DETAILED ACTION
This communication is in response to the Amendments and Arguments filed on March 17, 2026. Claims 1-20 are pending and have been examined. Hence, this action has been made FINAL.
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 .
Response to Arguments
The reply filed on March 17, 2026 has been entered. Applicant’s arguments with respect to claims 1-20 have been considered but are not persuasive and moot in view of new ground(s) of rejection caused by the amendments.
With respect to the applicant’s arguments to claim rejections under 35 U.S.C § 101, Applicant has amended each of the independent claims and asserts that “The rejection dissects the claims to identify four disjoint phrases from the independent claims that the Examiner alleges, without proof or documentation (e.g., a reference to some other example or case), are mental processes. Further, the way in which the elements are split appears to be solely to allege a mental process rather than to recognize that such a procedure does not make sense if a person does it, but rather enables a computer system to ascertain a request based on a document repository. This is supported by the erroneous analogy offered by the Examiner, that of a user listening in on a conversation and hearing "that a user made a request" and "compare the user's request to documents from a filing cabinet," etc. 3 The claim does not state this. The claim requires a semantic analysis of the text to perform a "semantic match" between the text ( e.g., transcript) of a portion of the conversation and "documents in a repository." Only then can the actual request be determined. When these elements are combined with, for example, the embedding vector recited in claim 5, the position taken in the rejection is untenable. Accordingly, the proposed "mental processes" are unreasonable.” The examiner respectfully disagrees with these assertions. The examiner sees no difference in claimed process between a human performing the asserted “semantic analysis of the text to perform a "semantic match" between the text ( e.g., transcript) of a portion of the conversation and "documents in a repository"” and a machine. Specifically, a human (e.g. receptionist, call analyst, etc.) is capable of capturing a portion of a conversation between two other humans (e.g. a customer service representative and a customer), and using that portion of conversation to identify documents from a repository. The exact nature of the repository is irrelevant when considering the claim language under the broadest reasonable interpretation. With regard to claim 5, while impractical, a computer scientist would clearly be able to walk through the process of embedding text from a conversation into a numerical vector and performing mathematical similarity calculations between the conversation vectors and the document vectors.
Applicant further asserts that “The rest of the analysis simply makes unsupported assertions that the other elements of the claims are merely extra solution activities (e.g., not needed for the solution), generic computer components (e.g., "apply it"), or merely linking the idea to a technological environment. However, none of these decisions are explained in such a way as to give the Applicant a full and fair opportunity to respond. For example, what evidence does the Examiner provide that vector embeddings can be practically created by humans by hand? The rejection cannot be maintained by mere opinion.” The examiner respectfully disagrees with these assertions. As per MPEP 2144.03, “To adequately traverse a finding based on official notice, an applicant must specifically point out the supposed errors in the examiner’s action, which would include stating why the noticed fact is not considered to be common knowledge or well-known in the art. A mere request by the applicant that the examiner provide documentary evidence in support of an officially-noticed fact is not a proper traversal. See 37 CFR 1.111(b). See also Chevenard, 139 F.2d at 713, 60 USPQ at 241.” The applicant’s assertions request for documentary evidence that vector embeddings can be practically created by humans by hand, but this request cannot be considered an adequate traversal as per the above cited MPEP section. Accordingly, the applicant fails to state why generation of vector embeddings by hand is not considered common knowledge, and thus, traversal of claim rejections under 35 USC § 101 using this rationale is improper.
Applicant further asserts that “The present claims are clearly an attempt to improve computer technology. The MPEP states that such claims are eligible.” The examiner respectfully disagrees with these assertions. The applicant asserts that the present claims are clearly an attempt to improve computer technology. However, the applicant fails to subsequently explain said improvements to computer technology. Accordingly, the examiner is given no reason to withdraw the claim rejections under 35 USC § 101 using this rationale. As amended, there is no language in the independent claims that would prevent a human from performing these steps, as addressed in further detail below with respect to claim rejections under 35 USC § 101.
With respect to the applicant’s arguments to claim rejections under 35 U.S.C § 103, the applicant’s arguments with respect to claims 1-20 have been considered but are moot in view of new ground(s) of rejection caused by the amendments.
Claim Interpretation
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.
The following terms in the claims have been given the following interpretations in light of the specification:
Cardinality: ¶ [0024], “the cardinality parameter specifies, for example, ten documents. Thus, the top ten documents most similar to the part of text under the similarity metric are in the semantic group A 160.”
Thus, a cardinality is some specific number or amount. This definition is used for purposes of searching for prior art, but cannot be incorporated into the claims.
Request: ¶ [0025], “At this stage, the documents in the semantic group A 160 can provide a list of available requests that the second person can address. For example, if the documents in the semantic group A 160 are different kind of loan applications, then the request is determined to be a loan. If the documents are different types of wills, then the request can be considered to be the creation or amendment of a will.”
Thus, a request is any type, concept, or category of document that a person can assist a second person with.
Should applicant wish different definitions, Applicant should point to the portions of the specification that clearly show a different definition.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. All of the claims are method claims (13-20), apparatus/machine claims (1-12) or manufacture claim under (Step 1), but under Step 2A all of these claims recite abstract ideas and specifically mental processes. These mental processes are more particularly recited in claims 1 and 13 as:
determining the request from the text…
evaluating a semantic match between parts of the text and documents in a repository…
selecting a document from the repository…
populating a field in the set of fields from the text…
Under Step 2A Prong One, claims 1 and 13 are directed to an abstract idea and specifically a mental process. As detailed above, the steps of determining, evaluating, selecting, populating, etc. may be practically performed in the human mind with the use of a physical aid such as a pen and paper. For example, a human could listen in to a conversation between a human customer support agent and a user, hear that the user made a request, compare the user’s request to documents in a filing cabinet, select the document with the highest similarity match to the user’s request, and then pre-populate fields on the document based on user utterances.
Under Step 2A Prong Two, this judicial exception is not integrated into a practical application because claims 1-20 do not recite additional elements that integrate the exception into a practical application. In particular, claims 1 and 13 recite the additional elements of non-transitory machine readable media including instructions (¶ [0046]) and processing circuitry (¶ [0042]). These additional elements are recited at a high level of generality and merely equate to “apply it” or otherwise merely uses a generic computer as a tool to perform an abstract which are not indicative of integration into a practical application as per MPEP 2106.05(f). Further, claims 1-20 recite the additional elements of “obtaining…”, both of which amount to insignificant extra-solution activities which are not indicative of integration into a practical application as per MPEP 2106.05(g). Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
Under Step 2B, the claims do not recite additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to the integration of the abstract idea into a practical application, the additional elements of using a computer is noted as a general computer {non-transitory machine readable media including instructions (¶ [0046]); processing circuitry (¶ [0042])}. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. Further, the additional limitations in the claims noted above are directed towards insignificant extra-solution activities. The claims are not patent eligible.
With respect to claims 2-4, 14, and 15, the claim relates to identifying part-of-speech (POS) tokens and word combinations using a natural language technique. This relates to a human labelling the transcript of a customer service conversation using POS tags and elementary discourse units. The limitation of “large language model (LLM) artificial neural network (ANN)” is recited at a high level of generality (¶ [0023]) and merely equates to “apply it” or otherwise merely uses a generic computer as a tool to perform an abstract which are not indicative of integration into a practical application as per MPEP 2106.05(f). No additional limitations are present. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
With respect to claims 5 and 16, the claim relates to calculating embedding vectors for parts of the conversation. This relates to a human creating vectors of word embeddings by hand. No additional limitations are present. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
With respect to claims 6-9, 17, and 18, the claim relates to various methods of evaluating similarity between the conversation and documents in a repository. This relates to a human using a cosine similarity metric to select a top 10 most similar documents beyond a certain similarity threshold. No additional limitations are present. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
With respect to claim 10, the claim relates to filling out a second part of a document using a second text from the conversation. This relates to a human continuing to listen to a conversation and filling out additional fields based on the continued conversation. No additional limitations are present. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
With respect to claim 11 and 19, the claim relates to fetching data from an external source upon failing to populate a second field. This relates to a human contacting a tax form expert to help correctly fill out fields they were unable to by themselves. No additional limitations are present. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
With respect to claim 12 and 20, the claim relates to fetching data from an external source upon failing to populate a second field. This relates to a human asking the customer service agent to ask the customer a specific question in order to fill out a field that was not satisfied by the initial conversation or external search. No additional limitations are present. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
For all of the above reasons, taken alone or in combination, claims 1-20 recite a non-statutory mental process.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 5-8, 10, 13, and 16-18 are rejected under 35 U.S.C. 103 as obvious over US Patent 11743378 B1 (Johnston et al.) in view of US Patent 12393617 B1 (Niu et al.) in view of US Patent Publication 20190065991 A1 (Guggilla et al.).
Claim 1
Regarding claim 1, Johnston et al. disclose non-transitory machine readable medium including instructions for natural language processing over a document repository (Johnston et al. ¶ (123), "This apparatus may be specially constructed for the purposes, e.g., a specific computer, or it may comprise a computer selectively activated or reconfigured by a computer program stored in the computer. Such a computer program may be stored in a computer readable storage medium"), the instructions, when executed by processing circuitry, cause the processing circuitry to perform operations comprising:
obtaining text from a portion of a conversation between a first person and a second person (Johnston et al. ¶ (106), "The communication server 1310 allows utterances to flow freely between a customer 1305 and an agent 1390, but also allows the agent assistance module to “listen in” on the conversation. ... A speech recognizer 1320 is used to convert the audio signal to a text transcript."), the first person making a request in the portion of the conversation to the second person (Johnston et al. ¶ (47)-(50), "For example, assume the following reservation conversation between an agent and a consumer: ... Consumer: Yeah, I am calling from my work. My number is 555-555-5555. Also you can add my work number into the system, which is 222-222-2222.");
selecting a document [from the repository based on the request once determined], the document including a set of fields (Johnston et al. ¶ (100), "The AI Assist to HI module 1360 “listens in” in real time on the conversation between customer and agent, recognizes what is being communicated, tracks the conversation, and represents recognized information as objects in a graphical user interface (hereinafter “agent desktop interface”) that can be used to accelerate the process of accomplishing the customer's objective for the conversation. ... This includes easing the burden of agents filling in pertinent forms, such as by pre-populating results within the agent desktop interface where there is sufficient confidence in those results."); and
populating a field in the set of fields from the text (Johnston et al. ¶ (100), "This includes easing the burden of agents filling in pertinent forms, such as by pre-populating results within the agent desktop interface where there is sufficient confidence in those results.").
Johnston et al. do not explicitly disclose all of retrieving documents from a repository.
However, Niu et al. disclose obtaining text from a portion of a conversation between a first person and a second person (Niu et al. ¶ (93), "if the conversation between user 118C (e.g., a customer) and user 118B (e.g., an agent) is a telephone call, the transcription engine 134 executes one or more automatic speech recognition algorithms to reformat the audio data into text data (e.g., a transcript of the conversation, a conversational log)."), [the first person making a request in the portion of the conversation to the second person];
determining the request from the text by evaluating a semantic match between parts of the text and documents in a repository to determine a group of documents (Niu et al. ¶ (115), "the retrieval module 184 encodes an input (e.g., a trigger utterance), encodes assistance information, and estimates a relevance using a similarity measure between the embedded input and embedded assistance information." ¶ (113), "Examples of assistance information includes documents, articles, phone numbers, website addresses, physical addresses, contact information, template emails, refund forms, (or other documents), etc. that are highly relevant to a trigger utterance in the input" Assistance information is considered analogous to a group of documents), [the group of documents having a corresponding list of requests, the request being in the corresponding list of requests]; and
selecting a document from the repository based on the request once determined (Niu et al. ¶ (136)-(138), "the score selector 708 identifies the document associated with the highest similarity score 706 as relevant the most relevant document to the trigger utterance (e.g., relevant assistance information 720). ... the retrieval module 184 transmits the relevant assistance information (or the subset of relevant assistance information) to the compute device 122 of user 118B at circle (6).").
It would have been obvious to a person having ordinary skill in the art before the time of the effective filing date of the claimed invention of the instant application to modify Johnston et al.’s agent assistance system to incorporate Niu et al.’s document retrieval.
The suggestion/motivation for doing so would have been that, “the identification of relevant assistance information (e.g., what assistance to provide to an agent) reduces computing resources required by a system and beneficially provides superior chat assistance in the form of relevant recommended subject matter at relevant time(s) in the conversation,” as noted by the Niu et al. disclosure in paragraph [0016].
Johnston et al. in view of Niu et al. do not explicitly disclose all of the retrieved group of documents corresponding to a list of requests.
However, Guggilla et al. disclose obtaining text from a portion of a conversation between a first person and a [second person] computer (Guggilla et al. ¶ [0044], "At 305, the document interpreter 131 may obtain user input via the network interface. In an example, the user input may be an audio input and the system 100 may use natural language processing to determine the intent of the user and identify at least one document feature."), the first person making a request in the portion of the conversation (Guggilla et al. ¶ [0044], "The user input may be a query to identify documents of interest to a user such as invoices, responses to certain queries to customer support, documents with personally identifiable information and the like." Querying the type of a document is considered analogous to a request; see claim interpretation) [to the second person];
determining the request from the text by evaluating a semantic match between parts of the text and documents in a repository to determine a group of documents (Guggilla et al. ¶ [0028], "the document interpreter 131 may use features, for example, received from the user to identify a subset of documents that match the user inputs. The output of the document comparator 161 may include similarity values 162 that represents how the structural feature category 160, the semantic feature category 151, the document feature category 275 or a combination thereof and the features of features received from the user."), the group of documents having a corresponding list of requests, the request being in the corresponding list of requests (Guggilla et al. ¶ [0044]-[0045], " The user input may be a query to identify documents of interest to a user such as invoices, responses to certain queries to customer support, documents with personally identifiable information and the like. … At 306, the document interpreter 131 may identify a subset of documents that match the query document feature category 275 to identify documents that match the query document feature category 1101 in the data repository 175. For example, the document interpreter 131 may identify documents that are similar to the documents that are determined have a similar document feature category 275 by the document decoder 130." A document feature category is considered analogous to a request; see claim interpretation. Thus, the query document feature category 1101 is considered analogous to the user request. Further, the set of similar document feature categories 275 is considered analogous to a list of requests); and
selecting a document from the repository based on the request once determined (Guggilla et al. ¶ [0045], "The system 100 may then transmit the matching subset of documents to a client application for display on the client application."), the document including a set of fields (Johnston et al. ¶ (100), "The AI Assist to HI module 1360 “listens in” in real time on the conversation between customer and agent, recognizes what is being communicated, tracks the conversation, and represents recognized information as objects in a graphical user interface (hereinafter “agent desktop interface”) that can be used to accelerate the process of accomplishing the customer's objective for the conversation. ... This includes easing the burden of agents filling in pertinent forms, such as by pre-populating results within the agent desktop interface where there is sufficient confidence in those results.").
It would have been obvious to a person having ordinary skill in the art before the time of the effective filing date of the claimed invention of the instant application to modify Johnston et al. in view of Niu et al. to include Guggilla et al.’s request-based document categorization because such a modification is the result of simple substitution of one known element for another producing a predictable result. More specifically, Niu et al.’s assistance information and Guggilla et al.’s categorized documents perform the same general and predictable function, the predictable function being providing relevant information to the user session. Since each individual element and its function are shown in the prior art, albeit shown in separate references, the difference between the claimed subject matter and the prior art rests not on any individual element or function but in the very combination itself - that is in the substitution of Niu et al.’s assistance information by replacing it with Guggilla et al.’s categorized documents. Thus, the simple substitution of one known element for another producing a predictable result renders the claim obvious.
Claim 5
Regarding claim 5, the rejection of claim 1 is incorporated.
Niu et al. further disclose wherein evaluating the semantic match between parts of the text and documents in the repository includes calculating an embedding vector for the parts of the text (Niu et al. ¶ (115), "In operation, the retrieval module 184 encodes an input (e.g., a trigger utterance), encodes assistance information, and estimates a relevance using a similarity measure between the embedded input and embedded assistance information.").
Claim 6
Regarding claim 6, the rejection of claim 5 is incorporated.
Niu et al. further disclose wherein evaluating the semantic match between parts of the text and documents in the repository includes identifying a set of documents based on a similarity metric between the embedding vector of a part of the text and documents in the set of documents (Niu et al. ¶ (115), "In operation, the retrieval module 184 encodes an input (e.g., a trigger utterance), encodes assistance information, and estimates a relevance using a similarity measure between the embedded input and embedded assistance information.").
Claim 7
Regarding claim 7, the rejection of claim 6 is incorporated.
Niu et al. further disclose wherein the similarity metric is one of cosine similarity, Euclidean distance, or dot product similarity (Niu et al. ¶ (131), "the similarity module 704 is configured to compute the cosine similarity between the embeddings.").
Claim 8
Regarding claim 8, the rejection of claim 6 is incorporated.
Niu et al. further disclose wherein documents are selected from the repository to be included in the set of documents based on a defined cardinality of the set of documents, the set of documents having a highest rank under the similarity metric (Niu et al. ¶ (137), “the score selector 708 is configured to select a top-k highest similarity scores using the k nearest neighbor algorithm. In these embodiments, the score selector 708 identifies the top-k most relevant documents to the trigger utterance." Selecting the top-k most similar documents is considered analogous to selecting a set of documents based on a defined cardinality having highest rank under a similarity metric. See Claim Interpretation).
Claim 10
Regarding claim 10, the rejection of claim 1 is incorporated.
Johnston et al. further disclose wherein a second field of the set of fields is populated from a second text of a second portion of the conversation (Johnston et al. ¶ (47)-(55), "assume the following reservation conversation between an agent and a consumer: ... Consumer: Yeah, I am calling from my work. My number is 555-555-5555. Also you can add my work number into the system, which is 222-222-2222. ... By analyzing 410 the script, the system can decompose the script into several components: ... (2) Task: collect consumer identification A. Collect phone number B. Collect reward number; (3) Task: Collect Phone Number to search; (4) Task: Collect Alternate Phone Number which also uses (3)." ¶ (30), "Each task has form fill requirements, like a hotel property name, which would include AI search for the property 655, reservation dates, number of rooms, etc.").
Claim 13
Regarding claim 13, the limitations of claim 13 are similar in scope to that of claim 1 and therefore are rejected for similar reasons as described above.
Claim 16
Regarding claim 16, the rejection of claim 13 is incorporated. The limitations of claim 16 are similar in scope to that of claim 5 and therefore are rejected for similar reasons as described above.
Claim 17
Regarding claim 17, the rejection of claim 16 is incorporated. The limitations of claim 17 are similar in scope to that of claim 6 and therefore are rejected for similar reasons as described above.
Claim 18
Regarding claim 18, the rejection of claim 17 is incorporated. The limitations of claim 18 are similar in scope to that of claim 7 and therefore are rejected for similar reasons as described above.
Claims 2, 3, 11, 12, 14, 15, 19, and 20 are rejected under 35 U.S.C. 103 as obvious over Johnston et al. in view of Niu et al. in view of Guggilla et al. as applied to claims 1 and 13 above, and further in view of US Patent Publication 20220058114 A1 (Ranjan Jena et al.).
Claim 2
Regarding claim 2, the rejection of claim 1 is incorporated. Johnston et al. in view of Niu et al. in view of Guggilla et al. disclose all the elements of the claimed invention as stated above.
Niu et al. further disclose wherein determining the request from the text includes determining the parts of text using a natural language processing technique (Niu et al. ¶ (116), "the retrieval module 184 is configured to perform at least two tasks to identify (and retrieve) relevant assistance information. First, the retrieval module 184 is configured to match tokens in an input (e.g., an utterance, a sentence, a turn, etc.) to tokens in assistance information. ... Second, the retrieval module 184 is configured to identify semantically related tokens (or words/phrases).") [to identify a noun, an adjective, a verb, or an adverb in the text].
Johnston et al. in view of Niu et al. in view of Guggilla et al. do not explicitly disclose all of identifying a noun, adjective, verb, or adverb in text.
However, Ranjan Jena et al. disclose determining a parts of text using a natural language processing technique to identify a noun, an adjective, a verb, or an adverb in the text (Ranjan Jena et al. ¶ [0031], "POS token extractor 404 may include any suitable type of POS token extractor capable of extracting POS tokens from text. POS token extractor 404 may apply machine learning to use NLP to identify the POS token corresponding to each word of the test scenario(s)." Extracting Part-of-speech (POS) tokens from text is considered analogous to identifying nouns, adjectives, verbs, or adverbs in text).
It would have been obvious to a person having ordinary skill in the art before the time of the effective filing date of the claimed invention of the instant application to modify Johnston et al. in view of Niu et al. in view of Guggilla et al. to incorporate Ranjan Jena et al.’s part-of-speech tagging.
The suggestion/motivation for doing so would have been that, “using keyphrases and part-of-speech (POS) tokens accurately capture the intent of test scenarios and templates without the need for knowledge of business processes, application processes, and testing tools related to the test scenarios or templates,” as noted by the Ranjan Jena et al. disclosure in paragraph [0005].
Claim 3
Regarding claim 3, the rejection of claim 2 is incorporated. Johnston et al. in view of Niu et al. in view of Guggilla et al. in view of Ranjan Jena et al. disclose all the elements of the claimed invention as stated above.
Niu et al. further disclose wherein the parts of text are a combination of multiple related words (Niu et al. ¶ (116), "First, the retrieval module 184 is configured to match tokens in an input (e.g., an utterance, a sentence, a turn, etc.) to tokens in assistance information." A sentence is considered analogous to a combination of multiple related words).
Ranjan Jena et al. further disclose wherein the parts of text are a combination of multiple related words (Ranjan Jena et al. ¶ [0031], "Keyphrase extractor 402 may include any suitable type of algorithm capable of extracting keyphrases from text. Keyphrase extractor 402 may apply machine learning to use NLP to identify words or phrases representing the most relevant information contained in the document. For example, a keyphrase extracted from the above example could be “OSP Validation.”"), relationships between words determined by the natural language processing technique (Ranjan Jena et al. ¶ [0031], "Keyphrase extractor 402 may apply machine learning to use NLP to identify words or phrases representing the most relevant information contained in the document.").
Claim 11
Regarding claim 11, the rejection of claim 1 is incorporated.
Johnston et al. further disclose material in conversation (Johnston et al. ¶ (103), "In this case, the system has identified a room type (namely, a “king”), a hotel name (“hyatt regency”) and a city (“new York”) within the conversation (namely, the user statement “I wanna book a king room for the hyatt regency hotel in new york city”)").
Niu et al. further disclose retrieving data from an external source (Niu et al. ¶ (90), "the provider network 100 may instruct the retrieval model 184 to retrieve knowledge base data 160 bi-annually.")
Johnston et al. in view of Niu et al. in view of Guggilla et al. do not explicitly disclose all of retrieving data from an external source in response to empty fields.
However, Ranjan Jena et al. disclose retrieving data from an external source (Ranjan Jena et al. ¶ [0037], "Database search tool 612 may execute the generated SQL statement to search application database 212 for empty field data. Second field set filler 614 may use the empty field data to fill the empty fields and output fully filled template(s)." Application database 212 is considered analogous to an external source) in response to identifying a second field of the set of fields based on an inability to populate the second field (Ranjan Jena et al. ¶ [0036], "Any partially filled template(s) output by NLP filler 222 may be input into empty field identifier 604. Empty field identifier 604 may identify empty field(s) in the partially filled template(s). ... From template repository 210, database table and column identifier 606 may identify the database table and column mapped to each identified empty field.") [from material in the conversation].
It would have been obvious to a person having ordinary skill in the art before the time of the effective filing date of the claimed invention of the instant application to modify Johnston et al. in view of Niu et al. in view of Guggilla et al. to incorporate Ranjan Jena et al.’s empty field detection because such a modification is the result of combining prior art elements according to known methods to yield predictable results. More specifically, Niu et al.’s document retrieval as modified by Ranjan Jena et al.’s empty field detection can yield a predictable result of improving system flexibility since detecting empty fields would allow the system to recommend dialogue based off of the current form state. Thus, a person of ordinary skill would have appreciated including in Niu et al.’s document retrieval the ability to do Ranjan Jena et al.’s empty field detection since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable.
Claim 12
Regarding claim 12, the rejection of claim 11 is incorporated.
Johnston et al. further disclose identifying a third field of the set of fields that is unfilled (Johnston et al. ¶ (104), "the system might suggest that the agent ask the customer for the number of nights that the customer is staying, since that information is part of a hotel reservation but has not yet been specified during the conversation.") [following retrieval of data from the external source]; and
prompting the second person to request data from the first person to fill the third field (Johnston et al. ¶ (104), "the system might suggest that the agent ask the customer for the number of nights that the customer is staying, since that information is part of a hotel reservation but has not yet been specified during the conversation.")
Ranjan Jena et al. further disclose identifying a third field of the set of fields that is unfilled following retrieval of data from the external source (Ranjan Jena et al. ¶ [0037], "In some embodiments, certain empty fields may not be fillable by using NLP to fill fields in the test template or by using an application database lookup to fill identified empty fields in the test template."); and
prompting the second person to request data from the first person to fill the third field (Ranjan Jena et al. ¶ [0037], "A list of such fields may be generated, and human intervention may be requested to fill these fields.").
Claim 14
Regarding claim 14, the rejection of claim 13 is incorporated. The limitations of claim 14 are similar in scope to that of claim 2 and therefore are rejected for similar reasons as described above.
Claim 15
Regarding claim 15, the rejection of claim 14 is incorporated. The limitations of claim 15 are similar in scope to that of claim 13 and therefore are rejected for similar reasons as described above.
Claim 19
Regarding claim 19, the rejection of claim 13 is incorporated. The limitations of claim 19 are similar in scope to that of claim 11 and therefore are rejected for similar reasons as described above.
Claim 20
Regarding claim 20, the rejection of claim 19 is incorporated. The limitations of claim 20 are similar in scope to that of claim 12 and therefore are rejected for similar reasons as described above.
Claim 4 is rejected under 35 U.S.C. 103 as obvious over Johnston et al. in view of Niu et al. in view of Guggilla et al. in view of Ranjan Jena et al. as applied to claim 3 above, and further in view of US Patent Publication 20250086647 A1 (Gao et al.).
Claim 4
Regarding claim 4, the rejection of claim 3 is incorporated. Johnston et al. in view of Niu et al. in view of Guggilla et al. in view of Ranjan Jena et al. disclose all the elements of the claimed invention as stated above.
Niu et al. further disclose wherein the natural language processing technique is a [large language model (LLM)] artificial neural network (ANN) (Niu et al. ¶ (131), "the similarity module 704 may be a neural network trained to rank the similarity of embeddings 710 and 712").
Johnston et al. in view of Niu et al. in view of Guggilla et al. in view of Ranjan Jena et al. do not explicitly disclose all of a large language model.
However, Gao et al. disclose a large language model (LLM) (Gao et al. ¶ [0045], "The multi-agent generative AI system 120 includes a large language model (LLM) 116. The chat agent 106 and triage agent 108 are configured to interact with the LLM 116 while processing one or more user queries.")
It would have been obvious to a person having ordinary skill in the art before the time of the effective filing date of the claimed invention of the instant application to modify Johnston et al. in view of Niu et al. in view of Guggilla et al. in view of Ranjan Jena et al. to incorporate Gao et al.’s large language model because such a modification is the result of simple substitution of one known element for another producing a predictable result. More specifically, Niu et al.’s artificial neural network and Gao et al.’s large language model perform the same general and predictable function, the predictable function being processing user dialogue. Since each individual element and its function are shown in the prior art, albeit shown in separate references, the difference between the claimed subject matter and the prior art rests not on any individual element or function but in the very combination itself - that is in the substitution of Niu et al.’s artificial neural network by replacing it with Gao et al.’s large language model. Thus, the simple substitution of one known element for another producing a predictable result renders the claim obvious.
Claim 9 is rejected under 35 U.S.C. 103 as obvious over Johnston et al. in view of Niu et al. in view of Guggilla et al. as applied to claim 1 above, and further in view of US Patent Publication 20250086647 A1 (Gao et al.).
Claim 9
Regarding claim 9, the rejection of claim 1 is incorporated. Johnston et al. in view of Niu et al. disclose all the elements of the claimed invention as stated above.
Johnston et al. in view of Niu et al. do not explicitly disclose all of a similarity threshold.
However, Gao et al. disclose wherein documents are selected from the repository to be included in the set of documents based on a similarity metric being within a threshold of similarity (Gao et al. ¶ [0088], "The retrieval module 218 is further configured to query the vector database 112 using the query embedding to retrieve one or more documents. ... the respective document embedding matches the query embedding when a level of similarity between the respective document embedding and the query embedding exceeds a threshold level.").
It would have been obvious to a person having ordinary skill in the art before the time of the effective filing date of the claimed invention of the instant application to modify Johnston et al. in view of Niu et al. in view of Guggilla et al. to include Gao et al.’s similarity threshold because such a modification is the result of combining prior art elements according to known methods to yield predictable results. More specifically, Niu et al.’s similarity-based retrieval as modified by Gao et al.’s similarity threshold can yield a predictable result of improving system consistency since a similarity threshold would enforce a minimum level of similarity required for selection. Thus, a person of ordinary skill would have appreciated including in Niu et al.’s similarity-based retrieval the ability to do Gao et al.’s similarity threshold since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JACOB B VOGT whose telephone number is (571)272-7028. The examiner can normally be reached Monday - Friday 9:30am - 7pm EST.
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, Paras D Shah can be reached at (571)270-1650. 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.
/JACOB B VOGT/ Examiner, Art Unit 2653
/Paras D Shah/ Supervisory Patent Examiner, Art Unit 2653
05/07/2026
Prosecution Insights