiating process of notifying a sender of the question of the alternative condition and inquiring of the sender whether to approve of the alternative condition: These limitations further recite certain methods of organizing human activity of a commercial or legal interaction wherein replacing a condition on which the question was to be answered with an alternative condition, proposing that alternative condition to the party that asked the question, and asking that party whether it approves, which is a counter-proposal and a request for acceptance in forming an agreement between the party asking the question and the party answering it. The claim recites that exchange of proposal and approval without limiting the kind of condition, and the conditions the specification describes are terms of answering, such as a reward for providing an answer, an answering method, whether an additional question is permitted and an attribute of the respondent ( [0088]), one example being a requirement of a 10% increase in the offered reward in place of having the answer checked by the user ( [0148]). Proposing and approving the terms on which one party will answer for another is agreement formation and a business relation, within the commercial or legal interactions sub-grouping of certain methods of organizing human activity, and deciding on what terms a question asked of the user will be answered is also managing an interaction between people. See MPEP 2106.04(a)(2)(II)(B), (C). Formulating the replacement condition is also a mentally performable evaluation and judgment, with the aid of pen and paper. See MPEP 2106.04(a)(2)(III).
Thus, this claim further recites the judicial exception.
Step 2A Prong 2: The judicial exception is not integrated into a practical application. The additional elements of this claim are as follows:
wherein the at least one processor further carries out: These additional elements are recited at a high level of generality and merely represents generic computer machinery performing in its ordinary capacity to implement the underlying judicial exception. See MPEP 2106.05(f).
Thus, the way in which the additional elements use or interact with the judicial exception
when analyzed with this claim as a whole do not integrate the judicial exception into a
practical application.
Step 2B: The additional elements from Step 2A Prong 2 include invoking computers or
other machinery to apply the underlying judicial exception. Thus, the additional elements, viewed individually or in combination, do not provide an inventive concept or otherwise amount to significantly mor than the abstract idea itself. See MPEP 2106.05.
Claim 9
Step 1: a machine, as in claim 5.
Step 2A Prong 1: This claim recites the same abstract ideas as in claim 5 and further recites:
accepting a request for a survey in which answers to a predetermined question are collected, and conducting, negotiations for causing the language model to generate an answer to the predetermined question: These limitations recite acting as an intermediary between a party requesting a survey and a party asked to answer it, taking in a request for a survey in which answers to a question are to be collected, and conducting negotiations with the answering party toward the provision of an answer to the question. The claim recites the negotiations without any technical limitation. Under the claim interpretation under 35 U.S.C. 112(f) set out in this action, the negotiating includes notifying the answering party of the question, of a condition as to its answer, or of both, and receiving its reply, which may be an answer or a notice that it cannot answer ([0040], [0099] and [0113]), and, in the disclosed alternatives, exchanging inquiries about the question and the condition and proposing and approving alternative conditions ([0129]-[0132], [0150]-[0151] and [0156]-[0161]). Soliciting answers to a survey on a requester's behalf and negotiating with a respondent over whether, and on what conditions, the respondent will answer are commercial interactions in the form of business relations and agreements, and they manage interactions between people, namely the requester and the respondent, all of which fall within the certain methods of organizing human activity grouping. The specification's examples of a survey on marketing ([0039]) and of conditions such as a reward for providing an answer ([0100]) illustrate that activity but are not relied on to define it, and the respondent's being represented by an apparatus does not remove the activity from the grouping, which includes activity between a person and a computer. See MPEP 2106.04(a)(2)(II)(B), (C).
Thus, this claim further recites the judicial exception.
Step 2A Prong 2: The judicial exception is not integrated into a practical application. The additional elements of this claim are as follows:
A survey system, comprising: the information processing apparatus according to claim 5; and an intermediary apparatus for accepting…conducting with the information processing apparatus: These additional elements are recited at a high level of generality and merely represents generic computer machinery performing in its ordinary capacity to implement the underlying judicial exception. See MPEP 2106.05(f).
Thus, the way in which the additional elements use or interact with the judicial exception
when analyzed with this claim as a whole do not integrate the judicial exception into a
practical application.
Step 2B: The additional elements from Step 2A Prong 2 include invoking computers or
other machinery to apply the underlying judicial exception. Thus, the additional elements, viewed individually or in combination, do not provide an inventive concept or otherwise amount to significantly mor than the abstract idea itself. See MPEP 2106.05.
Claim 10
Step 1: This claim recites “An adjustment method, comprising”; therefore, it is directed to the statutory category of a process.
Step 2A Prong 1: This claim recites, inter alia:
designating data related to a job of a predetermined user as job-related data: These limitations recite a mentally performable process, with the aid of pen and paper, of looking over the data associated with a particular person's work, such as the person's daily reports, the documents the person has prepared and the mail the person has sent and received, and judging which of those items relate to the person's job, which is an observation, evaluation and judgment. See MPEP 2106.04(a)(2)(III).
Thus, this claim recites a judicial exception.
Step 2A Prong 2: The judicial exception is not integrated into a practical application. The additional elements of this claim are as follows:
at least one processor designating: These additional elements are recited at a high level of generality and merely represents generic computer machinery performing in its ordinary capacity to implement the underlying judicial exception. See MPEP 2106.05(f).
the at least one processor adjusting, with use of the job-related data, a language model having been trained by machine learning so as to output an answer to a query, such that the language model suits the predetermined user: These additional elements are recited at a high level of generality. The language model is identified only by what it has been trained to do, which is to output an answer to a query, and the adjusting by the at least one processor is identified only by its input, the job-related data selected by the abstract idea, and by its intended result, a language model that suits the predetermined user. The claim does not recite how that result is achieved, it names no training procedure, no model parameter that is changed, no model structure, and no manner of storing, locating or referring to the job-related data, so it covers every way of adjusting any language model with the selected data. The specification confirms that breadth, stating that "A method for adjusting the language model with use of the job-related data is not particularly limited provided that the method enables the language model which suits the predetermined user to be generated" ([0018]), and giving as examples retraining the model on text contained in the job-related data, retraining it on that text associated with ground truth data, and registering the job-related data as data to be referred to in generating an answer ([0018]-[0020]). A limitation that covers any solution to an identified problem, with no restriction on how the result is accomplished and no description of the mechanism for accomplishing it, is equivalent to the words "apply it." See MPEP 2106.05(f).
Thus, the way in which the additional elements use or interact with the judicial exception
when analyzed with this claim as a whole do not integrate the judicial exception into a
practical application.
Step 2B: The additional elements from Step 2A Prong 2 include invoking computers or
other machinery to apply the underlying judicial exception and adding the words equivalent to “apply it” with the judicial exception. Thus, the additional elements, viewed individually or in combination, do not provide an inventive concept or otherwise amount to significantly mor than the abstract idea itself. See MPEP 2106.05.
Claim Rejections - 35 U.S.C. 102
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 the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1, 2, 5, and 10 are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Belkin et al. (hereinafter Belkin), US 2025/0021919 A1.
Regarding independent claim 1, Belkin discloses an information processing apparatus, comprising (Belkin: [0043], "a system as disclosed herein includes one or more of the following"; [0044], "An online server that could be deployed either in the public or private (where so requested by the customer) cloud that performs one or more of the following functions"; the system (an information processing apparatus) includes the online server, which runs the inference layer, and the data extraction layer, which organizes each user's data for that inference layer, so the two operate together as parts of the one system) at least one processor, the at least one processor carrying out (Belkin: [0051], "Inference layer responsible for running machine learning models trained with the appropriate dataset to answer questions"; while not explicitly stated, the online server that runs these models and the periodically run process of the data extraction layer necessarily execute on a processor (at least one processor), because software processes and models are executed by a processor): a designating process of designating data related to a job of a predetermined user (Belkin: [0059], "Data extraction layer: a periodically run process that extracts and organizes relevant data from other systems for each user"; the data extraction layer (a designating process) extracts and organizes the relevant data of each user (a predetermined user); [0061], "the user is prompted to connect their datasets (such as email, google drive, slack and other similar systems) into the system"; [0049], "Supports functions necessary for users to manage data connections to systems users have access to"; the relevant data comes from the work systems the user connects, such as email, shared drives and collaboration tools, so it is data of the user's work (data related to a job of a predetermined user)) as job-related data (Belkin: [0059], "to be used for context creation by the inference layer"; the relevant data so extracted and organized for the user is identified as the data (job-related data) to be used for that user's context creation); and an adjusting process of adjusting, with use of the job-related data, (Belkin: [0059], "Data extraction layer processes domain specific data (e.g., extracts data records from a financial system, ticketing system, source code etc. and converts it into a representation usable by the model)"; the relevant data (the job-related data) is converted into a representation usable by the model; [0041], "houses vector representation of user data 226 along with metadata necessary to retrieve context for the model and for the user"; [0050], "Data management layer segregated for each user that manages datasets available to the inference layer"; the user's data is held as vector representations, with the metadata needed to retrieve it as context for the model, in a data management layer segregated for that user, so that the context creation (an adjusting process) changes the data available to the model when it answers for that user) a language model (Belkin: [0040], "Inference layer 222 houses the main model (LLM)"; FIG. 2 shows Vector Database 224, holding the user data slices 226, feeding Inference Layer 222, which houses the LLM (a language model)) having been trained by machine learning (Belkin: [0080], "a system as disclosed herein uses several base large language models (LLMs) trained on large public data sets (i.e., a model like GPT-3)"; the LLM is a base large language model trained by machine learning on large public data sets) so as to output an answer to a query (Belkin: [0051], "Inference layer responsible for running machine learning models trained with the appropriate dataset to answer questions"; [0019], "Digital twins handle routine queries and tasks"; the trained models are run to produce answers (an answer) to the routine queries (a query) that the digital twins handle), such that the language model suits the predetermined user (Belkin: [0113], "An AI (artificial intelligence) based system that represents and mimics interactions from a point of view of a specific individual (an employee of the company) using data (currently, previously, etc.) accessible to that employee"; [0114], "A system that strictly segregates individual data, such that responses given and actions taken are always from a context and access capabilities of an individual employee represented by the AI digital twin"; the LLM answers from the data of the one user (the predetermined user) that the digital twin represents, rather than generically).
Regarding dependent claim 2, Belkin discloses the information processing apparatus according to claim 1, wherein in the adjusting process, the at least one processor adjusts the language model (Belkin: [0059], "extracts and organizes relevant data from other systems for each user to be used for context creation by the inference layer", [0080], "a system as disclosed herein uses several base large language models (LLMs) trained on large public data sets (i.e., a model like GPT-3)"; the context creation (the adjusting process) for each user is carried out by the processor (the at least one processor) that, as explained above for claim 1, the system necessarily includes, and changes the data the LLM (the language model) draws on when it answers) such that the language model suits the predetermined user, (Belkin: [0114], "responses given and actions taken are always from a context and access capabilities of an individual employee represented by the AI digital twin"; the LLM answers from the data of the user (the predetermined user) that the digital twin represents) by registering the job-related data (Belkin: [0059], "converts it into a representation usable by the model"; [0041], "houses vector representation of user data 226"; the user's relevant data (the job-related data) is converted into vector representations and stored in the vector database (registering)) as data to be referred to in generating an answer via the language model (Belkin: [0041], "along with metadata necessary to retrieve context for the model and for the user"; [0050], "Data management layer segregated for each user that manages datasets available to the inference layer", [0114], "responses given and actions taken are always from a context and access capabilities of an individual employee represented by the AI digital twin"; the vector representations are stored with the metadata needed to retrieve them as context (data to be referred to) for the LLM, and FIG. 2 shows Vector Database 224 feeding Inference Layer 222, which houses the LLM, so the registered data is available to the inference layer when the LLM generates an answer (generating an answer via the language model)).
Regarding independent claim 5, Belkin discloses an information processing apparatus, comprising (Belkin: [0043], "a system as disclosed herein includes one or more of the following"; [0044], "An online server that could be deployed either in the public or private (where so requested by the customer) cloud that performs one or more of the following functions"; the system (an information processing apparatus) includes the online server, which runs the inference layer, and the data extraction layer, which organizes each user's data for that inference layer, so the two operate together as parts of the one system) at least one processor, the at least one processor carrying out: (Belkin: [0051], "Inference layer responsible for running machine learning models trained with the appropriate dataset to answer questions"; while not explicitly stated, the online server that runs these models and the periodically run process of the data extraction layer necessarily execute on a processor (at least one processor), because software processes and models are executed by a processor) an accepting process of accepting an input of a query (Belkin: [0052], "User interaction layer that records user dialogs with digital twins"; [0037], "a business logic component of the system that governs user interaction flows (i.e. messages, conversations, responses, etc.)"; the user interaction layer (an accepting process) receives each message a user enters in a dialog with a digital twin; [0019], "Digital twins handle routine queries and tasks"; the message is one of the queries (a query) the digital twin handles) from a predetermined user (Belkin: [0070], "the user will be able to ask his/her own digital twin something the user believes he knew or likely knew but forgot"; the query is entered by the owner of the digital twin (a predetermined user)); and a responding process of generating an answer to the query accepted in the accepting process, via a language model having been trained by machine learning so as to output an answer to a query, the language model having been adjusted to suit the predetermined user with use of job-related data related to a job of the predetermined user, or generating an answer to the query accepted in the accepting process, (Belkin: [0051], "Inference layer responsible for running machine learning models trained with the appropriate dataset to answer questions"; [0114], "responses given and actions taken are always from a context and access capabilities of an individual employee represented by the AI digital twin"; the inference layer (a responding process) generates the answers (an answer) of the owner's digital twin to what the owner asks through the user interaction layer; the limitation recites two alternatives joined by "or", only one of which is required, and this rejection relies on the second alternative, generating the answer with use of the job-related data and a language model) with use of the job-related data (Belkin: [0059], "extracts and organizes relevant data from other systems for each user to be used for context creation by the inference layer"; [0041], "houses vector representation of user data 226 along with metadata necessary to retrieve context for the model and for the user"; the answer is generated with the context retrieved from the relevant data (the job-related data) extracted from the owner's work systems) and a language model (Belkin: [0040], "Inference layer 222 houses the main model (LLM)"; the answer is generated by the LLM (a language model) housed in the inference layer).
Regarding independent claim 10, Belkin discloses an adjustment method, comprising (Belkin: [0013], "Techniques are disclosed to create, maintain, and use Artificial Intelligence (AI)-based digital twins of company employees"; the techniques (an adjustment method) create and maintain a digital twin for each employee): at least one processor (Belkin: [0051], "Inference layer responsible for running machine learning models trained with the appropriate dataset to answer questions"; while not explicitly stated, the online server that runs these models and the periodically run process of the data extraction layer necessarily execute on a processor (at least one processor), because software processes and models are executed by a processor) designating data related to a job of a predetermined user (Belkin: [0059], "Data extraction layer: a periodically run process that extracts and organizes relevant data from other systems for each user"; [0061], "the user is prompted to connect their datasets (such as email, google drive, slack and other similar systems) into the system"; the process extracts and organizes, for each user (a predetermined user), the relevant data from the work systems that user connects (data related to a job of a predetermined user)) as job-related data (Belkin: [0059], "to be used for context creation by the inference layer"; the relevant data so organized is identified as the data (job-related data) used for that user's context creation); and the at least one processor adjusting, with use of the job-related data, (Belkin: [0059], "converts it into a representation usable by the model"; [0041], "houses vector representation of user data 226 along with metadata necessary to retrieve context for the model and for the user"; the processor (the at least one processor) converts the relevant data (the job-related data) into vector representations held with the metadata needed to retrieve them as context for the model, which changes the data the model draws on when it answers for that user (adjusting)) a language model (Belkin: [0040], "Inference layer 222 houses the main model (LLM)"; the context is created for the LLM (a language model)) having been trained by machine learning (Belkin: [0080], "a system as disclosed herein uses several base large language models (LLMs) trained on large public data sets"; the LLM is trained by machine learning) so as to output an answer to a query (Belkin: [0051], "machine learning models trained with the appropriate dataset to answer questions"; [0019], "Digital twins handle routine queries and tasks"; the models produce answers (an answer) to the routine queries (a query) that the digital twins handle), such that the language model suits the predetermined user (Belkin: [0114], "responses given and actions taken are always from a context and access capabilities of an individual employee represented by the AI digital twin"; the LLM answers from the data of the user (the predetermined user) that the digital twin represents).
Claim Rejections - 35 U.S.C. 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 3 is rejected under 35 U.S.C. 103 as being unpatentable over Belkin, as applied in the rejection of claim 1, in view of Karimzadehgan et al. (hereinafter Karimzadehgan) "Enhancing Expert Finding Using Organizational Hierarchies" (2009).
Regarding dependent claim 3, Belkin teaches the information processing apparatus according to claim 1, as set forth above for claim 1.
Belkin does not expressly teach wherein in the designating process, the at least one processor uses hierarchy information indicating a hierarchy of an organization to which the predetermined user belongs, to designate a level having a predetermined relationship with a level to which the predetermined user belongs, and designates the job-related data related to the level designated.
However, Karimzadehgan teaches wherein in the designating process, the at least one processor uses hierarchy information (Karimzadehgan: page 182, Section 4.2, "Since it is unlikely that we will have this information for all members, we propose using the organizational hierarchy as an additional data source"; the hierarchy-based algorithm (the designating process) uses the organizational hierarchy (hierarchy information) as a data source; page 180, Section 3, "By crawling these lists, we were able to create expert profiles for 24% of all people in the organization"; while not explicitly stated, the expert finding system that crawls the discussion lists, builds the profiles and runs the algorithm necessarily runs on a processor (the at least one processor)) indicating a hierarchy of an organization to which the predetermined user belongs, (Karimzadehgan: page 178, "Figure 1 shows an example organizational hierarchy. The nodes represent employees and the links between them represent managerial reporting relationships"; the organizational hierarchy (a hierarchy of an organization) records the reporting relationships among the employees of the organization, including the employee (the predetermined user) whose expertise is estimated) to designate a level having a predetermined relationship with a level to which the predetermined user belongs, (Karimzadehgan: page 180, Section 3, "propagating expertise scores among neighbors (e.g., managers, subordinates, and peers)"; page 178, "Two members are considered peers if they share the same direct manager"; for an employee, the algorithm identifies the neighbors (a level having a predetermined relationship with a level to which the predetermined user belongs) joined to the employee by a reporting or peer link, namely the manager at the level immediately above the employee's level, the subordinates at the level immediately below it, and the peers at the employee's own level) and designates the job-related data related to the level designated (Karimzadehgan: page 183, Section 5.2, "We attempted to build a profile for each person in the organization by considering the emails they sent to the list in reply to posted questions"; the expert profiles (the job-related data) are built from the work email that each member sent to internal discussion lists; page 187, Section 7, "we also investigated propagating the keywords in the expert profiles to the neighbors and scoring employees based on these expanded profiles"; the keywords in the neighbors' expert profiles (the job-related data) are added to the employee's expanded profile; page 187, "neighbors in an organization tend to have similar expertise"; the neighbors' profiles relate to the employee's own work).
Because Karimzadehgan is reasonably pertinent to the problem faced by the inventor of selecting, from the members of an organization, the work data indicative of what a particular employee knows, a selection Belkin makes in extracting the data used for each user's context creation, accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to apply Karimzadehgan's use of the organizational hierarchy to identify an employee's direct manager, direct subordinates and peers and to use their email-based expert profiles for that employee to the data extraction layer of Belkin's system, with a reasonable expectation of success, by having the data extraction layer, when it extracts the relevant data used for context creation for a user's digital twin, also extract the relevant data of that user's direct manager, direct subordinates and peers identified from the organizational hierarchy, limited to data such as discussion-list email that is visible to all employees, so that the context stays within the data accessible to that user (Belkin: [0059], [0114]), in the same way that Belkin already adds an answer given by one digital twin to the knowledge/data set of another digital twin (Belkin: [0077]) and already sends a question to an employee with the same job title, description or credentials (Belkin: [0079]); Karimzadehgan's report that propagating profile keywords ranked experts less accurately than propagating scores compares two ways of using the neighbors' information, does not criticize or discourage using the neighbors' profiles, and is followed by Karimzadehgan's plan to propagate more of the neighbors' information (Karimzadehgan: page 187), to teach wherein in the designating process, the at least one processor uses hierarchy information indicating a hierarchy of an organization to which the predetermined user belongs, to designate a level having a predetermined relationship with a level to which the predetermined user belongs, and designates the job-related data related to the level designated. This modification would have been motivated by the desire to refine the estimate of what an employee knows, particularly for an employee about whom little or no information is available (Karimzadehgan: page 181).
Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Belkin, as applied in the rejection of claim 1 above, in view of Al Bawab et al. (hereinafter AlBawab), US 2020/0349920 A1.
Regarding dependent claim 4, Belkin teaches the information processing apparatus according to claim 1, as set forth above for claim 1.
Belkin does not expressly teach wherein in the designating process, the at least one processor designates the job-related data according to a set degree of confidentiality of each piece of data related to [[a]]the job (interpreted per the Claim Objections set forth above) of the predetermined user.
However, AlBawab teaches wherein in the designating process, the at least one processor designates the job-related data (AlBawab: [0054], "For data labeled as private, the processor 620 may filter out this data and only store an ID of the data without storing the private data itself. Meanwhile, the processor 620 may store both the ID and the content of data determined as public"; in the filtering process (the designating process), the processor 620 (the at least one processor) keeps the content of the public data (the job-related data) for use in training and excludes the private data) according to a set degree of confidentiality (AlBawab: [0054], "the processor 620 may determine whether the data is public data or private data based on a label of the source of the data"; [0016], "the users may create a channel or other group identifier that is labeled as private"; the public or private label (a set degree of confidentiality) is set for the channel from which the data comes, and the processor keeps or excludes the data according to that label) of each piece of data related to the job of the predetermined user (AlBawab: [0049], "the aggregating may include tagging a piece of data within the organizational data with a respective channel identification"; [0026], "The unique IDs may correspond to respective data channels (e.g., meeting groups, projects, email chains, etc.) where the data has been collected from"; [0031], "the system may check on the privacy of each data"; each piece of data (each piece of data) collected from the user's work channels, such as meeting groups, projects and email chains (related to the job of the predetermined user), carries the label of its own channel, and the privacy of each piece is checked, so each piece is kept or excluded according to its own label).
Because Belkin and AlBawab are analogous art with both addressing which of an organization's work data, such as email, documents and meeting content, may be used with a language model without exposing private information, accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to apply AlBawab's filtering of collected work data by the public or private label of each piece of data to the data extraction layer of Belkin's system, with a reasonable expectation of success, by having the data extraction layer check the label carried by each piece of a user's data and include in the relevant data used for context creation for that user's digital twin only the pieces labeled public, which moves into the extraction itself the concern with private or sensitive data that Belkin's company administrators address by searching the data used to train the digital twins (Belkin: [0074]) and keeps private data in one user's digital twin from reaching the other users who ask that digital twin questions (Belkin: [0066]), accepting, as AlBawab does for a model used across an organization, that the pieces labeled private are left out of the data the digital twin draws on, to teach wherein in the designating process, the at least one processor designates the job-related data according to a set degree of confidentiality of each piece of data related to the job of the predetermined user. This modification would have been motivated by the desire to prevent leakage of private information when business data is used with a language model (AlBawab: [0022]).
Claim 6 is rejected under 35 U.S.C. 103 as being unpatentable over Belkin, as applied in the rejection of claim 5 above, in view of Rey et al. (hereinafter Rey), US 2022/0166731 A1.
Regarding dependent claim 6, Belkin teaches the information processing apparatus according to claim 5, wherein the language model used in the responding process (interpreted per the 35 U.S.C. 112(b) rejection set forth above) is a vicariously answering language model (Belkin: [0070], "The user will be able to interact with his/her own digital twin"; the owner's digital twin's LLM (the language model used in the responding process) answers the owner's queries with context created from the owner's relevant data, as relied upon above for the second alternative of the responding process of claim 5; [0069], "the user is able to see interactions other users are having with his/her digital twin"; [0020], "The digital twins are available around the clock, providing answers and assistance even when the human counterparts are not available"; the same digital twin also answers other users in the owner's place, so the twin's LLM (a vicariously answering language model) answers vicariously) having been trained by machine learning (Belkin: [0080], "a system as disclosed herein uses several base large language models (LLMs) trained on large public data sets"; [0015], "a given employee's emails, texts, files, and other data may be used to train a model that can be used to provide an AI-based digital twin"; the digital twin's LLM is a model trained by machine learning) so as to be capable of generating, as a surrogate for the predetermined user, an answer to a question inputted, (Belkin: [0015], "capable of generating and providing responses that reflect the knowledge and/or expertise of the given employee"; [0014], "replicate the knowledge, skills, and work behavior of their human counterparts, allowing for anytime access to the employee's expertise, even after they leave the company"; the digital twin generates answers (an answer) to the question (a question) put to it that reflect the owner's knowledge, in the owner's place (as a surrogate for the predetermined user)) in the accepting process, the at least one processor accepts an input of a question (Belkin: [0066], "a user can discover available digital twins (including for former users that are no longer with the company) already present in the system (created by other users) and interact with them"; [0066], "by asking the same question of all of them"; [0052], "User interaction layer that records user dialogs with digital twins"; the user interaction layer (the accepting process), run by the processor (the at least one processor), receives the question (a question) that another user puts to the owner's digital twin) asked the predetermined user, (Belkin: [0068], "The user may choose to escalate the question-in which case the owner of the digital twin will be notified to take an action and to answer the question that received an unsatisfactory answer from the system"; the question is one asked of the owner of the digital twin (the predetermined user), because escalation sends the same question to the owner to answer) and
Belkin does not expressly teach the at least one processor further carries out an answering allowance judging process of judging whether to cause the vicariously answering language model to generate an answer to the question.
However, Rey teaches the at least one processor further carries out (Rey: [0005], "a data processing system having a processor and a memory in communication with the processor wherein the memory stores executable instructions that, when executed by the processor, cause the data processing system to perform multiple functions"; the processor (the at least one processor) executes the functions described below) an answering allowance judging process of judging whether to cause (Rey: [0074], "method 500 may proceed to determine if access to the requested information and/or document is available to the sender, at 525"; [0047], "the access determination service 116 may receive a request for accessing information as an input and may provide a determination as to whether access can be granted to the information as an output"; the access determination service 116 (an answering allowance judging process) determines, before any response is created, whether the sender may access the requested information) the vicariously answering language model (Rey: [0022], "an autonomous software program (e.g., a bot) that can automatically interact with a user when the person to whom the user's query is directed is unavailable"; [0076], "This may involve use of one or more NLP models that utilize AQA techniques to generate a human-level response"; the NLP models (the vicariously answering language model) of the conversation agent generate the response in place of the unavailable person) to generate an answer to the question (Rey: [0076], "When, however, it is determined that access to the requested information and/or document is available to the user (yes at 525), method 500 may proceed to create a response containing the requested information, at 535"; [0075], "If it is determined that access to the requested information and/or document is not available to the sender (no at 525), method 500 may proceed to provide a notification to the sender 530"; the response (an answer) to the query (the question) is created only on a yes determination, and otherwise the sender receives a notification instead).
Because Belkin and Rey are analogous art with both addressing automatically answering, in place of a person, questions that other people direct to that person using that person's work data, accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to apply Rey's determination, made before a response is created, of whether the sender of a query may access the requested information to Belkin's online server, with a reasonable expectation of success, by having the online server, when another user's question reaches the owner's digital twin, first determine whether that user may access the information the question requests, cause the owner's digital twin's LLM to generate the answer only when the user may, and otherwise notify the user that the digital twin cannot respond, which adds, to the owner's settings of the users and the types of questions to which the digital twin is allowed to respond (Belkin: [0062], [0063]), an express determination, made for each question before the LLM generates an answer, of whether the asking user may access the information the question requests, to teach the at least one processor further carries out an answering allowance judging process of judging whether to cause the vicariously answering language model to generate an answer to the question. This modification would have been motivated by the desire to respond automatically and safely to queries directed at an individual who is unavailable without revealing confidential or private data to which the requester should not have access (Rey: [0019]-[0020]).
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over Belkin, as applied in the rejection of claim 1 above, in view of Froman et al. (hereinafter Froman), US 2021/0090097 A1.
Regarding dependent claim 9, Belkin teaches the information processing apparatus according to claim 5 (Belkin: [0070], "the user will be able to ask his/her own digital twin something the user believes he knew or likely knew but forgot"; the system (the information processing apparatus) accepts the owner's questions to his/her own digital twin; [0059], "extracts and organizes relevant data from other systems for each user to be used for context creation by the inference layer"; and answers them with the LLM using context created from the owner's relevant data, as set forth above for claim 5).
Belkin does not expressly teach a survey system, comprising: an intermediary apparatus for accepting a request for a survey in which answers to a predetermined question are collected, and conducting, with the information processing apparatus, negotiations for causing the language model used in the responding process (interpreted per the 35 U.S.C. 112(b) rejection set forth above) to generate an answer to the predetermined question.
However, Froman teaches a survey system, comprising: (Froman: [0072], "The system 10 may be an automated market research system"; [0076], "The server 12 generates a market research project (e.g. survey, concept test, or the like) implementing a data collection method based on the input data received from the researcher device 16"; the system 10 (a survey system) is an automated market research system that conducts surveys) an intermediary apparatus (Froman: [0073], "a research automation server platform 12, which communicates with a plurality of market researcher devices 16 and a plurality of research participant devices 18 via a network 20"; [0091], "The devices 12, 14, 16, 18, 22 may include one or more of a memory, a secondary storage device, a processor, an input device, a display device, and an output device"; the research automation server platform 12 (an intermediary apparatus), a computing device with a processor, stands between the researcher devices and the participant devices) for accepting a request for a survey (Froman: [0075], "The server 12 receives input data from the researcher device 16"; the server 12 accepts the input data from the researcher device 16 (a request for a survey), from which it generates the survey) in which answers to a predetermined question are collected, (Froman: [0082], "market research data (e.g. marketing assets to be tested, survey questions)"; [0080], "The response data may be raw response data from participants (e.g. answers to questions)"; the input specifies the survey questions (a predetermined question), and the server 12 collects the participants' answers (answers) to them) and conducting, with the information processing apparatus, negotiations (Froman: [0077], "The server 12 may send data to and receive data from the participant device 18 or servers 14, 22 in the execution of the project and collection of response data"; [0080], "The server 12 may deliver project data (i.e. the customized data collection method or market research project) to the participant device 18 and receive response data therefrom automatically and without human input"; [0087], "the interface may be an API or other such interface configured to coordinate communication and data transfer between the software component implementing the virtual participant and the software component implementing the data collection method"; the server 12 notifies the participant device or server that implements the virtual participant of the survey question and receives that apparatus's reply to the notification through the interface (negotiations)) for causing the language model used in the responding process to generate an answer to the predetermined question (Froman: [0079], "the research participant at the participant device 18 may not be a human but may be an automated or virtual persona"; [0030], "The response data may be generated by a machine learning model configured to simulate a human response to a particular question or stimuli"; the delivered survey causes the machine learning model of the automated participant to generate the participant's answers (an answer) to the survey questions (the predetermined question)).
Because Belkin and Froman are analogous art with both addressing computer-generated answers to questions on behalf of people, and Froman is reasonably pertinent to collecting such answers through a survey intermediary, accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to apply Froman's survey server and automatic participant interface to Belkin's system, with a reasonable expectation of success, by providing a separate programmed survey server that accepts the researcher's survey request containing a question, sends that question through Belkin's API layer (Belkin: [0045]) to the digital twins selected for the survey, and receives from each either the answer its LLM generates using its owner's relevant job data or the notice Belkin gives when a digital twin is unable to respond to a question (Belkin: [0079]), retaining the same owner's digital twin that answers the owner's own questions (Belkin: [0059], [0069]-[0070]) and extending Belkin's existing facility for asking the same question of several digital twins (Belkin: [0066]) with Froman's API-coordinated survey exchange (Froman: [0087]), to teach a survey system, comprising: the information processing apparatus according to claim 5; and an intermediary apparatus for accepting a request for a survey in which answers to a predetermined question are collected, and conducting, with the information processing apparatus, negotiations for causing the language model used in the responding process to generate an answer to the predetermined question. This modification would have been motivated by the desire to reduce the time needed to gather respondents' answers through automation (Froman: [0064]).
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Belkin in view of Rey, as applied in the rejection of claim 6 above, and further in view of Ollikainen et al. (hereinafter Ollikainen), US 2017/0221081 A1.
Regarding dependent claim 7, Belkin, in view of Rey, teach the information processing apparatus according to claim 6, wherein in the answering allowance judging process, (Rey: [0047], "the access determination service 116 may receive a request for accessing information as an input and may provide a determination as to whether access can be granted to the information as an output"; the access determination service 116 (the answering allowance judging process) makes the determination relied upon for claim 6) the at least one processor judges that an answer to the question should be generated (Rey: [0076], "When, however, it is determined that access to the requested information and/or document is available to the user (yes at 525), method 500 may proceed to create a response containing the requested information, at 535"; on a yes determination, the processor (the at least one processor) judges that the response (an answer) to the query (the question) should be created) by the vicariously answering language model corresponding to the predetermined user (Rey: [0076], "This may involve use of one or more NLP models that utilize AQA techniques to generate a human-level response"; [0022], "an autonomous software program (e.g., a bot) that can automatically interact with a user when the person to whom the user's query is directed is unavailable"; the response is generated by the NLP models (the vicariously answering language model) of the conversation agent, which responds for the person to whom the query is directed (corresponding to the predetermined user)).
Belkin and Rey do not expressly teach in a case where the predetermined user satisfies a condition associated with the question.
However, Ollikainen teaches in a case where the predetermined user satisfies (Ollikainen: [0046], "The Responding Agents have access to this information, and they may check whether the user they represent meets the set criteria"; [0086], "The Responding Agent may then evaluate the match criteria and determine whether an individual associated with the Responding Agent is a part of the target group for this survey"; [0103], "the user is a past customer of the organization according to a purchase history available to the Responding Agent"; the Responding Agent determines whether the individual associated with the Responding Agent (the predetermined user) meets the criteria (satisfies), for example whether that individual is a past customer of the requesting organization; [0104], "After determining that the user associated with the Responding Agent should participate in this survey, the Responding Agent implants the requested information into a message"; that determination decides whether the agent provides the requested information) a condition associated with the question (Ollikainen: [0046], "The Marketplace publishes (2) the criteria that the repliers need to meet"; [0085], "a criteria to be included to this survey"; the match criteria (a condition associated with the question) that repliers must meet are included with the survey request).
Because Ollikainen is reasonably pertinent to the problem faced by the inventor of deciding whether an agent that answers on behalf of a represented user should answer a request that carries a condition on who may answer it, accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to apply Ollikainen's determination of whether the represented user meets the criteria published with a survey request to the answering allowance judgment of Belkin's online server as modified by Rey, with a reasonable expectation of success, by associating a criterion, such as an area of expertise or experience, with a question that an asking user puts to several digital twins at once (Belkin: [0066]), comparing that criterion with the expertise and experience recorded for each digital twin's owner (Belkin: [0048]), and, when the owner meets the criterion and the access determination permits answering, judging, before any answer is created, that the owner's digital twin's LLM should generate the answer (Rey: [0074]-[0076]), which extends Belkin's existing direction of a question to an employee with the same job title, description or credentials (Belkin: [0079]), to teach wherein in the answering allowance judging process, in a case where the predetermined user satisfies a condition associated with the question, the at least one processor judges that an answer to the question should be generated by the vicariously answering language model corresponding to the predetermined user. This modification would have been motivated by the desire to obtain the requested information only from the types of users whose data is of interest to the requesting party (Ollikainen: [0045]).
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Belkin in view of Rey, as applied in the rejection of claim 6 above, and further in view of Kussmaul et al. (hereinafter Kussmaul), US 2022/0100893 A1.
Regarding dependent claim 8, Belkin, in view of Rey, teach the information processing apparatus according to claim 6, as set forth above for claim 6.
Belkin and Rey do not expressly teach wherein the at least one processor further carries out: an alternative condition generating process of generating an alternative condition which replaces a condition associated with the question; and a negotiating process of notifying a sender of the question of the alternative condition and inquiring of the sender whether to approve of the alternative condition.
However, Kussmaul teaches wherein the at least one processor further carries out: (Kussmaul: [0005], "an application (App) BOT comprising a memory and a processor"; the processor (the at least one processor) of the App BOT carries out the negotiation described below) an alternative condition generating process of generating an alternative condition (Kussmaul: [0097], "The App BOT may calculate the privacy leak based on the data-context, user preference, and the proposed incentives, and come up with a decision to agree to the MRD BOT's request or to propose a new negotiation offer"; [0130], "the App BOT may send an updated negotiation request in the form of a counter-offer in order to reduce the privacy leak score and/or increase the incentives for providing access to data"; when the App BOT (an alternative condition generating process) does not agree to a request as made, it generates a counteroffer (an alternative condition)) which replaces a condition associated with the question (Kussmaul: [0003], "The data access request offer comprising a request offer for a set of application specific permissions that are requested permissions to access a labelled user information entity"; the request that the App BOT answers carries requested permissions (a condition associated with the question) that set how much of the user's information is to be provided; [0131], "MRD BOT I want to have Eye Gaze permission. App BOT I can give you Eye Gaze data of Natural Objects"; the counteroffer offers narrower data in place of the requested permission (replaces)); and a negotiating process of notifying a sender of the question of the alternative condition (Kussmaul: [0003], "In a negotiation phase, the method comprises negotiating by receiving a data access request offer from a mixed reality data (MRD)-BOT"; [0005], "sending a counteroffer to the offer to the MRD BOT"; in the negotiation phase (a negotiating process), the counteroffer (the alternative condition) is sent to the MRD BOT (a sender of the question), which sent the request) and inquiring of the sender whether to approve of the alternative condition (Kussmaul: [0004], "receiving a counteroffer in response to the data request offer"; [0004], "sending an acceptance of the counteroffer to the App BOT"; [0131], "MRD BOT I do not want that"; the MRD BOT (the sender) that receives the counteroffer accepts or rejects it, so sending the counteroffer asks the MRD BOT whether it approves the counteroffer (inquiring of the sender whether to approve of the alternative condition); [0096], "Once negotiations are complete, meaning agreement has been reached, or an offer/counter offer has been accepted"; the negotiation ends when a counteroffer is accepted).
Because Kussmaul is reasonably pertinent to the problem faced by the inventor of negotiating, on behalf of the person whose information is requested, a changed condition on which a request will be answered, accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to apply Kussmaul's counteroffer, by which the App BOT answers a request it will not grant as made with an offer of narrower data and sends that offer to the requesting bot for acceptance (Kussmaul: [0130], [0131]), to the answering allowance judgment of Belkin's online server as modified by Rey, with a reasonable expectation of success, by treating the scope of the information that a question requests (Rey: [0074]) as a condition associated with that question and, when that scope is not available to the asking user, having the online server generate a counteroffer that replaces it with the narrower scope of information that the asking user's access level permits (Rey: [0043]), send the counteroffer to the asking user, and ask whether the asking user accepts it before the owner's digital twin's LLM answers, in place of a notification that the conversation agent cannot respond or that the recipient's permission is required (Rey: [0075]), much as Rey already asks the sender whether the sender would like a response from the conversation agent (Rey: [0057]), to teach wherein the at least one processor further carries out: an alternative condition generating process of generating an alternative condition which replaces a condition associated with the question; and a negotiating process of notifying a sender of the question of the alternative condition and inquiring of the sender whether to approve of the alternative condition. This modification would have been obvious as the use of a known technique, Kussmaul's counteroffer that narrows a request for a user's information, to improve a similar system, Rey's automatic answering in place of an unavailable person, in the same way, so that the asking user may receive an answer within the scope that user may access rather than no answer (MPEP 2143, subsection I.C).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Spataro “Introducing Microsoft 365 Copilot – your copilot for work” (Sept. 5, 2023) (Page 1 Humans are hard-wired to dream, to create, to innovate. Each of us seeks to do work that gives us purpose — to write a great novel, to make a discovery, to build strong communities, to care for the sick. The urge to connect to the core of our work lives in all of us. But today, we spend too much time consumed by the drudgery of work on tasks that zap our time, creativity and energy. To reconnect to the soul of our work, we don’t just need a better way of doing the same things. We need a whole new way to work.
Today, we are bringing the power of next-generation AI to work. Introducing Microsoft 365 Copilot — your copilot for work . It combines the power of large language models (LLMs) with your data in the Microsoft Graph and the Microsoft 365 apps to turn your words into the most powerful productivity tool on the planet).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KUANG FU CHEN whose telephone number is (571)272-1393. The examiner can normally be reached M-F 9:00-5:30pm ET.
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, Jennifer Welch can be reached on (571) 272-7212. 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.
/KC CHEN/Primary Patent Examiner, Art Unit 2143
Prosecution Insights