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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 2 February 2026 has been entered.
Claim Rejections - 35 USC § 103
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Claim(s) 1, 4-6, 8-10, 12-14, 17-18, and 21-28 is/are rejected under 35 U.S.C. 103 as being unpatentable over Vukich et al (From IDS: US Pub. No. 11,082,387, published Aug. 2021, hereinafter “Vukich”) in view of Gratton et al (US Pub. No. 2021/0081559, published March 2021, hereinafter “Gratton”), further in view of Ma et al. (NPL: Modeling Task Relationships in Multi-task Learning with Multi-gate Mixture-of-Experts, published Aug. 2018, hereinafter “Ma”) and Nguyen et al. (US Pub. No. 2020/0159916, published May 2020, hereinafter “Nguyen”).
Regarding claim 1, Vukich teaches a computer-implemented method comprising:
obtaining event-related communication data generated in connection with one or more systems associated with at least one enterprise (Vukich, Col. 9 Lines 9-17 – “the parse engine 221 may transform the data, such as, e.g., message subject, message recipients, message sender, listed attachments, related calendar events, related tasks, etc., as well as the user interaction 224 including setting a follow-up flag, a message category flag, or other user interaction 224, into, e.g., feature vectors or feature maps such that the message prioritization machine learning engine 222 may generate message priority predictions based on features of the data” – teaches obtaining event-related communication data generated in connection with one or more systems associated with at least one enterprise, as supported in Col. 20 Lines 8-13 – “the exemplary network 505 may also include, for instance, at least one of a local area network (LAN), a wide area network (WAN), the Internet, a virtual LAN (VLAN), an enterprise LAN, a layer 3 virtual private network (VPN), an enterprise IP network, or any combination thereof.” – teaches enterprise network access for obtaining communication data);
comparing identifying information pertaining to one or more event notifications within the event-related communication data to identifying information pertaining to multiple historical event notifications stored in at least one database (Vukich, Col. 12 Lines 10-16 – “the email database 205 may utilize the personnel data 212, subject attributes of the message data 217, or linked message data identified in the message data 217, or a combination thereof, to determine related messages in a message history and associated historical message data and user interaction data for each message in the message history of the email database 205.” – teaches comparing identifying information pertaining to one or more event notifications to identifying data pertaining to multiple historical event notifications stored in at least one database (determines related messages in a message history and associated historical message data));
Vukich fails to explicitly teach predicting, for the one or more event notifications upon determining that the identifying information pertaining to the one or more event notifications differs from the identifying information pertaining to the multiple historical event notifications, at least one communication channel and at least one communication format by processing at least a portion of the obtained event-related communication data using one or more machine learning techniques; wherein the multiple types of outputs comprise a first output identifying the at least one communication channel to be used for at least one of the one or more event notifications, and a second output identifying the at least one communication format to be used for at least one of the one or more event notifications; and performing one or more automated actions based at least in part on the at least one predicted communication channel and the at least one predicted communication format, wherein performing one or more automated actions comprises: generating, using at least one automated communication system, the at least one event notification in the at least one predicted communication format; and outputting, using the at least one automated communication system, the at least one generated event notification to one or more external devices via the at least one predicted communication channel; wherein the method is performed by at least one processing device comprising a processor coupled to a memory.
However, analogous to the field of the claimed invention, Gratton teaches:
predicting, for the one or more event notifications upon determining that the identifying information pertaining to the one or more event notifications differs from the identifying information pertaining to the multiple historical event notifications, at least one communication channel and at least one communication format by processing at least a portion of the obtained event-related communication data using one or more machine learning techniques (Gratton, [0344] – “An event notification module refers to notification preferences, to determine how to notify an entity (either about events relevant to them or events relevant to another entity). An entity can be notified via email, via text message, through other messaging infrastructures, by storing an event in durable storage, etc. An event can be formatted for compatibility with entity systems. For example, an event can be stored in a data format requested by an entity.” and in [0370] – “Event identification module 1118 can access preferences 1426. Event identification module 1118 can (possibly using artificial intelligence and/or machine learning) determine that event 1424 is occurring in area 1444 (and otherwise satisfies preferences 1426). Event notification module 116 can send notification 1471 to entity 1421 to notify entity 1421 of event 1424.” – teaches predicting at least one communication channel (e.g., email, text) and at least one communication format by processing at least a portion of the obtained event-related communication data using one or more machine learning techniques (event notification module can use artificial intelligence and/or machine learning). Gratton further teaches in [0376] – “The impact prediction module can maintain an event history database of prior events and corresponding impacts. As new events are detected, the impact prediction module can refer to the event history database and compare the new events to prior events.” – teaches identifying information pertaining to one or more event notifications that differ from the information pertaining to historical event notifications (as new events are detected, impact prediction model can refer to event history database and compare new events to prior events, thus identifying information pertaining to one or more events that differ from the information pertaining to historical events));
wherein the multiple types of outputs comprise a first output identifying the at least one communication channel to be used for at least one of the one or more event notifications, and a second output identifying the at least one communication format to be used for at least one of the one or more event notifications (Gratton, [0820] – “Ingestion modules 101 can send normalized signals 4222 to event detection infrastructure 103. Event detection infrastructure 103 and event notification 116 can adhere to data privacy settings of the multiple organizations. Event detection infrastructure 103 can detect event 4235 from normalized signals 4222. Event detection infrastructure 103 can send event 4235 to event notification module 116. Event notification module 116 can notify one or more entities about event 4235”, and in [0344] – “An event notification module refers to notification preferences, to determine how to notify an entity (either about events relevant to them or events relevant to another entity). An entity can be notified via email, via text message, through other messaging infrastructures, by storing an event in durable storage, etc. An event can be formatted for compatibility with entity systems. For example, an event can be stored in a data format requested by an entity.” – teaches an event notification module that can notify one or more entities with multiple outputs, wherein the outputs comprise a communication channel (e.g., email, text) and a communication format (event can be formatted for compatibility)), and
performing one or more automated actions based at least in part on the at least one predicted communication channel and the at least one predicted communication format (Gratton, [0370] – “Event identification module 1118 can (possibly using artificial intelligence and/or machine learning) determine that event 1424 is occurring in area 1444 (and otherwise satisfies preferences 1426). Event notification module 116 can send notification 1471 to entity 1421 to notify entity 1421 of event 1424.” – teaches performing one or more automated actions (notifying an entity) based in part on the at least one predicted communication channel and format (based on entity preferences)), wherein performing one or more automated actions comprises: generating, using at least one automated communication system, the at least one event notification in the at least one predicted communication format; and outputting, using the at least one automated communication system, the at least one generated event notification to one or more external devices via the at least one predicted communication channel (Gratton, [0820] – “Ingestion modules 101 can send normalized signals 4222 to event detection infrastructure 103. Event detection infrastructure 103 and event notification 116 can adhere to data privacy settings of the multiple organizations. Event detection infrastructure 103 can detect event 4235 from normalized signals 4222. Event detection infrastructure 103 can send event 4235 to event notification module 116. Event notification module 116 can notify one or more entities about event 4235”, and in [0344] – “An event notification module refers to notification preferences, to determine how to notify an entity (either about events relevant to them or events relevant to another entity). An entity can be notified via email, via text message, through other messaging infrastructures, by storing an event in durable storage, etc. An event can be formatted for compatibility with entity systems. For example, an event can be stored in a data format requested by an entity.” – teaches generating, using at least one automated communication system (event notification module 116), the at least one event notification in the at least one predicted communication format (event notification module determines how to notify entity including formatting events for compatibility with entity systems, thus generating at least one event notification in the at least one predicted communication format) and teaches outputting, using the at least one communication system (event notification module 116), the least one generated event notification to one or more external devices (event notification module 116 can notify one or more entities about event) via the at least one predicted communication channel (entity can be notified by event notification module 116 via email, text, other messaging infrastructures));
wherein the method is performed by at least one processing device comprising a processor coupled to a memory (Gratton, [0109] – “Implementations can comprise or utilize a special purpose or general-purpose computer including computer hardware, such as, for example, one or more computer and/or hardware processors (including any of Central Processing Units (CPUs), and/or Graphical Processing Units (GPUs), general-purpose GPUs (GPGPUs), Field Programmable Gate Arrays (FPGAs), application specific integrated circuits (ASICs), Tensor Processing Units (TPUs)) and system memory” – teaches utilizing central processing units and system memory to perform the methods).
Therefore, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to incorporate the prediction of communication channel and format, multiple outputs, and automated actions of Gratton to the data gathering and data comparison of Vukich in order to predict communication channels and format for automated actions. Doing so would enable systems to notify users to be aware of relevant events as close as possible to the event’s occurrence, determine relevancy of the events, and establish how users would like to receive event notifications (Gratton, [0096]).
The combination of Vukich and Gratton fails to explicitly teach wherein processing at least a portion of the obtained event-related communication data using one or more machine learning techniques comprises processing at least a portion of the obtained event-related communication data using at least one deep neural network comprising at least one input layer, multiple hidden layers, and multiple output layers, wherein the multiple hidden layers and the multiple output layers are arranged in multiple parallel branches, with each of the multiple parallel branches corresponding to one of multiple types of outputs.
However, analogous to the field of the claimed invention, Ma teaches:
wherein processing at least a portion of the obtained event-related communication data using one or more machine learning techniques comprises processing at least a portion of the obtained event-related communication data using at least one deep neural network comprising at least one input layer, multiple hidden layers, and multiple output layers (Ma, Fig. 1 and Pg. 2, Left Column, Paragraph 2 – “Instead of having one bottom network shared by all tasks, our model, shown in Figure 1(c), has a group of bottom networks, each of which is called an expert. In our paper, each expert is a feed-forward network.” – teaches wherein processing at least a portion of the obtained data using one or more machine learning techniques comprise processing at least a portion of the obtained data using at least one deep neural network comprising at least one input layer, multiple hidden layers, and multiple output layers (Fig. 1 shows machine learning techniques comprising processing obtained data using at least one deep neural network comprising at least one input layer, as in Fig. 1 “Input”, multiple hidden layers, as in Fig. 1 the hidden layers are experts which are feed-forward networks, and multiple output layers, as in Fig. 1 “Output A” and “Output B”)),
wherein the multiple hidden layers and the multiple output layers are arranged in multiple parallel branches, with each of the multiple parallel branches corresponding to one of multiple types of outputs (Ma, Fig. 1 – teaches wherein the multiple hidden layers and the multiple outputs layers are arranged in multiple parallel branches, with each of the multiple parallel branches corresponding to one of multiple types of outputs (Fig. 1 shows the hidden layers and output layers arranged in multiple parallel branches, with each of the branches corresponding to one of multiple outputs, the outputs received from different tasks)),
Therefore, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to incorporate the hidden layers and output layers arranged in parallel branches of Ma to the event-related communication data, predicted communication channels, and predicted communication formats of Vukich and Gratton. Doing so would enable machine learning models to learn task-specific functionalities and generate outputs for different tasks using one input layer (Ma, Introduction).
The combination of Vukich, Gratton, and Ma fails to explicitly teach wherein at least one of the multiple output layers corresponding to the first output comprises a first number of neurons determined as a function of at least a designated number of possible communication channels, and wherein at least another of the multiple output layers corresponding to the second output comprises a second number of neurons determined as a function of at least a designated number of communication formats.
However, analogous to the field of the claimed invention, Nguyen teaches:
wherein at least one of the multiple output layers corresponding to the first output comprises a first number of neurons determined as a function of at least a designated number of possible communication channels, and wherein at least another of the multiple output layers corresponding to the second output comprises a second number of neurons determined as a function of at least a designated number of possible communication formats (Nguyen, [0100] – “The order in which the operations are described in each example flow diagram or technique is not intended to be construed as a limitation, and any number of the described operations can be combined in any order or in parallel to implement each technique.”, [0195] – “The classification section 1008 can be configured to determine the indication 830 (and likewise indication 328) as a probability based at least in part on the at least two filter outputs 818, 824, 908, e.g., based on filter values 1016, 1020. For example, the classification section 1008 can include a fully-connected layer 1024 whose neurons' outputs represent the probabilities of various classifications, e.g., scaled using softmax. This is not limiting; other connectivity patterns of an output layer can be used. Shown are two neurons 1026, 1028 having respective outputs 1030, 1032, although any number ≥1 of neurons can be used.” and in [0197] – “The outputs from the scaling operation 1034 are, e.g., values 1036, 1038 representing probabilities that the event 808 falls into respective categories, e.g., associated with a security violation or not associated with a security violation. In some examples, the number of neuron outputs is unity, representing one of those options, or more than two, e.g., representing “not associated,” “associated with a violation of type 1,” “associated with a violation of type 2,” etc. ” – teaches wherein at least one of the multiple output layers corresponding to the first output comprises a first number of neurons determined as a function of at least a designated number of possible communication channels, and wherein at least another of the multiple output layers corresponding to a second output comprises a second number of neurons determined as a function of at least a designated number of communication formats (classification includes a fully-connected layer with multiple outputs, wherein each output corresponds to a number of neurons. Any number of neurons may be used, and the number of neurons correspond to respective outputs. Neuron outputs represent probability of various classifications scaled using softmax, thus the number of neurons are determined as a function of at least a designated number of possible classifications));
Therefore, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to incorporate the neurons corresponding to first and second outputs and neurons determined as a function of a designated number of possible outputs of Nguyen to the parallel branches, multiple output layers, possible communication channels, and possible communication formats of Vukich, Gratton, and Ma. Doing so would enable models to classify events occurring at a monitored computing device (Nguyen, [0018]) and provide a number of neurons corresponding to respective outputs (Nguyen, [0195]).
Claims 14 and 18 incorporate substantively all the limitations of claim 1 in a non-transitory processor-readable storage medium and an apparatus, and are rejected on similar grounds as above.
Regarding claim 4, the combination of Vukich, Gratton, Ma, and Nguyen teaches the computer-implemented method of claim 1, wherein the multiple hidden layers comprise at least one activation function, and wherein the at least one activation function of the multiple hidden layers comprises at least one rectified linear unit activation function (Ma, Pg. 5, Left Column, Paragraph 2 – “Our implementation consists of identical multilayer perceptrons with ReLU activations.” – teaches wherein at least one hidden layer comprises at least one activation function, and wherein the at least one activation function of the at least one hidden layer comprises at least one rectified linear unit activation function (hidden layers of multilayer perceptrons utilize ReLU activations, thus at least one hidden layer comprises at least one rectified linear unit activation function)).
Therefore, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to incorporate the hidden layers comprising at least one rectified linear unit activation function of Ma to further modify the hidden layers of Vukich, Gratton, Ma, and Nguyen in order to utilize a rectified linear unit activation function in the hidden layer. Doing so would enable machine learning models to learn task-specific outputs for different tasks (Ma, Introduction).
Claims 21 and 25 are similar to claim 4, hence similarly rejected.
Regarding claim 5, the combination of Vukich, Gratton, Ma, and Nguyen teaches the computer-implemented method of claim 1, wherein the multiple output layers comprise at least one activation function, and wherein the at least one activation function of the multiple output layers comprises at least one softmax activation function (Nguyen, [0152] – “The NN can include a softmax, threshold, clipping, clamping, or other operation to output(s) from neuron outputs of the output layer” – teaches wherein the multiple output layers comprise at least one activation function, and wherein the at least one activation function of the multiple output layers comprises at least one softmax activation function (NN can include a softmax to output(s) from neuron outputs of the output layer, thus at least one activation function of the multiple output layers comprises at least one softmax activation function)).
Therefore, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to incorporate the output layers comprising at least one softmax activation function of Nguyen to further modify the multiple output layers of Vukich, Gratton, Ma, and Nguyen in order to utilize a softmax activation function at the output layer. Doing so would enable the output layers of the model to output a representation of categories to which the event associated with an event record belongs (Nguyen, [0042]).
Claims 22 and 26 are similar to claim 5, hence similarly rejected.
Regarding claim 6, the combination of Vukich, Gratton, Ma, and Nguyen teaches the computer-implemented method of claim 1, wherein processing at least a portion of the obtained event-related communication data using one or more machine learning techniques comprises processing a set of input data from the obtained event-related communication data, wherein the set of input data comprises two or more of event source-related data, event type-related data, event status-related data, destination-related data, and language- related data (Gratton, [0896] – “Through natural language processing and/or image analysis, geo determination module 104 can infer an event location from the text and image.” – teaches inferring location (destination-related data) based on text (language-related data), and in [0598] – “The listing of events may be automatically arranged in panel 2504 according to one or more criteria such as timestamp, severity, truthfulness, status, event type, location, or other characteristic(s).” – teaches processing input data comprising event-status related data, event-type related data, and event-source related data).
Therefore, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to incorporate the input data comprising various event-related data of Gratton to further modify the method of Vukich, Gratton, Ma, and Nguyen in order to process input comprising event-related data using one or more machine learning techniques. Doing so would allow the event-related data to be provided to event detection systems for event detection and be identified based on rules defining events of interest (Gratton, [0716]).
Claims 23 and 27 are similar to claim 6, hence similarly rejected.
Regarding claim 8, the combination of Vukich, Gratton, Ma, and Nguyen teaches the computer-implemented method of claim 1, further comprising: generating cryptographic information attributed to at least a portion of the event-related communication data by processing the at least a portion of the event-related communication data using at least one cryptographic function, wherein the identifying information pertaining to one or more event notifications within the event-related communication data comprises at least a portion of the generated cryptographic information (Gratton, [0122] – “In this description and the following claims, “geohash” is defined as a geocoding system which encodes a geographic location into a short string of letters and digits. Geohash is a hierarchical spatial data structure which subdivides space into buckets of grid shape (e.g., a square). Geohashes offer properties like arbitrary precision and the possibility of gradually removing characters from the end of the code to reduce its size (and gradually lose precision). As a consequence of the gradual precision degradation, nearby places will often (but not always) present similar prefixes. The longer a shared prefix is, the closer the two places are. geo cells can be used as a unique identifier and to approximate point data (e.g., in databases)” – teaches generating cryptographic information attributed to a portion of the event-related communication data, in this case the location/destination, wherein the identifying information pertaining to one or more event notifications comprises at least a portion of the generated cryptographic information).
Therefore, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to incorporate the cryptographic information attributed to the event-related communication data through use of a cryptographic function of Gratton to further modify the method of Vukich, Gratton, Ma, and Nguyen, in order to apply a cryptographic function to at least a portion of the event-related communication data. Doing so would offer properties like arbitrary precision (Gratton, [0122]) and enable systems to determine event truthfulness and severity (Gratton, [0196]).
Claim 17 is similar to claim 8, hence similarly rejected.
Regarding claim 9, the combination of Vukich, Gratton, Ma, and Nguyen teaches the computer-implemented method of claim 8, wherein the at least one cryptographic function comprises at least one secure hash algorithm (Gratton, [0122] – “In this description and the following claims, “geohash” is defined as a geocoding system which encodes a geographic location into a short string of letters and digits. Geohash is a hierarchical spatial data structure which subdivides space into buckets of grid shape (e.g., a square). Geohashes offer properties like arbitrary precision and the possibility of gradually removing characters from the end of the code to reduce its size (and gradually lose precision). As a consequence of the gradual precision degradation, nearby places will often (but not always) present similar prefixes. The longer a shared prefix is, the closer the two places are. Geo cells can be used as a unique identifier and to approximate point data (e.g., in databases)” – teaches wherein at least one cryptographic function comprises a secure hash algorithm).
Regarding claim 10, the combination of Vukich, Gratton, Ma, and Nguyen teaches the computer-implemented method of claim 1, wherein comparing identifying information pertaining to one or more event notifications within the event-related communication data to identifying information pertaining to multiple historical event notifications stored in at least one database comprises comparing at least one hash attributed to the one or more event notifications within the event-related communication data to at least one hash attributed to each of the multiple historical event notifications stored in the at least one database (Vukich, Col. 30 Lines 1-11 – “the exemplary inventive computer-based components of the present disclosure may be configured to securely store or transmit data by utilizing one or more of encryption techniques (e.g., private/public key pair, Triple Data Encryption Standard (3DES), block cipher algorithms (e.g., IDEA, RC2, RCS, CAST and Skipjack), cryptographic hash algorithms (e.g., MD5, RIPEMD-160, RTRO, SHA-1, SHA-2, Tiger (TTH), WHIRLPOOL, RNGs).” – teaches stored event notifications comprising a hash attributed to the one or more event notifications and each of the multiple historical event notifications in at least one database, and in Col. 12, Lines 50-53 – “by comparing the subject or body text of the message data 217 with subjects and names of tasks in the task database 206, the task database 206 or the parse engine 221 may identify similar or related work tasks.” – teaches comparing the subject of the event notification with the subjects and names in the database storing historical event notifications, thus comparing hashes attributed to the event notification data to the hashes attributed to the historical event notification data).
Claims 24 and 28 are similar to claim 10, hence similarly rejected.
Regarding claim 12, the combination of Vukich, Gratton, Ma, and Nguyen teach the computer-implemented method of claim 1, wherein performing one or more automated actions comprises automatically training the one or more machine learning techniques using feedback generated in connection with one or more of the at least one predicted communication channel and the at least one predicted communication format (Vukich, Col. 10 Lines 52-61 – “ If the user interacts with the first message before the second message, the user acted as though the first message had higher priority, and thus, the optimizer 223 may train the message prioritization machine learning engine 222 according to the error resulting from the priority prediction of each of the first and second messages. In some embodiments, the optimizer 223 may backpropagate the error to the parse engine 221, the message prioritization machine learning engine 222, or both to train each engine in an on-line fashion.” – teaches automatically training the one or more machine learning techniques using feedback generated in connection with one or more of the at least one predicted communication channel and format).
Regarding claim 13, the combination of Vukich, Gratton, Ma, and Nguyen teach the computer-implemented method of claim 1, further comprising: automatically training the one or more machine learning techniques using historical event notification data and corresponding context-related information (Vukich, Col. 17 Lines 50-54 – “the email database 205 may be continually updated with information concerning the priority and user interactions relative to each message associated with the user to better train the priority model 225 according to historical messages.” – teaches training the one or more machine learning techniques using historical data and corresponding context-related information, and in Col. 17 Lines 41-49 – “the user interaction 252 or the corresponding message and relative priority of the message may be provided to the emailing database 205 to update the message history and message data. In some embodiment, each user interaction 252 may be added to the message history of a given message affected by the user interaction 252, e.g., by affected the priority of the message or through a response or other action taken relative to the message. “ – teaches training the one or more machine learning techniques using corresponding context related information (such as relative priority and user interaction)).
Response to Arguments
Applicant’s arguments, see pages 1-5 of Remarks, filed 31 December 2025, with respect to the rejection(s) of claim(s) 1, 4-6, 8-10, 12-14, 17-18, and 21-28 under 35 U.S.C. 103 have been fully considered and are persuasive. Therefore, the rejection has been withdrawn. However, upon further consideration, a new ground(s) of rejection is made over Vukich in view of Gratton, further in view of Ma et al. (NPL: Modeling Task Relationships in Multi-task Learning with Multi-gate Mixture-of-Experts, published Aug. 2018) and Nguyen et al. (US Pub. No. 2020/0159916, published May 2020). Ma teaches the amended limitations of claim 1 regarding “wherein processing at least a portion of the obtained event-related communication data comprises… at least one deep neural network comprising at least one input layer, multiple hidden layers, and multiple output layers, wherein the multiple hidden layers and the multiple output layers are arranged in multiple parallel branches, with each of the multiple parallel branches corresponding to one of multiple types of outputs”. Nguyen teaches the amended limitations of claim 1 regarding “wherein at least one of the multiple output layers corresponding to the first output comprises a first number of neurons determined as a function of at least a designated number of possible communication channels, and wherein at least another of the multiple output layers corresponding to the second output comprises a second number of neurons determined as a function of at least a designated number of possible communication formats”.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Kumar et al. (US Pub. No. 2023/0334249, filed April 2022) teaches methods for analyzing a plurality of natural language inputs associated with at least one user. Teaches hidden layers with a rectified linear unit activation function. Teaches output layers with a softmax activation function. Teaches wherein entities, or users, are of an enterprise, or business. Teaches bi-directional RNN for processing communication, or language data, in parallel.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to LOUIS C NYE whose telephone number is 571-272-0636. The examiner can normally be reached Monday - Friday 9:00AM - 5:00PM.
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, MATT ELL can be reached at 571-270-3264. 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.
/LOUIS CHRISTOPHER NYE/Examiner, Art Unit 2141
/MATTHEW ELL/Supervisory Patent Examiner, Art Unit 2141
Prosecution Insights