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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 02/25/2026 has been entered.
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Response to Amendment
This communication is responsive to the applicant’s amendment dated 02/25/2026. The applicant(s) amended claims 1, 8, and 16.
The 35 USC 101 Rejection is being maintained because the newly amended limitations do not include additional elements that are sufficient to amount to significantly more than the judicial exception. The merging step is a mathematical calculation and the electronic device is recited as performing generic computer functions routinely used in computer applications.
Response to Arguments
Applicant's arguments with respect to independent claims 1, 8, and 16 have been considered but are moot in view of the new ground(s) of rejection because the arguments pertain to the newly amended limitations.
Claim Rejections - 35 USC § 101
Claims 16-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim(s) recite(s) obtaining vectors and generating a response, which appear to be mathematical calculations and mental processes. The claim(s) do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements (electronic device), which is recited at a high level of generality, provide conventional computer functions that do not add meaningful limits to practicing the abstract idea.
Claim 16 recites, in part, based on an utterance, obtaining two vectors, obtaining a synthetic vector, and providing a response. The vector obtaining steps appear to be mathematical calculations, which are specifically identified in the 2019 PEG as an exemplar in the “mathematical concepts” grouping of abstract ideas. The providing of a response step can be performed in the human mind by thinking of a response to the utterance. The electronic device is recited at a high level of generality and is recited as performing generic computer functions routinely used in computer applications. Generic computer components recited as performing generic computer functions that are well-understood, routine and conventional activities amount to no more than implementing the abstract idea with a computerized system. Therefore, the claim as a whole does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Claims 17-20 are rejected under the same rationale.
Claim Rejections - 35 USC § 103
Claim(s) 1-7 and 16-20 are rejected under 35 U.S.C. 103 as being unpatentable over Madhusudhan et al. (US 20220238103 A1) in view of Lee (US 20200160841 A1).
Regarding claims 1 and 16, Madhusudhan teaches:
“An intelligent server comprising: a processor; and a memory electrically connected to the processor and storing instructions which, when executed by the processor” (par. 0078; ‘Client instance 42 is supported by virtual servers similar to those explained with respect to FIG. 2, and is illustrated here to show support for the disclosed functionality described herein within the client instance 42.’), cause the processor to:
“based on a user's utterance, obtain a named entity vector and a sentence vector” (par. 0083; ‘In certain embodiments, the NLU engine 116 is designed to perform a number of functions of the NLU framework 104, including generating word vectors (e.g., intent vectors, subject or entity vectors, subtree vectors) from word or phrases of utterances, as well as determining distances (e.g., Euclidean distances) between these vectors.’; par. 0163; ‘As such, in block 500, a new word vector would be generated for the term “Everest” that would be relatively close in one or more dimensions to word vectors for terms such as “conference”, “meeting”, “presentation”, and so forth.’; par. 0166; ‘Returning to the “Everest” example, if the utterance 122 is “I'm not sure, but we have a meeting scheduled in Everest at 2:30 pm this afternoon to discuss what to do”, the other words 522 in the utterance 122 may be used to determine that the use of the word “Everest” in the utterance 122 was referring to the conference room.’), and
“based on the synthetic vector, provide a response corresponding to the user's utterance” (par. 0085; ‘For the illustrated embodiment, the RA/BE 102 also provides a response 124 (e.g., a virtual agent utterance or confirmation) to the client device 14D via the network 18, for example, indicating actions performed by the RA/BE 102 in response to the received user utterance 122.’).
However, Madhusudhan does not expressly teach:
“obtain a synthetic vector by merging the named entity vector and the sentence vector in the order of the named entity vector and the sentence vector.”
In a similar field of endeavor (utterance processing), Lee teaches:
“obtain a synthetic vector by merging the named entity vector and the sentence vector in the order of the named entity vector and the sentence vector” (par. 0039; ‘The block 412 may receive features from the entity and/or entity tracking information, an utterance embedding based on a user utterance from the sentence encoder 208, one or more of the biasing vectors e.sup.s or e.sup.u from the NLR inferencer 220, a previously fully-formed action in a sentence embedding from the sentence encoder 248, and content based on an API call. The received embeddings in vector form, may then be combined into a feature vector for input into the context RNN 224, such that an output may be generated.’).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the entity vector of Madhusudhan with an utterance vector, as taught by Lee, into a feature vector for input into a context RNN, such that an output may be generated. (Lee: par. 0039)
Regarding claims 2 (dep. on claim 1) and 17 (dep. on claim 16), the combination of Madhusudhan in view of Lee further teaches:
“based on the synthetic vector, obtain intent information corresponding to the user's utterance, in response to the user's utterance being an inquiry” (M: par. 0060; ‘As mentioned, a computing platform may include a chat agent, or another similar virtual agent, that is designed to automatically respond to user requests to perform functions or address issues on the platform.’; par. 0083; ‘As such, a similarity measure or distance between two different utterances can be calculated using the respective intent vectors produced by the NLU engine 116 for the two intents, wherein the similarity measure provides an indication of similarity in meaning between the two intents.’),
“generate an answer corresponding to the inquiry based on the synthetic vector” (M: par. 0085; ‘For the illustrated embodiment, the RA/BE 102 also provides a response 124 (e.g., a virtual agent utterance or confirmation) to the client device 14D via the network 18, for example, indicating actions performed by the RA/BE 102 in response to the received user utterance 122.’), and
“based on the intent information or the answer, provide the response” (M: par. 0085; ‘For the illustrated embodiment, the RA/BE 102 also provides a response 124 (e.g., a virtual agent utterance or confirmation) to the client device 14D via the network 18, for example, indicating actions performed by the RA/BE 102 in response to the received user utterance 122.’).
Regarding claims 3 (dep. on claim 2) and 18 (dep. on claim 17), the combination of Madhusudhan in view of Lee further teaches:
“based on the user's utterance, obtain inquiry intent information” (M: par. 0062; ‘In this manner, present embodiments extract intents/entities from the user utterance, such that a virtual agent can suitably respond to these intent/entities.’);
“based on the inquiry intent information, retrieve and arrange information associated with the inquiry” (M: par. 0080; ‘The intent/entity model 108 stores associations or relationships between particular intents and particular sample utterances.’);, and
“based on a retrieval result obtained from the retrieving and an arrangement result obtained from the arranging, extract the answer corresponding to the inquiry” (M: par. 0082; ‘For the embodiment illustrated in FIG. 4A, the conversation model 110 stores associations between intents of the intent/entity model 108 and particular responses and/or actions, which generally define the behavior of the RA/BE 102.’).
Regarding claim 4 (dep. on claim 1), the combination of Madhusudhan in view of Lee further teaches:
“wherein the named entity vector is obtained as named entity information extracted from the user's utterance obtained through a conversion to text that is encoded” (M: par. 0053; ‘As used herein an “intent-entity model” refers to a model that associates particular intents with particular entities and particular sample utterances, wherein entities associated with the intent may be encoded as a parameter of the intent within the sample utterances of the model.’).
Regarding claim 5 (dep. on claim 1), the combination of Madhusudhan in view of Lee further teaches:
“wherein the sentence vector is obtained as sentence information extracted from the user's utterance obtained through a conversion to text is encoded” (M: par. 0053; ‘As used herein an “intent-entity model” refers to a model that associates particular intents with particular entities and particular sample utterances, wherein entities associated with the intent may be encoded as a parameter of the intent within the sample utterances of the model.’).
Regarding claims 6 (dep. on claim 1) and 20 (dep. on claim 16), the combination of Madhusudhan in view of Lee further teaches:
“wherein the synthetic vector is a single vector into which the named entity vector and the sentence vector are merged” (M: par. 0156; ‘The multi-vector aggregation algorithms 452 include one or more algorithms for deriving a single word vector from a collection of word vectors.’).
Regarding claim 7 (dep. on claim 1), the combination of Madhusudhan in view of Lee further teaches:
“based on a result of providing the response, update a database (DB) comprising named entity-related information, and based on the updated DB, perform voice recognition” (M: par. 0058; ‘For example, components of the NLU framework (e.g., the structure subsystem or the vocabulary subsystem of the meaning extraction subsystem) may be continuously updated based on new utterances, such as exchanges between users and a virtual agent, to enhance the adaptability of the NLU framework to changes in the use of certain terms and phrases over time.’).
Regarding claim 19 (dep. on claim 16), the combination of Madhusudhan in view of Lee further teaches:
“wherein the named entity vector is obtained as named entity information extracted from the user's utterance obtained through a conversion to text is encoded, and wherein the sentence vector is obtained as sentence information extracted from the user's utterance obtained through the conversion to text is encoded” (see claims 4 and 5).
Claim(s) 8-15 are rejected under 35 U.S.C. 103 as being unpatentable over Madhusudhan in view of Mueller et al. (US 20200151583 A1), further in view of Lee.
Regarding claim 8, Madhusudhan teaches:
“An intelligent server comprising: a processor; and a memory electrically connected to the processor and storing instructions which, when executed by the processor” (par. 0078; ‘Client instance 42 is supported by virtual servers similar to those explained with respect to FIG. 2, and is illustrated here to show support for the disclosed functionality described herein within the client instance 42.’), cause the processor to:
“obtain first intent information corresponding to a user's utterance” (par. 0062; ‘In this manner, present embodiments extract intents/entities from the user utterance, such that a virtual agent can suitably respond to these intent/entities.’);
“[[in response to a reliability of the first intent information being less than a threshold value,]] obtain a named entity vector and a sentence vector based on the user's utterance” (par. 0163; ‘As such, in block 500, a new word vector would be generated for the term “Everest” that would be relatively close in one or more dimensions to word vectors for terms such as “conference”, “meeting”, “presentation”, and so forth.’; par. 0166; ‘Returning to the “Everest” example, if the utterance 122 is “I'm not sure, but we have a meeting scheduled in Everest at 2:30 pm this afternoon to discuss what to do”, the other words 522 in the utterance 122 may be used to determine that the use of the word “Everest” in the utterance 122 was referring to the conference room.’),
“obtain a synthetic vector [[by merging the named entity vector and the sentence vector in the order of the named entity vector and the sentence vector]]” (par. 0156; ‘The multi-vector aggregation algorithms 452 include one or more algorithms for deriving a single word vector from a collection of word vectors.’), and
“based on the synthetic vector, provide a response corresponding to the user's utterance” (par. 0085; ‘For the illustrated embodiment, the RA/BE 102 also provides a response 124 (e.g., a virtual agent utterance or confirmation) to the client device 14D via the network 18, for example, indicating actions performed by the RA/BE 102 in response to the received user utterance 122.’).
However, Madhusudhan does not expressly teach:
“in response to a reliability of the first intent information being less than a threshold value, obtain a named entity vector and a sentence vector based on the user's utterance” (emphasis added); and
“obtain a synthetic vector by merging the named entity vector and the sentence vector in the order of the named entity vector and the sentence vector” (emphasis added).
Mueller teaches:
“in response to a reliability of the first intent information being less than a threshold value, obtain a named entity vector and a sentence vector based on the user's utterance” (par. 0039; ‘For example, the vague utterance “I like new coffeemaker plz!” may be recognized as the intent “Purchase espresso machine” with a confidence level below a specified threshold.’; par. 0040; ‘Fact verification may consist of retrieving entity values that are stored in various components such as databases, and comparing the stored entity values against stated entity values mentioned by the customer. For example, if the customer states “I've paid $600 for this [espresso machine name] and now I see that it's available online from [vendor name] for only $400!”, verification can consist of checking the customer's transaction history to confirm that they did purchase the mentioned espresso machine for $600.’).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the named entity vector and sentence vector taught by Madhusudhan by incorporating Mueller’s fact verification in order to obtain the vectors based on a low confidence level for an intent. The combination would provide a verify a customer’s concern. (Mueller: par. 0040)
However, Madhusudhan and Mueller does not expressly teach:
“obtain a synthetic vector by merging the named entity vector and the sentence vector in the order of the named entity vector and the sentence vector.”
In a similar field of endeavor (utterance processing), Lee teaches:
“obtain a synthetic vector by merging the named entity vector and the sentence vector in the order of the named entity vector and the sentence vector” (par. 0039; ‘The block 412 may receive features from the entity and/or entity tracking information, an utterance embedding based on a user utterance from the sentence encoder 208, one or more of the biasing vectors e.sup.s or e.sup.u from the NLR inferencer 220, a previously fully-formed action in a sentence embedding from the sentence encoder 248, and content based on an API call. The received embeddings in vector form, may then be combined into a feature vector for input into the context RNN 224, such that an output may be generated.’).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the entity vector of Madhusudhan with an utterance vector, as taught by Lee, into a feature vector for input into a context RNN, such that an output may be generated. (Lee: par. 0039)
Regarding claim 9 (dep. on claim 8), the combination of Madhusudhan in view of Mueller and Lee further teaches:
“based on the synthetic vector, obtain second intent information corresponding to the user's utterance” (Madhusudhan: par. 0060; ‘As mentioned, a computing platform may include a chat agent, or another similar virtual agent, that is designed to automatically respond to user requests to perform functions or address issues on the platform.’; par. 0083; ‘As such, a similarity measure or distance between two different utterances can be calculated using the respective intent vectors produced by the NLU engine 116 for the two intents, wherein the similarity measure provides an indication of similarity in meaning between the two intents.’),,
“in response to the user's utterance being an inquiry, generate an answer corresponding to the inquiry based on the synthetic vector” (Madhusudhan: par. 0085; ‘For the illustrated embodiment, the RA/BE 102 also provides a response 124 (e.g., a virtual agent utterance or confirmation) to the client device 14D via the network 18, for example, indicating actions performed by the RA/BE 102 in response to the received user utterance 122.’), and
“based on the second intent information or the answer, provide the response” (Madhusudhan: par. 0085; ‘For the illustrated embodiment, the RA/BE 102 also provides a response 124 (e.g., a virtual agent utterance or confirmation) to the client device 14D via the network 18, for example, indicating actions performed by the RA/BE 102 in response to the received user utterance 122.’).
Regarding claim 10 (dep. on claim 9), the combination of Madhusudhan in view of Mueller and Lee further teaches:
“based on the user's utterance, obtain inquiry intent information” (see claim 3);
“based on the inquiry intent information, retrieve and arrange information associated with the inquiry” (see claim 3); and
“based on a retrieval result obtained from the retrieving and an arrangement result obtained from the arranging, extract the answer corresponding to the inquiry” (see claim 3).
Regarding claim 11 (dep. on claim 8), the combination of Madhusudhan in view of Mueller and Lee further teaches:
“wherein the named entity vector is obtained as named entity information extracted from the user's utterance obtained through a conversion to text is encoded” (see claim 4).
Regarding claim 12 (dep. on claim 8), the combination of Madhusudhan in view of Mueller and Lee further teaches:
“wherein the sentence vector is obtained as sentence information extracted from the user's utterance obtained through a conversion to text that is encoded” (see claim 5).
Regarding claim 13 (dep. on claim 8), the combination of Madhusudhan in view of Mueller and Lee further teaches:
“wherein the synthetic vector is a single vector into which the named entity vector and the sentence vector are merged” (see claim 6).
Regarding claim 14 (dep. on claim 8), the combination of Madhusudhan in view of Mueller and Lee further teaches:
“based on a result of providing the response, update a database (DB) comprising named entity-related information, and based on the updated DB, perform voice recognition” (see claim 7).
Regarding claim 15 (dep. on claim 8), the combination of Madhusudhan in view of Mueller and Lee further teaches:
“wherein the reliability is calculated based on an equation set based on at least one of a confidence value or an uncertainty value” (Mueller: par. 0039; ‘For example, the vague utterance “I like new coffeemaker plz!” may be recognized as the intent “Purchase espresso machine” with a confidence level below a specified threshold.’);
“wherein the confidence value comprises a numerical representation of a degree of confidence with which an intent information acquisition circuitry obtains the first intent information when obtaining the first intent information” (Mueller: par. 0039; ‘For example, the vague utterance “I like new coffeemaker plz!” may be recognized as the intent “Purchase espresso machine” with a confidence level below a specified threshold.’); and
“wherein the uncertainty value comprises a value obtained based on a Bayesian model” (well-known in the art, as evident by Short et al. (US 20150287422 A1) (par: 0302; ‘This ambiguity measure may be used in a Kalman filter, a Bayesian decision process, a scoring function or a similar process whereby the certainty/ambiguity measure is used to determine which tracklets or coherent groups should be extracted or enhanced.’)).
Claim(s) 21 is rejected under 35 U.S.C. 103 as being unpatentable over Madhusuhan in view of Lee, further in view of Wang et al. (US 20210082585 A1)
Regarding claim 21 (dep. on claim 1), Madhusudhan teaches parallel processing of vector translator module (par. 0271; ‘For embodiments in which multiple portions of the NLU-processed utterance 1020 are provided as input to the DAVE system 1004, the vector translator module 1022 may perform the repeat the illustrated process 1061 for each portion of the of the NLU-processed utterance 1020 (e.g., in series, or in parallel for reduced latency).’).
However, Madhusudhan and Lee do not expressly teach:
“wherein the named entity vector and the sentence vector are obtained in parallel.”
Wang teaches:
“wherein the named entity vector and the sentence vector are obtained in parallel” (par. 0034; ‘Continuing with reference to FIG. 3, each generated word vector is processed using sentence embedding 321 to yield a sentence vector 325, and, in parallel, each identified medical entity 330 is processed using word embedding 331 to yield an entity vector 335.’).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Madhusudhan’s (in view of Lee) utterance embeddings (par. 0254) and entity vector (par. 0163) by incorporating Wang’s sentence embedding and word embedding in order to obtain the vectors in parallel. The combination would allow a deep learning model to learn abstract features with context from the input features. (Wang: par. 0034)
Conclusion
Other pertinent prior art are cited in the PTO-892 for the applicant's consideration.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MARK VILLENA whose telephone number is (571)270-3191. The examiner can normally be reached 10 am - 6pm EST Monday through Friday.
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, Richemond Dorvil can be reached at (571) 272-7602. 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.
MARK . VILLENA
Examiner
Art Unit 2658
/MARK VILLENA/Examiner, Art Unit 2658