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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 06/05/2026 has been entered.
Examiner’s Note
According to the Remarks (filed on 3/18/2024), the term “micro-application” is defined in the specification and drawings (e.g., fig 27 and paragraph 170). In addition, the term does not denote a size of the application, but it is used for naming a building block as shown in figs 26-27. Thus, the rejections under 35 USC 112(b) have been withdrawn.
The Examiner encourages Applicant to schedule an interview (via Automated Interview Request (AIR)) to discuss issues related to, for example, the rejections noted below under 35 U.S.C § 101 and § 103, for moving toward allowance. (e.g., integrating dependent claims, elaborating claim languages with specificities, and elaborating claims toward a practical application)
Providing supporting paragraph(s) for each limitation of amended/new claim(s) in Remarks is strongly requested for clear and definite claim interpretations by Examiner (e.g., to avoid rejections under 35 U.S.C § 112(a) “Lack of written description”)
Applicant can schedule interviews (via Automated Interview Request (AIR)) at any stage of the prosecution (e.g., Non-Final, Final, and After-Final) to discuss any issues related to, for example, rejections under 35 U.S.C § 101 and § 102/103, for moving toward allowance.
If a limitation has bold brackets (i.e. [·]) around claim languages, the bracketed claim languages indicate that they have not been taught yet by the current prior art reference but they will be taught by another prior art reference afterwards.
If a limitation has one or more bold underlines, the one or more bold underlined claim languages indicate that they are taught by the current prior art reference, while the one or more non-underlined claim languages indicate that they have been taught already by one or more previous art references.
Priority
Acknowledgment is made of applicant's claim for the provisional application (62/719,849) filed on 08/20/2018.
Response to Arguments
Applicant's arguments filed on 06/05/2026 have been fully considered but they are not persuasive.
In Remarks, regarding 35 U.S.C § 101, Applicant contends:
“Therefore, the additional element, under its broadest reasonable interpretation, reads on obtaining at least one machine learning model through creation of a new model rather than obtaining at least one machine learning model through mere data gathering.”
“Applicant respectfully submits that especially "the at least one machine learning model including an input layer, a data processing layer, and an output layer," "receiving, at the input layer, video and image data from the sensor," and "selecting, at the processing layer, from a repository of computational models, a computational model for processing the video and image data based on a pattern recognized by the at least one machine learning model" is not a well known process according to MPEP 2106.05(d)(I)(2)”.
“Therefore, just as Ex parte Desjardins presented an improvement to the functioning of a computer with a machine learning system, wherein the improvement was set forth in the specification and reflected in the claims, the present invention similarly presents an improvement to the functioning of a computer by "[ minimizing] bias and variance," "[broadening] the applicability of the solution," "[providing] flexibility and adaptability," and minimizing the delay of the workflow. Accordingly, and assuming arguendo that a judicial exception exists, the claimed invention integrates the judicial exception into a practical application. Applicant thusly respectfully submits that the rejection under 35 U.S.C. 101 is overcome and respectfully requests it be withdrawn.”
Examiner’s response:
The examiner understands the applicant’s assertion.
However, it appears that each processing step is just applying the abstract idea to a general field of endeavor with additional elements. In addition, improvements to technology or technical field are not necessarily reflected in the claims. Thus, the claim does not integrate the judicial exception into a practical application, and the claim does not amount to significantly more than the judicial exception.
The examiner understands the applicant’s assertion “Therefore, the additional element, under its broadest reasonable interpretation, reads on obtaining at least one machine learning model through creation of a new model rather than obtaining at least one machine learning model through mere data gathering.”
However, as rejected under Claim Rejections - 35 USC § 101, it can still be interpreted as receiving data, and the data is just obtained based on a sensor. It doesn’t appear that the limitation is about obtaining at least one machine learning model through creation of a new model. Thus, it appears that the limitation is still an insignificant extra-solution activity.
The examiner understands the applicant’s assertion “Applicant respectfully submits that especially "the at least one machine learning model including an input layer, a data processing layer, and an output layer," "receiving, at the input layer, video and image data from the sensor," and "selecting, at the processing layer, from a repository of computational models, a computational model for processing the video and image data based on a pattern recognized by the at least one machine learning model" is not a well known process according to MPEP 2106.05(d)(I)(2)”.
However, as rejected under Claim Rejections - 35 USC § 101, the limitations can be interpreted as a combination of abstract ideas and additional elements since they are just about an ML model structure, receiving data and selecting a model. It appears that they are still is a well-known, well-understood, routine, or conventional process or activity.
The examiner understands the applicant’s assertion “Therefore, just as Ex parte Desjardins presented an improvement to the functioning of a computer with a machine learning system, wherein the improvement was set forth in the specification and reflected in the claims, the present invention similarly presents an improvement to the functioning of a computer by "[ minimizing] bias and variance," "[broadening] the applicability of the solution," "[providing] flexibility and adaptability," and minimizing the delay of the workflow. Accordingly, and assuming arguendo that a judicial exception exists, the claimed invention integrates the judicial exception into a practical application. Applicant thusly respectfully submits that the rejection under 35 U.S.C. 101 is overcome and respectfully requests it be withdrawn.”
However, par 91 states “From ensemble learning methods we know that a combination of lower accuracy models may perform better than a higher accuracy model due to overcoming bias” and par 127 says “Model Combination. For any given situation, Selector 452 may not be constrained to using a single model, but may activate a combination of models for ensemble learning, for example, to minimize bias and variance.” Based on the par 91, it appears that it is known that ensemble learning methods minimize bias, and based on the par 127, it appears that “minimizing bias and variance” is mentioned just an example of the ensemble learning, but not a key inventive concept of the whole invention. Assuming, arguendo, that they could be improvements of the invention, it is not clear how the limitations provide the asserted improvements. Basically, it appears that the claim limitations are not detailed enough to provide the asserted improvements.
In addition, par 128 says “Accordingly, cos θ may be used as a metric of congruence between different models. However, embodiments may also use less correlated models, which learn different things, to broaden the applicability of the solution.” Based on the par 128, it appears that the invention just may use another option of “less correlated models”, and the “less correlated models” may broaden the applicability of the solution. However, the specification does not provide what the “less correlated models” are and how the “less correlated models” broaden the applicability of the solution. In addition, it appears that the claims do not specify and elaborate the “less correlated models”. Thus, even assuming, arguendo, it is considered improvements, still it is not clear how the claims reflect the asserted improvements.
Furthermore, it appears that minimizing the delay of the workflow for the inventive system may be an improvement by determining the best layer for execution for each step of the solution plan and providing the requisite processing power for each step while providing flexibility and adaptability. However, the limitations do not provide how to determine a best layer for execution for each step of the solution plan. In addition, it is not clear why/how determining the best layer provides flexibility and adaptability in terms of the technologies of the inventive system.
The examiner understands the applicant’s assertion regarding Step 2B.
However, as rejected as rejected under Claim Rejections - 35 USC § 101, when read individually and/or in combination as a whole, it appears that the claim does not provide improvements for the whole invention. Thus, the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
The Remarks state that the claimed invention provides improvements for each technology mentioned above, but details of how the claimed invention provides improvements for those technologies are missing. Showing details of how the claimed invention provides improvements for a specific technology toward a practical application may help overcome the rejections under 35 USC § 101.
Currently, it does not appear that the limitations clearly show e.g., improvements in computer technology and improvements to other technical fields. Rather, it appears that the improvements in Remarks are about just improving the abstract ideas of the independent claims. It doesn’t seem that the specification and/or the independent claims clearly show how the inventive concept of the claims enables improvements and how they are tied together. The applicant may need to amend the claims to show how the claim languages and improvements are tied together.
To find a valid improvement to a technology, MPEP 2106.04(d)(1) says the specification must explain the improvement and that the claim must reflect the disclosed improvement. Furthermore, the improvement should not be merely a consequence of the abstract idea. See MPEP 2106.05(a). An improvement in the abstract idea itself is not an improvement to technology.
However, it appears that elaborating the last limitation toward a practical application may overcome the existing 101 rejections. The Examiner encourages Applicant to schedule an interview to discuss issues related to, for example, the rejections noted below under 35 U.S.C § 101.
For at least these reasons, Applicant's arguments are not convincing.
Applicant’s arguments regarding 35 USC § 102/103 with respect to the independent claims have been considered but are moot because the arguments are directed to amended limitation(s) that has/have not been previously examined.
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, 3-7, 9-13, 15-18 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Regarding claim 1
Claim 1 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: The claim recites a method; therefore, it falls into the statutory category of processes.
Step 2A Prong 1:
The limitations of
“A method implemented in …, the method comprising:
…;
generating, at …, a description of the problem, wherein the description conforms to defined format;
obtaining, at …, at least one … relevant to the problem;
selecting, …, from a repository of computational models, a computational model for processing the video and image data based on a pattern recognized by the at least one machine learning model;
selecting, at …, … upon which to execute the at least one … relevant to the problem, wherein …; and
executing, at …, the at least one … relevant to the problem using the selected … to generate at least one recommendation relevant to the problem.”, as drafted, are a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind. That is, nothing in the claim element precludes the step from practically being performed in the mind. For example, the limitations in the context of this claim encompass the user mentally thinking with a physical aid (e.g., pencil and paper).
If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: This judicial exception is not integrated into a practical application.
In particular, the claim recites an additional elements (“receiving, at the computer system, data relating to a problem to be solved, the data obtained via a sensor measuring environmental data”, “receiving, at the input layer, video and image data from the sensor:”) – the act of receiving data. The claim is adding an insignificant extra-solution activity to the judicial exception – see MPEP 2106.05(g). The act of receiving data is recited at a high-level of generality (i.e., as a generic act of receiving performing a generic act function of receiving data) such that it amounts no more than a mere act to apply the exception using a generic act of receiving. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
In particular, the claim recites additional elements (“a computer system comprising a processor, memory accessible by the processor, and computer program instructions stored in the memory and executable by the processor”, “the computer system”, “machine learning model”, “at the processing layer”, “computing infrastructure”, “the selected computing infrastructure comprises a mesh of interconnected micro-applications,”) – using a computing system and a machine learning model to process data. The computing system and the machine learning model in each step are recited at a high-level of generality (i.e., as a generic computer performing a generic computer function of processing data) such that it amounts no more than mere instructions to apply the exception using a generic computer component. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
In particular, the claim recites an additional element (“the at least one machine learning model including an input layer, a data processing layer, and an output layer”, “wherein each micro-application comprises analog input adapted to receive analog signals at edges of the mesh of micro-applications, digital input adapted to receive digital data from other micro-applications, event ingestion processing adapted to receive digital events and process and format the digital events for consumption, event consumption processing adapted to process the digital events to obtain information from the digital events and perform processing of the digital events, event generation processing adapted to generate another event to be output from micro-application to other micro-applications or out of the mesh of micro-applications, analog output adapted to output an analog signal representing the generated event, and digital output adapted to output a digital signal representing a digital event”). This is a recitation of a particular type of application to be used in performing the abstract idea. Limiting the abstract idea to a particular type of application is an attempt to limit the abstract idea to a particular field of use or technological environment, which does not integrate the abstract idea into a practical application. See MPEP 2106.05(h)
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
The claim is appending a well-understood, routine, conventional activity previously known to the industry, specified at a high level of generality, to the judicial exception - see MPEP 2106.05(d)(II) – “Receiving or transmitting data over a network, e.g., using the Internet to gather data” is Well-Understood, Routine, and Conventional Activity (MPEP 2106.05(d)). As discussed above with respect to integration of the abstract idea into a practical application, the additional element of the act of receiving/transmitting data amounts to no more than a mere act to apply the exception using a generic act of receiving/transmitting. A mere act to apply an exception using a generic act of receiving/transmitting cannot provide an inventive concept. The claim is not patent eligible.
As discussed above, with respect to integration of the abstract idea into a practical application, the additional elements of using a generic computer to perform each step amount to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. The claim is not patent eligible.
This is a recitation of a particular type of classifier to be used in performing the abstract idea. Limiting the abstract idea to a particular type of application is an attempt to limit the abstract idea to a particular field of use or technological environment, which does not amount to significantly more than the abstract idea. See MPEP 2106.05(h).
Similarly, Claims 7, 13 is/are rejected under 35 U.S.C. 101, mutatis mutandis, as reciting an abstract idea without adding significantly more than the judicial exception.
Regarding claim 3
Claim 3 is rejected under 35 U.S.C. 101 because it only modifies the abstract idea by data from different data sources, which also does not add significantly more or provide a specific application of the judicial exception.
Similarly, Claim(s) 9, 15 is/are rejected under 35 U.S.C. 101, mutatis mutandis, as reciting an abstract idea without adding significantly more than the judicial exception.
Regarding claim 4
Claim 4 is rejected under 35 U.S.C. 101 because it only modifies the abstract idea by selecting a model from existing models and generating a new model, which also does not add significantly more or provide a specific application of the judicial exception.
Similarly, Claims 10, 16 is/are rejected under 35 U.S.C. 101, mutatis mutandis, as reciting an abstract idea without adding significantly more than the judicial exception.
Regarding claim 5
Claim 5 is rejected under 35 U.S.C. 101 because it only modifies the abstract idea by determining and assembling a combination of models for a higher accuracy, which also does not add significantly more or provide a specific application of the judicial exception.
Similar to the combination of claims 4 and 5, Claims 11, 17 is/are rejected under 35 U.S.C. 101, mutatis mutandis, as reciting an abstract idea without adding significantly more than the judicial exception.
Regarding claim 6
Claim 6 is rejected under 35 U.S.C. 101 because it only modifies the abstract idea by determining a combination of models for a higher accuracy based on trained heuristics, which also does not add significantly more or provide a specific application of the judicial exception.
Similarly, Claims 12, 18 is/are rejected under 35 U.S.C. 101, mutatis mutandis, as reciting an abstract idea without adding significantly more than the judicial exception.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1, 3-5, 7, 9-11, 13, 15-17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Sun et al. (US 2017/0011308 A1) in view of Ross et al. (US 2018/0232663 A1) in view of TEIG et al. (US 20180025268 A1) further in view of WETMORE et al. (US 20140057232 A1)
Regarding claim 1
Sun teaches
A method implemented in a computer system comprising a processor, memory accessible by the processor, and computer program instructions stored in the memory and executable by the processor, the method comprising:
(Sun, [figs 1])
receiving, at the computer system, data relating to a problem to be solved, the data obtained [via] a sensor measuring environmental data;
(Sun, [figs 2-3] [pars 40-66] “Referring now to FIG. 3A, in connection with FIG. 2, a flow diagram depicts one embodiment of a method 300 for receiving a description of a problem and applying machine learning to automatically solve the problem. In brief overview, the method 300 includes receiving, by a first computing device, from a second computing device, via a user interface component, a description of a first problem (302). The method 300 includes assigning, by a clustering engine executing on the first computing device, the first problem to a class (304).” [par(s) 26] “A wide variety of I/O devices 130a-130n may be present in the computing device 100. Input devices include keyboards, mice, trackpads, trackballs, microphones, scanners, cameras, and drawing tablets. Output devices include video displays, speakers, inkjet printers, laser printers, and dye-sublimation printers. The I/O devices may be controlled by an I/O controller 123 as shown in FIG. 1B.”;)
generating, at the computer system, a description of the problem, wherein the description conforms to defined format;
(Sun, [figs 2-3] [pars 40-66] “the machine learning interface 204 may be able to receive input from, for example, the user interface component 202 and may format the input into a format that is understandable by (e.g., able to be processed by) one or more machine learning components, without an administrator of the system having to modify the user interface component 202 to be able to perform that formatting function.”; e.g., “machine learning interface 204 may be able to receive input from, for example, the user interface component 202 and may format the input into a format that is understandable by (e.g., able to be processed by) one or more machine learning components” may read on “description of the problem, wherein the description conforms to defined format”.)
obtaining, at the computer system, at least one machine learning model relevant to the problem;
(Sun, [figs 2-3] “Machine Learning Engine 208” [pars 40-66] “The machine learning engine 208 may include an ensemble of classifiers, for example. In an embodiment in which the machine learning engine 208 includes functionality for executing a clustering algorithm to identify patterns in data (e.g., the clustering engine 210), a variety of algorithms may be used, including, without limitation, K-Means and Expectation-Maximization algorithms. … the machine learning interface 204 may be configured to access one or more machine learning components of any type (e.g., the clustering engine 210 and the correlation engine 220) and receive some or all of an answer to one or more questions, without requiring users to interact with the underlying machine learning components. … the clustering engine 210 provides the description of the first problem and the assigned class to the correlation engine 220. In some embodiments, by determining a class of the first problem, the clustering engine 210 enables the correlation engine 220 to identify related databases, problems (e.g., problems of the same class), and problem resolutions. … In one embodiment, the correlation engine 220 accesses a database of solutions to previous problems of the same type to identify a solution”;)
selecting, at the processing layer, from a repository of computational models, a computational model for processing the [video and image] data based on a pattern recognized by the at least one machine learning model;
(Sun, [figs 2-3] “Machine Learning Engine 208” [pars 40-66] “The machine learning engine 208 may include an ensemble of classifiers, for example. In an embodiment in which the machine learning engine 208 includes functionality for executing a clustering algorithm to identify patterns in data (e.g., the clustering engine 210), a variety of algorithms may be used, including, without limitation, K-Means and Expectation-Maximization algorithms. … the machine learning interface 204 may be configured to access one or more machine learning components of any type (e.g., the clustering engine 210 and the correlation engine 220) and receive some or all of an answer to one or more questions, without requiring users to interact with the underlying machine learning components. … the clustering engine 210 provides the description of the first problem and the assigned class to the correlation engine 220. In some embodiments, by determining a class of the first problem, the clustering engine 210 enables the correlation engine 220 to identify related databases, problems (e.g., problems of the same class), and problem resolutions. … In one embodiment, the correlation engine 220 accesses a database of solutions to previous problems of the same type to identify a solution”;)
selecting, at the computer system, computing infrastructure upon which to execute the at least one machine learning model relevant to the problem, wherein the selected computing infrastructure comprises [a mesh of interconnected micro-applications, wherein each micro-application comprises analog input adapted to receive analog signals at edges of the mesh of micro-applications, digital input adapted to receive digital data from other micro-applications, event ingestion processing adapted to receive digital events and process and format the events for consumption, event consumption processing adapted to process the event to obtain information from the event and perform processing of the event, event generation processing adapted to generate another event to be output from micro-application to other micro-applications or out of the mesh of micro-applications, analog output adapted to output an analog signal representing the generated event, and digital output adapted to output a digital signal representing a digital event];
(Sun, [figs 2-3] “Machine Learning Engine 208” [pars 40-66] “In one embodiment, the clustering engine 210 uses machine learning to identify a class—or type—of problem being described in order to identify what resources (e.g., databases, machines, or people) to access in order to solve the problem, and any related, as-yet undescribed problems. In some embodiments, the clustering engine 210 performs a keyword identification to be able to identify what resources may be relevant. In some embodiments, the clustering engine 210 uses a variety of data, including user input via text, live data (e.g., streaming data), and Voice data, to determine what type of data to look for. … In one embodiment, the correlation engine 220 accesses a database of solutions to previous problems of the same type to identify a solution.”; e.g., “correlation engine 220 accesses a database of solutions to previous problems of the same type to identify a solution” may read on “machine learning model relevant to the problem”.)
executing, at the computer system, the at least one machine learning model relevant to the problem using the selected computing infrastructure to generate at least one recommendation relevant to the problem.
(Sun, [figs 2-3] “Machine Learning Engine 208” [pars 40-66] “the machine learning interface 204 may be configured to access one or more machine learning components of any type (e.g., the clustering engine 210 and the correlation engine 220) and receive some or all of an answer to one or more questions, without requiring users to interact with the underlying machine learning components. … The method 300 includes providing, by the first computing device, via the user interface component, a suggestion for solving the first problem, based on the retrieved first data (310). In one embodiment, the correlation engine 220 accesses a database of solutions to previous problems of the same type to identify a solution. In another embodiment, the correlation engine 220 identifies a description of a resolution associated with a problem that is substantially similar to the first problem and provides the description of the resolution to the user interface component 202 for display to the user. In still another embodiment, the correlation engine 220 applies a machine learning model to identify a second problem in the class; identifies a resolution associated with the second problem; and determines that the resolution to the second problem resolves the first problem.”; e.g., “correlation engine 220 accesses a database of solutions to previous problems of the same type to identify a solution” may read on “machine learning model relevant to the problem”.)
However, Sun does not appear to explicitly teach:
the data obtained [via] a sensor measuring environmental data;
the at least one machine learning model including an input layer, a data processing layer, and an output layer:
receiving, at the input layer, video and image data from the sensor:
selecting, at the processing layer, from a repository of computational models, a computational model for processing the [video and image] data based on a pattern recognized by the at least one machine learning model;
selecting, at the computer system, computing infrastructure upon which to execute the at least one machine learning model relevant to the problem, wherein the selected computing infrastructure comprises [a mesh of interconnected micro-applications, wherein each micro-application comprises analog input adapted to receive analog signals at edges of the mesh of micro-applications, digital input adapted to receive digital data from other micro-applications, event ingestion processing adapted to receive digital events and process and format the digital events for consumption, event consumption processing adapted to process the digital events to obtain information from the digital events and perform processing of the digital events, event generation processing adapted to generate another event to be output from micro-application to other micro-applications or out of the mesh of micro-applications, analog output adapted to output an analog signal representing the generated event, and digital output adapted to output a digital signal representing a digital event]; and
Ross teaches
the data obtained via a sensor measuring environmental data;
(Ross [par(s) 49] “FIG. 3 shows an example input into a machine learning model. The example input can be an image 300 recorded by a camera of a self-driving car. One or more machine learning models receive the input image and identify various objects of interest within the scene, such as "street" 310, "pedestrian" 320, 330, "stop sign" 340, "car" 350, etc. One or more machine learning models can also identify the region of the image within which each object of interest is located, such as street region 360, pedestrian region 370, 380, stop sign region 390, car region 395, etc. The region of the image 360, 370, 380, 390, 395 can be a bounding box of the identified feature in the image. In addition, one or more of machine learning models can identify a confidence level associated with each object of interest, such as 0.9 confidence level for street, 0.8 confidence level for pedestrian, 1 confidence level for stop sign, 0.85 confidence level for car. The machine learning models can be combined in parallel, in serial, hierarchically, etc., to identify each object of interest with an appropriate output label, such as "street" 310, "pedestrian" 320, 330, "stop sign" 340, "car" 350, etc. Further, after combining the existing machine learning models, the server can train each of the constituent machine learning model separately, thus reducing the memory and processor requirements needed to train the machine learning model.”;)
the at least one machine learning model including an input layer, a data processing layer, and an output layer:
(Ross [par(s) 44] “FIG. 2 is an example of a machine learning model, according to one embodiment. The machine learning model is an artificial neural network 200 ("neural network"), receiving input data at the input layer 210, processing the input through various layers 210, 220, 230, and outputting an output label at the output layer 230. The neural network 200 can contain multiple layers in addition to the layer 220, between the input layer 210, and the output layer 230. Each layer 210, 220, 230 includes one or more neurons, such as a neuron 240 in the input layer 210, a neuron 250 in the layer 220, a neuron 260 in the output layer 230. The number of neurons among the layers can be the same, or can differ. Each neuron in the layer 210, 220 can be connected to each neuron in the subsequent layer 220, 230, respectively. Alternatively, each neuron in the layer 210,220 can be connected to a subset of neurons in the subsequent layer 220, 230, respectively. Neurons are connected via connections 270 (only one labeled for brevity), where each connection includes a weight, where weight is a scalar number.”;)
receiving, at the input layer, video and image data from the sensor:
(Ross [par(s) 35] “The user interface module 110 receives from a first user device 150 a label describing a feature to identify within an input data. The input data can be video, audio, alphanumeric data, any combination of the foregoing, etc. The label can be a word such as "person", "stop sign", "car", "barking sound", etc.” [par(s) 46] “The input layer 210 receives the input data, such as video, audio, alphanumeric text, etc. For example, the neuron 240 receives a group of one or more pixels from the input data. The input data is processed through the layers 210, 220, 230, and the layer 230 outputs an output label, such as "a pedestrian." The output label can be accompanied by a confidence level associated with the output label, and/or an input data region, such as an image region, within which the neural network 200 has identified the output label. The confidence level is expressed within a normalized range, such as 0% to 100%, or Oto 1, indicating how confident the neural network 200 is that the input data region contains the identified output label.”;)
selecting, at the processing layer, from a repository of computational models, a computational model for processing the video and image data based on a pattern recognized by the at least one machine learning model;
(Ross [fig(s) 5A, 5B, 6] [par(s) 35] “The user interface module 110 receives from a first user device 150 a label describing a feature to identify within an input data. The input data can be video, audio, alphanumeric data, any combination of the foregoing, etc. The label can be a word such as "person", "stop sign", "car", "barking sound", etc.” [par(s) 46] “The input layer 210 receives the input data, such as video, audio, alphanumeric text, etc. For example, the neuron 240 receives a group of one or more pixels from the input data. The input data is processed through the layers 210, 220, 230, and the layer 230 outputs an output label, such as "a pedestrian." The output label can be accompanied by a confidence level associated with the output label, and/or an input data region, such as an image region, within which the neural network 200 has identified the output label. The confidence level is expressed within a normalized range, such as 0% to 100%, or Oto 1, indicating how confident the neural network 200 is that the input data region contains the identified output label.” [par(s) 56] “FIG. 6 shows two machine learning models hierarchically combined. An output layer 610, of a machine learning model 600 is connected to multiple machine learning models 620, 630. The output layer 610 can be connected directly to the input layer's 640, 650 of the machine learning model 620, 630 respectively, or can be connected to the input layer 640, 650 through an interface mechanism 660, 670, respectively. When the output layer 610 is directly connected to the input layer 640, 650, a neuron of the output layer 610 can be connected to a neuron in the input layers 640, 650 in a one-to-one, one-to-many, many-to-one, or one-to-none mapping.”;)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the recommendation system of Sun with the video and image data of Ross.
Doing so would lead to improving the accuracy of the machine learning model by combining the existing machine learning models in serial, in parallel, or hierarchically, to create a resulting machine learning model.
(Ross, [pars 28-43] “the system can improve the accuracy of the machine learning model by combining the existing machine learning models in serial, in parallel, or hierarchically, to create a resulting machine learning model.”;)
However, the combination of Sun, Ross does not appear to explicitly teach:
selecting, at the computer system, computing infrastructure upon which to execute the at least one machine learning model relevant to the problem, wherein the selected computing infrastructure comprises [a mesh of interconnected micro-applications, wherein each micro-application comprises analog input adapted to receive analog signals at edges of the mesh of micro-applications, digital input adapted to receive digital data from other micro-applications, event ingestion processing adapted to receive digital events and process and format the digital events for consumption, event consumption processing adapted to process the digital events to obtain information from the digital events and perform processing of the digital events, event generation processing adapted to generate another event to be output from micro-application to other micro-applications or out of the mesh of micro-applications, analog output adapted to output an analog signal representing the generated event, and digital output adapted to output a digital signal representing a digital event]; and
TEIG teaches
selecting, at the computer system, computing infrastructure upon which to execute the at least one machine learning model relevant to the problem, wherein the selected computing infrastructure comprises a mesh of interconnected micro-applications,
(TEIG [fig(s) 5-6, 14] [par(s) 31] “FIG. 1 shows an example machine learning assembly 100. In an implementation, the example machine learning assembly 100 can be configured from pluggable, interchangeable modules that may each have at least one compatible port for interconnecting with each other. A “core” module 102 contains at least one machine learning kernel embodied in a neural network, and multiple core modules 102 can be coupled together in series, in parallel, or as a cluster to expand the neural network. Peripheral modules 104, 106, 108, 110, 112 & 114 are attachable to the core module 102, or to each other, but all peripheral modules can be in communication with at least one core module 102.” [par(s) 71-73] “FIG. 14 shows an example configuration of a machine learning assembly 1400, with core modules 102 & 314, each containing at least one set of core components 500 & 500′, coupled together via interconnection module 1402. Sensor module 202 is coupled with core module 102, utility module 1404 is coupled to core module 102 via interconnection module 1406, sensor modules 1408 & 1410 are coupled to core module 102 and to utility module 1404 via interconnection module 1406. Sensor module 1412 is coupled to core module 102 via interconnection module 1414.” see also [par(s) 40-50];)
wherein each micro-application comprises analog input adapted to receive analog signals at edges of the mesh of micro-applications,
digital input adapted to receive digital data from other micro-applications,
(TEIG [fig(s) 5-6, 14] [par(s) 32] “the interconnection modules 210 may include analog-to-digital converters to render analog output of the sensor modules 202, 204 & 206, into digital signals for processing at the core module 102.” [par(s) 40-50] “For control and security, if any raw digital data or raw analog data is selected to leave the system 200, the data being released is only released from the processing unit 502, and the decision to release the data is reserved to the processing unit 502. … The example core components 500 may include discrete analog-to-digital converters 606 to render analog sensor input 510 into digital signals for the processing unit 502.”;)
event ingestion processing adapted to receive digital events and process and format the digital events for consumption,
event consumption processing adapted to process the digital events to obtain information from the digital events and perform processing of the digital events,
event generation processing adapted to generate another event to be output from micro-application to other micro-applications or out of the mesh of micro-applications,
(TEIG [fig(s) 5-6, 14] [par(s) 32] “the interconnection modules 210 may include analog-to-digital converters to render analog output of the sensor modules 202, 204 & 206, into digital signals for processing at the core module 102.” [par(s) 83-89] “An example machine-trained (MT) neural network 508 that may be used in some embodiments may utilize novel processing nodes with novel activation functions 610 that allow the MT neural network 508 to efficiently define a scenario with fewer processing node layers to solve a particular problem (e.g., face recognition, speech recognition, pattern recognition, and so forth). … assume that a sensing system 1704 is an imaging system located in an automobile and has been trained to recognize pedestrians. … If the neural network 508 recognizes or learns a pattern that indicates the presence of a pedestrian near a travel path of the automobile, the executive output 512 is used to alert the driver that a pedestrian is nearby.” [par(s) 72] “The utility module 1404 can also be an executive component performing the executive output 512 of the neural network 508, such as … a digital to analog converter” [par(s) 85] “The processing unit 502 can embody special purpose logic circuitry, e.g., an FPGA (field programmable gate array 604) or an ASIC (application-specific integrated circuit).”;)
analog output adapted to output an analog signal representing the generated event, and
digital output adapted to output a digital signal representing a digital event; and
(TEIG [fig(s) 5-6, 14] [par(s) 32] “the interconnection modules 210 may include analog-to-digital converters to render analog output of the sensor modules 202, 204 & 206, into digital signals for processing at the core module 102.” [par(s) 40-50] “Significantly, in an implementation, all the raw data 510 input from sensors to a core 102 remains within the system 200. For control and security, if any raw digital data or raw analog data is selected to leave the system 200, the data being released is only released from the processing unit 502, and the decision to release the data is reserved to the processing unit 502. The outgoing executive output 512 is generally an indicator, decision, or control directive generated by the neural network 508 that also remains within the device or machine hosting the example machine learning assembly 200, and is not raw data 510 incoming from sensors.” [par(s) 72] “The utility module 1404 can also be an executive component performing the executive output 512 of the neural network 508, such as a controller, an actuator, an alarm, a solenoid, a navigation system, a user interface intermediary, a display driver, camera electronics, a transmitter, an electrode, a digital to analog converter, an implantable medical device interface, an insulin pump controller for a medical patient, a pacemaker trigger, an implantable cardioverter-defibrillator interface, a hospital IV pump controller, a pager, a cell phone element, a heating-air-conditioning-and-ventilation governor, and so forth, as examples.”;)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the recommendation system of Sun, Ross with the interconnected micro-applications of TEIG.
One of ordinary skill in the art would have been motived to combine in order to draw low power for use in smart phones, watches, drones, automobiles, and medical devices.
(TEIG [par(s) 5] “Configurable machine learning assemblies for autonomous operation in personal devices are provided. Example systems implement machine learning based on neural networks that draw low power for use in smart phones, watches, drones, automobiles, and medical devices.”)
In the alternative, WETMORE can also be interpreted to teach the following limitation:
WETMORE teaches
analog input adapted to receive analog signals at edges of the mesh of micro-applications,
digital input adapted to receive digital data from other micro-applications,
analog output adapted to output an analog signal representing the generated event, and
digital output adapted to output a digital signal representing a digital event; and
(WETMORE [fig(s) 11] [par(s) 158] “By applying machine learning and clustering techniques to look at previous training data for a particular user, or population of users, the rate of memory consolidation and the resolution of sleep phase detection can be optimized for a user or users. To classify such signals any of the following methods may be used, including, but not limited to: Independent Component Analysis; Principal Component Analysis; Latent Dirichlet Allocation; Naive Bayesian Classifier; K-means clustering; and Neural networks.” [par(s) 172] “In this embodiment, a programmable microcontroller board applies control logic that determines device function based on user inputs, the user's sleep state or state of wakefulness, and previous device use by the user. For example, an Arduino open source microcontroller framework is an effective programmable microcontroller board used in this embodiment. The Arduino system includes digital inputs and outputs, analog inputs and outputs, serial receiver, serial transmission, and power (both 5V and 3.3V). The microcontroller system in this example is programmed with custom software for controlling the various elements of the system for memory enhancement.”;)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the recommendation system of Sun, Ross, TEIG with the inputs and outputs of a micro-application of WETMORE.
One of ordinary skill in the art would have been motived to combine in order to improve the quality of memory enhancement and/or sleep detection for the user.
(WETMORE [par(s) 128-140] “In some variations the system or device uses data mining and/or other statistical techniques to analyze stored data in order to improve or otherwise modify the quality of memory enhancement and/or sleep detection for the user or for other users (FIG. 10).”)
Regarding claim 3
The combination of Sun, Ross, TEIG, WETMORE teaches claim 1.
Sun further teaches
the data relating to the problem to be solved comprises at least one of data from sensors, data from devices, data from servers, data from robots, and data from humans.
(Sun, [figs 1] [pars 40-66] “Referring now to FIG. 3A, in connection with FIG. 2, a flow diagram depicts one embodiment of a method 300 for receiving a description of a problem and applying machine learning to automatically solve the problem. In brief overview, the method 300 includes receiving, by a first computing device, from a second computing device, via a user interface component, a description of a first problem (302). The method 300 includes assigning, by a clustering engine executing on the first computing device, the first problem to a class (304). … In one embodiment, the user interface component 202 receives the description of the first problem directly from a user of the first machine 106. In another embodiment, the machine 106 provides the client machine 102 with access to a user interface component 202 through which the user of the client machine 102 may provide the description of the first problem. .. For example, the system 200 may receive a description of a first problem (e.g., from a human or from another machine) and proceed to identify related problems and solutions with little or no human intervention.”;)
Regarding claim 4
The combination of Sun, Ross, TEIG, WETMORE teaches claim 3.
Sun further teaches
the at least one machine learning model relevant to the problem is obtained by at least one of:
selecting, at the computer system, at least one model from among previously used processed models stored at the computer system;
selecting, at the computer system, at least one model from among models obtained from public sources, proprietary sources, or both; and
generating, at the computer system, a new model based on type, morphology, and parameter information.
(Sun, [figs 2-3] “Machine Learning Engine 208” [pars 40-66] “The machine learning engine 208 may include an ensemble of classifiers, for example. In an embodiment in which the machine learning engine 208 includes functionality for executing a clustering algorithm to identify patterns in data (e.g., the clustering engine 210), a variety of algorithms may be used, including, without limitation, K-Means and Expectation-Maximization algorithms. … the machine learning interface 204 may be configured to access one or more machine learning components of any type (e.g., the clustering engine 210 and the correlation engine 220) and receive some or all of an answer to one or more questions, without requiring users to interact with the underlying machine learning components. … the clustering engine 210 provides the description of the first problem and the assigned class to the correlation engine 220. In some embodiments, by determining a class of the first problem, the clustering engine 210 enables the correlation engine 220 to identify related databases, problems (e.g., problems of the same class), and problem resolutions. … In one embodiment, the correlation engine 220 accesses a database of solutions to previous problems of the same type to identify a solution”; )
In the alternative, Ross can also be interpreted to teach the following limitation:
the at least one machine learning model relevant to the problem is obtained by at least one of:
selecting, at the computer system, at least one model from among previously used processed models stored at the computer system;
selecting, at the computer system, at least one model from among models obtained from public sources, proprietary sources, or both; and
generating, at the computer system, a new model based on type, morphology, and parameter information.
(Ross, [fig 1] [pars 28-43] “The system presented here can gather machine learning models from various sources, including at least one unreliable source. The machine learning models are identified by the input format that they receive, and an output label that they identify from the input format. For example, a machine learning model can take video as input format and produce an output label "chair". … If none of the retrieved machine learning models identify "chair" to within the predefined accuracy level, the system can improve the accuracy of the machine learning model by combining the existing machine learning models in serial, in parallel, or hierarchically, to create a resulting machine learning model.”;)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the recommendation system of Sun, Ross, TEIG, WETMORE with the selection of existing models of Ross.
Doing so would lead to improving the accuracy of the machine learning model by combining the existing machine learning models in serial, in parallel, or hierarchically, to create a resulting machine learning model.
(Ross, [pars 28-43] “the system can improve the accuracy of the machine learning model by combining the existing machine learning models in serial, in parallel, or hierarchically, to create a resulting machine learning model.”;)
Regarding claim 5
The combination of Sun, Ross, TEIG, WETMORE teaches claim 4.
the at least one machine learning model relevant to the problem is further obtained by: (see the rejections of claim 1)
Ross teaches
determining, at the computer system, a combination of the selected and generated models that produces higher accuracy results than the selected and generated models; and
(Ross, [fig 1] [fig 8] “obtain from various sources including unreliable sources machine learning models trained to identify the output label from the input data” [pars 28-43] “A user of the system can request the system to provide the user with a machine learning model to identify "chair" to within a predefined accuracy level, from video input data. The predefined accuracy level can be user-specified, or automatically determined. The system then finds the machine learning models associated with the output label "chair". If none of the retrieved machine learning models identify "chair" to within the predefined accuracy level, the system can improve the accuracy of the machine learning model by combining the existing machine learning models in serial, in parallel, or hierarchically, to create a resulting machine learning model. Further, even after combining the existing machine learning models, the system can train each of the constituent machine learning model separately, thus reducing the memory and processor requirements needed to train the resulting machine learning model. … The gathering module 120 obtains from one or more sources 160, including at least one unreliable source, one or more learning models trained to identify the label. The gathering module 120 stores the received machine learning models in the database 195.” see also [pars 44-51] regarding accuracy; e.g., “If none of the retrieved machine learning models identify "chair" to within the predefined accuracy level, the system can improve the accuracy of the machine learning model by combining the existing machine learning models in serial, in parallel, or hierarchically, to create a resulting machine learning model” and “The gathering module 120 obtains … one or more learning models trained” may read on “determining, at the computer system, a combination of the selected and generated models that produces higher accuracy results than the selected and generated models”.)
assembling, at the computer system, a combination of the selected and generated models based on the determination of the combination of the selected and generated models that produces higher accuracy results than the selected and generated models.
(Ross, [fig 1] [pars 28-43] “A user of the system can request the system to provide the user with a machine learning model to identify "chair" to within a predefined accuracy level, from video input data. The predefined accuracy level can be user-specified, or automatically determined. The system then finds the machine learning models associated with the output label "chair". If none of the retrieved machine learning models identify "chair" to within the predefined accuracy level, the system can improve the accuracy of the machine learning model by combining the existing machine learning models in serial, in parallel, or hierarchically, to create a resulting machine learning model. Further, even after combining the existing machine learning models, the system can train each of the constituent machine learning model separately, thus reducing the memory and processor requirements needed to train the resulting machine learning model. The resulting machine learning model can be encrypted before being sent to the user.”; e.g., “combining the existing machine learning models in serial, in parallel, or hierarchically, to create a resulting machine learning model” may read on “assembling”.)
The combination of Sun, Ross, TEIG, WETMORE is combinable with Ross for the same rationale as set forth above with respect to claim 4.
Regarding claim 7
The claim is a system claim corresponding to the method claim 1, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejection of the method claim.
Note that Sun teaches a computer readable storage and a computer system.
(Sun [fig(s) 1])
Regarding claim 9
The claim is a system claim corresponding to the method claim 3, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejection of the method claim.
Regarding claim 10
The claim is a system claim corresponding to the method claim 4, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejection of the method claim.
Regarding claim 11
The claim is a system claim corresponding to the method claims 4 and 5, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejection of the method claims.
Regarding claim 13
The claim is a product claim corresponding to the method claim 1, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejection of the method claim.
Note that Sun teaches a processor and memory.
(Sun [fig(s) 1])
Regarding claim 15
The claim is a product claim corresponding to the method claim 3, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejection of the method claim.
Regarding claim 16
The claim is a product claim corresponding to the method claim 4, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejection of the method claim.
Regarding claim 17
The claim is a product claim corresponding to the method claims 4 and 5, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejection of the method claims.
Claim(s) 6, 12, 18 is/are rejected under 35 U.S.C. 103 as being unpatentable over Sun et al. (US 2017/0011308 A1) in view of Ross et al. (US 2018/0232663 A1) in view of TEIG et al. (US 20180025268 A1) further in view of WETMORE et al. (US 20140057232 A1) further in view of Martinez-Munoz et al. (An Analysis of Ensemble Pruning Techniques Based on Ordered Aggregation)
Regarding claim 6
The combination of Sun, Ross, TEIG, WETMORE teaches claim 5.
the combination of the selected and generated models that produces higher accuracy results than the selected and generated models may be determined (see the rejections of claim 5) [by selected and trained heuristics or by a machine learning model].
However, the combination of Sun, Ross, TEIG, WETMORE does not appear to distinctly disclose:
the combination of the selected and generated models that produces higher accuracy results than the selected and generated models may be determined [by selected and trained heuristics or by a machine learning model].
Martinez-Munoz teaches
the combination of the selected and generated models that produces higher accuracy results than the selected and generated models may be determined by selected and trained heuristics or by a machine learning model.
(Martinez-Munoz, [tables 5-7] [sec Abs] “Several pruning strategies that can be used to reduce the size and increase the accuracy of bagging ensembles are analyzed. These heuristics select subsets of complementary classifiers that, when combined, can perform better than the whole ensemble. The pruning methods investigated are based on modifying the order of aggregation of classifiers in the ensemble. … The performance of these pruned ensembles is evaluated in several benchmark classification tasks under different training conditions. The results of this empirical investigation show that ordered aggregation can be used for the efficient generation of pruned ensembles that are competitive, in terms of performance and robustness of classification, with computationally more costly methods that directly select optimal or near-optimal subensembles.” [sec 3] “we introduce a family of pruning methods based on modifying the order in which classifiers are aggregated in a bagging ensemble.” [sec 4.5] “The performance of the different pruning heuristics is evaluated in a series of experiments on 28 classification tasks from the UCI repository [45]. Each experiment consists in 100 executions for each data set. For the synthetic problems (Led24, Ringnorm, Twonorm, and Waveform) random training and testing samples are generated”; Note that Ross teaches “the combination of the selected and generated models that produces higher accuracy results than the selected and generated models may be determined [by selected and trained heuristics or by a machine learning model]”.)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the recommendation system of Sun, Ross, TEIG, WETMORE with the selected and trained heuristics of Martinez-Munoz.
Doing so would lead to providing accurate and efficient classifiers which can be used for the efficient generation of pruned ensembles that are competitive, in terms of performance and robustness of classification, with computationally more costly methods that directly select optimal or near-optimal subensembles.
(Martinez-Munoz, [sec Abs] “The results of this empirical investigation show that ordered aggregation can be used for the efficient generation of pruned ensembles that are competitive, in terms of performance and robustness of classification, with computationally more costly methods that directly select optimal or near-optimal subensembles.”)
Regarding claim 12
The claim is a system claim corresponding to the method claim 5, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejection of the method claim.
Regarding claim 18
The claim is a product claim corresponding to the method claim 5, and is directed to largely the same subject matter. Thus, it is rejected for the same reasons as given in the rejection of the method claim.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SEHWAN KIM whose telephone number is (571)270-7409. The examiner can normally be reached Mon - Thu 7:00 AM - 5:00 PM.
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, Michael J Huntley can be reached on (303) 297-4307. 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.
/SEHWAN KIM/Examiner, Art Unit 2129
Prosecution Insights