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 .
Remarks
This Office Action is responsive to Applicants' Amendment filed on 03/05/2026, in which claims 1, 6-8, 13-15, and 17 have been amended. Claim 20 has been cancelled. New claim 21 has been added.
Claims 1-19, and 21 are currently pending.
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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 03/05/2026 has been entered.
Response to Arguments
With regards to the objections to claims 7, 14, and 17 due to minor informalities, Applicant’s arguments that the amendments correct the noted informalities is partially persuasive. Claims 7 and 17 have all previously noted informalities corrected, and thus the objections are withdrawn. However, Claim 14 still recites one previously noted informality, and thus the objection is maintained.
With regards to the rejections of claims 1-5, 7-12, and 14-19 under 35 U.S.C. 101 for being directed towards abstract ideas, Applicant’s arguments that the claims as amended overcome the rejections at least at Step 2B of the Subject Matter Eligibility Test are persuasive. Independent claims 1, 8, and 15 now recite subject matter similar to that identified previously in dependent claims 6, 13, and now-cancelled claim 20 as being substantially more than any recited judicial exceptions.
With regards to the rejections of claims 1, 7, 8, 14, and 15 under 35 U.S.C. 103 as being unpatentable over Li, in view of Wenzel, Applicant argues that the combination of Li and Wenzel does not properly teach the rejection as amended. While Examiner agrees and a new combination of art is cited for the rejections of the current Office Action, points salient to art previously cited and still relied upon are discussed below.
First, Applicant argues on pages 16 and 17 of the Remarks that Li does not teach “deriving” or “derivation” as used within claim 1, and that the Office Action uses the terms “deriving” and “derivation” inconsistently in general. Applicant further argues that the specification at paragraph [0047] explains the meanings of “derivation” and “derive”, that the mapping comment “’correlations between a data objects [sic] and data fields corresponds to associations between a data object and derivations from a first set of data values’” on page 12 of the Office Action is not consistent with the specification’s explanation of “derivation”, that the word derivation is omitted when mapping to Wenzel on page 13 of the Office Action, and that in the mapping to claim 7 on page 14 of the Office Action, mapping “’deriving values from variables’” to “deriving a set of values” is inconsistent interpretation of “deriving”.
Examiner respectfully disagrees. First, paragraph [0047] of the specification does not define “deriving” or “derivation”. Paragraph [0047] defines the behavior of “data derivation manager 125”, which is not recited within the claims. The broadest reasonable interpretation of “deriving” or “derivation” is broader than the particular behavior than Applicant’s exact module as described in the specification.
Further, Examiner disagrees that the Office Action uses “deriving” or “derivation” inconsistently. With regards to the mapping of the former limitation of claim 1 “and storing an association between the data object and a derivation of the first set of data values”, the cited paragraph from Li, Li [0038], reads more fully:
“In some examples, when a user device transmits a particular data object, one or more variables (e.g., data fields within the data object) can be correlated to a set of stored data objects. The supporting information associated with the set of stored data objects can be used to identify supporting information that is specific to the particular data object”.
Li states that stored supporting information can be used to derive further supporting information both above, and again at (Li [0037]) “if there is particular supporting information that is frequently associated with data objects that have been previously submitted, those correlations can be used to predict future correlations. For example, if a data object includes a transportation event using a particular method (e.g., train), the supporting information can include travel schedules associated with the particular method of travel (e.g., train schedules)”. So, Li does teach derivation from a first set of data values in the paragraph of Li cited by Examiner.
However, the point is moot because the new art of Onishi is used to map derivation of emissions values using a lookup value, as claim 1 is amended to recite, see the rejections below.
In the mapping to Wenzel for claim 7, no comment on derivation was necessary, because it was sufficiently clear that the cited “calculat[ion] therefrom consumption values and/or emission values” from “driving profiles” at Wenzel [0063] was a derivation. Wenzel is no longer cited, so the point is moot.
Applicant further argues on pages 17 and 18 of the Remarks that the obviousness rejection of claim 1 of the previous Office Action was not properly supported. Applicant states:
“the rejection does not set forth the proposed modification Li, in view of Wenzel, necessary to arrive at the claimed subject matter. See MPEP 2142 Legal Concept of Prima Facie Obviousness, requirement C…The Office Action does not explain how Li's node-identification framework would be modified to generate a first set of data values for a categorized data object and then derive values from the first set of data values, nor does it articulate how Wenzel's traffic-based emissions modeling would be incorporated into Li's architecture to produce and store derived emissions values in association with the data object”.
Examiner respectfully disagrees that the previous obviousness rejection of claim 1 was not supported. Wenzel was only relied upon for teaching that a derivation of a set of data values should be emissions values. Substituting the generic “supporting information” that Li derives for the more specific “emissions values” of Wenzel would not require any architectural changes to Li, just use with different data. The previous Office Action stated at page 13 how such a substitution would take place and why such a substitution would confer a predictable benefit. However, the point is technically moot as Wenzel is no longer relied upon to teach any limitations within claim 1, see the rejections below.
Claim Objections
Claim 14 is objected to because of the following informality: based on a subset of the set of data values; should read “based on a subset of the first set of data values;”. Appropriate correction is required.
Claim Rejections - 35 USC § 112b
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-19 and 21 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Regarding claim 1,
Claim 1 recites the limitation wherein at least one of the first set of data values is used as a lookup value for at least one of the one or more derived emissions values defined for the subcategory;. However, the “one or more emissions values” that are earlier recited in the claim are not stated to be defined for any subcategory. Therefore it is unclear whether “the one or more derived emissions values defined for the subcategory” has antecedent basis or not. Therefore, the scope of the claim is indefinite. For examination purposes, the limitation will be interpreted as reading “wherein at least one of the first set of data values is used as a lookup value for at least one of the one or more derived emissions values, and wherein the one or more derived emissions values are defined for the subcategory;”.
In reference to dependent claims 2-7, claims 2-7 do not cure the deficiencies noted in the rejection of claim 1. Therefore, claims 2-7 are rejected under the same rationale as claim 1.
Regarding claim 8,
Claim 8 recites a method for performing the function of the medium of claim 1. All limitations of claim 1 have substantial equivalents within claim 8, therefore claim 8 is considered indefinite with an equivalent rationale and is interpreted for examination purposes in the same way.
In reference to dependent claims 9-14, claims 9-14 do not cure the deficiencies noted in the rejection of claim 8. Therefore, claims 9-14 are rejected under the same rationale as claim 8.
Regarding claim 15,
Claim 15 recites a system for performing the function of the medium of claim 1. All limitations of claim 1 have substantial equivalents within claim 15, therefore claim 15 is considered indefinite with an equivalent rationale and is interpreted for examination purposes in the same way.
In reference to dependent claims 16-19 and 21, claims 16-19 and 21 do not cure the deficiencies noted in the rejection of claim 15. Therefore, claims 16-19 and 21 are rejected under the same rationale as claim 15.
Prior Art
The following references are used for prior art claim rejections:
Li et al. (U.S. Patent Application Publication No. 2018/0285777), hereinafter Li
Onishi et al. (U.S. Patent Application Publication No. 2011/0137768), hereinafter Onishi
Balasubramanian et al. (U.S. Patent Application Publication No. 2019/0303796), hereinafter Balasubramanian
Putrevu et al. (U.S. Patent Application Publication No. 2022/0391965), hereinafter Putrevu
Dong et al. “Vehicle Type Classification Using a Semisupervised Convolutional Neural Network”, hereinafter Dong
Aslandere et al. (U.S. Patent Application Publication No. 2021/0285779), hereinafter Aslandere
Biswas (U.S. Patent Application Publication No. 2022/0398485), hereinafter Biswas
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 7, 8, 14, and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Li, in view of Onishi, further in view of Balasubramanian.
Regarding claim 1,
Li teaches A non-transitory machine-readable medium storing a program executable by at least one processing unit of a device, the program comprising sets of instructions for: ((Li [0007]) “In some embodiments, a computer-program product is provided that is tangibly embodied in a non-transitory machine-readable storage medium. The computer-program product can include instructions configured to cause one or more data processors to perform part or all of a method disclosed herein”)
retrieving a data object associated with a defined category; ((Li [0034]) “The tree-learning model may be configured with two assumptions: 1) it may be assumed that a user who has initiated a process associated with a particular event (e.g., a travel plan, an incurred expense, a hotel reservation, a meal expense at a restaurant, and so on) will transmit at least one data object, and if so, what values may be included in the data object, and/or 2) it may be assumed that the user will submit a data object that includes one event in each category of a plurality of categorized event types (e.g., travel, lodging, dining, and so on), and if so, which values have previous users submitted for each event type”)
retrieving a set of machine learning models corresponding to a subcategory ((Li [0006]) “The method can also include defining one or more evaluation metrics using the one or more event parameters. Each evaluation metric can be used to classify the one or more events into an event type. Further, the method can include assessing the one or more evaluation metrics and the data set. Assessing can include executing the one or more machine-learning algorithms to generate the machine-learning model. The execution of the one or more machine-learning algorithms can generate a plurality of nodes and one or more correlations between at least two nodes of the plurality of nodes”, (Li [0037]) “the machine-learning models can be used to identify supporting documents or information associated with events”, event parameters correspond to subcategories, machine learning models that identify supporting information correspond to a set of machine learning models) of the defined category associated with the data object; ((Li [0035]) “a data object may include a receipt for a dinner, a receipt for a hotel stay, and a receipt for a flight expense. In some cases, a data object may not include an expense report, but rather, a request to define an event with anticipated expenses. An event can have one or more event parameters (e.g., location of the hotel, price of the dinner, price of the flight, destination city of the flight, etc.). Further, event parameters of various events can be classified into one or more event types (e.g., hotels, flights, meals, transportation, etc.)”, event parameters of a data object that are classified into event types correspond to subcategories of a defined category associated with the data object)
receiving, based at least in part on the set of machine learning models, a first set of data values, ((Li [0037]) “the machine-learning models can be used to identify supporting documents or information associated with events…Further, based on machine-learning analysis, metadata associated with the event can be automatically identified and presented on the interface”, supporting information or metadata can correspond to a first set of data values)
Onishi teaches the following further limitations that Li does not teach:
deriving, from the first set of data values, one or more emissions values, ((Onishi [0106]) “The emission calculating unit 102c applies the transportation facility (for example, a train) and the travel distance (for example, about 9.6km) extracted at Step SC-1 to the searched calculation standard to calculate greenhouse gas emissions (for example, about 182 grams) (Step SC-3)”, the values extracted at Step SC-1 correspond to a set of data values)
wherein at least one of the first set of data values is used as a lookup value for at least one of the one or more derived emissions values defined for the subcategory; (Onishi ([0105]) “The emission calculating unit 102c inquires the emission calculation-standard file 106b for emission calculation standards different for each of the extracted transportation facility (for example, 173 (g/km) in a case of a private car, 111 (g/km) in a case of an airplane, 51 (g/km) in a case of a bus, and 19 (g/km) in a case of a train) (Step SC-2)”, a value used to inquire a file corresponds to a lookup value, a transportation method is a subcategory)
and storing an association between the data object and the one or more derived emissions values in an application data storage or in the data object ((Onishi [0059]) “The history-information file 106d is a history-information storage unit that stores history information in which at least a part of a route search result (a route in which greenhouse gas emissions have been offset, travel expenses required at the time of using the route and the like)…and the offset result (offset greenhouse gas emissions, purchase amount of emission credits corresponding to the greenhouse gas emissions and the like) are associated with each other for each of members”, offset greenhouse gas emissions are an emissions value derived from the route information, a route search result corresponds to a data object)
At the time of filing, one of ordinary skill in the art would have motivation to combine Li and Onishi by taking the medium containing instructions for retrieving a data object with a category, including a transportation category, machine learning models with subcategories of the category, and receiving data values based on the set of machine learning models, taught by Li, and adding deriving emissions values from the data values using at least one data value as a lookup value, and storing an association between a data object and the derived emissions values, taught by Onishi, as Onishi teaches: (Onishi [0012]) “according to the present invention, when companies and the like declare to offset their carbon footprint voluntarily or are going to offset their carbon footprint according to the laws and regulations or the like, it is possible to verify whether their carbon footprint is being offset with respect to greenhouse gas emissions generated at the time of travel of the members such as employees”, that is, that deriving and storing emissions data in this way provides a predictable benefit for companies or other organizations to verify and record their emissions for accurate offsetting of the emissions later, either to provide evidence of voluntary environmental efforts or to cooperate with any relevant environmental regulations. Such a combination would be obvious.
Balasubramanian teaches the following further limitations that neither Li nor Onishi teach:
wherein at least one portion of the first set of data values are generated using a regular expression instead of the [set of] machine learning models if the at least one portion is successfully extracted using the regular expression; ((Balasubramanian [0308]-[0309]) “The input company name goes through the first module 606 to detect any static profanity. The module 606 marks a field IsFrivolousCompany to true or false based on regex 310 profanity comparison. These regex are pre-determined and stored in a SQL table or other data storage structure. If IsFrivolousCompany is true, the other two modules are skipped by the orchestration engine 614 and that input company name is marked as frivolous when returned by the service 604…If the field IsFrivolousCompany is false, the company name goes to the second module 610 to detect any other forms of profanity or keyboard gibberish….More details about the machine learning module 610 can be found elsewhere herein”, using a regex module first to generate a value and skipping a machine learning module if the regex module is successful in detection corresponds to generating at least a portion of a set of data values using a regular expression instead of a machine learning model, Li but not Subramanian explicitly teaches a set of machine learning models)
At the time of filing, one of ordinary skill in the art would have motivation to combine Li, Onishi, and Balasubramanian by taking the medium containing instructions for retrieving a data object with a category, including a transportation category, machine learning models with subcategories of the category, receiving data values based on the set of machine learning models, deriving emissions values from the data values using at least one data value as a lookup value, and storing an association between the data object and the derived emissions values, taught jointly by Li and Onishi, and attempting to determine data values based on a regular expression (also known as regex) before attempting to determine the data values using a machine learning model, taught by Balasubramanian, as regex is a much less computationally complex technique than many machine learning techniques, such as deep learning, and thus requires less time and computational power to use. Using regex in place of a machine learning model to determine data values when viable thus provides the predictable benefit of faster and cheaper computation of the data values. Such a combination would be obvious.
Regarding claim 7,
Li, Onishi, and Balasubramanian jointly teach The non-transitory machine-readable medium of claim 1, wherein the one or more derived emissions values are generated by:
Li further teaches:
determining a set of defined data from a plurality of sets of defined data based on a subset of the first set of data values; ((Li [0044]) “the data set and the evaluation metrics may be assessed. In some instances, assessing the data set and the evaluation metrics includes determining or identifying correlations between data objects within the data set. For example, assessing the data set and the evaluation metrics can include executing the one or more machine-learning algorithms to generate the machine-learning model…Further, executing the one or more machine-learning algorithms generates a plurality of nodes and one or more correlations between at least two nodes of the plurality of nodes”, a plurality of nodes is a set of defined data, assessing the data set and the evaluation metrics, including executing a machine-learning algorithm, corresponds to determining based on a subset of the set of data values)
and deriving a second set of data values based further on the set of defined data, ((Li [0006]) “each variable of the one or more variables can represent a characteristic of the particular event. The one or more variables can be mapped to the plurality of nodes of the machine-learning model. Based at least in part on the mapping, one or more nodes for each of the one or more variables can be identified. The one or more nodes can be included in the plurality of nodes of the machine-learning model. Further, the one or more nodes can be identified using the one or more correlations. One or more values associated with each of the nodes included in the one or more nodes can be retrieved”, deriving values from variables representing event characteristics mapped to nodes corresponds to deriving a set of data values based on the set of defined data nodes)
Onishi further teaches:
wherein the second set of data values comprises the one or more derived emissions values ((Onishi [0106]) “The emission calculating unit 102c applies the transportation facility (for example, a train) and the travel distance (for example, about 9.6km) extracted at Step SC-1 to the searched calculation standard to calculate greenhouse gas emissions (for example, about 182 grams) (Step SC-3)”, possible transportation types are a set of defined data)
At the time of filing, one of ordinary skill in the art would have motivation to combine the medium jointly taught by Li, Onishi, and Balasubramanian for the parent claim of claim 7, claim 1. No new embodiments are introduced, so the reason to combine is the same as for the parent claim.
Regarding claims 8 and 14,
Claims 8 and 14 recite a method for performing the function of the medium of claims 1 and 7, respectively. All other limitations in claims 8 and 14 are substantially the same as those in claims 1 and 7, therefore the same rationale for rejection applies.
Regarding claim 15,
Claim 15 recites a system for performing the function of the medium of claim 1. Specifically, claim 15 recites: A system comprising: a set of processing units; ((Li [0095]) “Processing unit 604, which can be implemented as one or more integrated circuits (e.g., a conventional microprocessor or microcontroller), controls the operation of computer system 600”)
All other limitations in claim 15 are substantially the same as those in claim 1, therefore the same rationale for rejection applies.
Claims 2, 9, and 16 are rejected under 35 U.S.C. 103 as being unpatentable over Li in view of Onishi, further in view of Balasubramanian, further in view of Putrevu.
Regarding claim 2,
Li, Onishi, and Balasubramanian jointly teach The non-transitory machine-readable medium of claim 1,
Li further teaches:
wherein the data object comprises a third set of data, ((Li [0008]) “the supporting information can be automatically attached to event records (e.g., expense entries) associated with particular entities (e.g., a train company) or locations (e.g., a destination city) by mapping the event record to those entities or locations (e.g., using text analysis in the image of a receipt), and having a separate mapping from entities or locations to supporting information”, entities and locations attached to event records correspond to a third set of data of a data object)
wherein retrieving the set of machine learning models comprises, upon determining that the subcategory of the defined category associated with the data object is a first subcategory in a plurality of defined subcategories, ((Li [0042]) “The machine-learning model may be generated using one or more machine-learning algorithms. For example, the data set can include one or more events (e.g., an incurred expense, flight reservation, booked hotel, paid meal, etc.) that have previously occurred. Each event of the one or more events can include or correspond to one or more event parameters that identify a characteristic of the event. Examples of an event parameter can include a location of the event (e.g., destination city of a flight), a vendor name associated with the event (e.g., restaurant name), a date and time of the event (e.g., time of the expense), distance of an event (e.g., travel distance when travelling by train or bus), and other suitable parameters of an event”, event parameters of events with types correspond to subcategories of defined categories, several event parameters are defined such as destination city, time, and distance)
Putrevu teaches the following further limitation that neither Li, nor Balasubramanian teaches, and more explicitly than Onishi teaches:
including a first model configured to predict distance values based on the third set of data of the data object in the set of machine learning models ((Putrevu [0007]) “The online concierge system leverages the obtained information describing distances between different pairs of starting locations and destination locations to generate and to train a distance prediction model that outputs a predicted travel distance between a starting location and a destination location”, pairs of starting locations and destination locations are a third set of data)
At the time of filing, one of ordinary skill in the art would have motivation to combine Li, Onishi, Balasubramanian, and Putrevu by taking the medium of claim 1, including a data object with associated data, jointly taught by Li, Onishi, and Balasubramanian, and adding using a machine learning model to predict distance values based on a set of data associated with the data object, taught by Putrevu, as Putrevu teaches: (Putrevu [0003]) “many conventional online concierge systems account for distances for a shopper to travel to fulfill different orders to minimize a total distance traveled by the shopper”, that is, that predicting distance values provides the predictable benefit of allowing for the use of a minimal distance, increasing efficiency of movement or transportation. Such a combination would be obvious.
Regarding claim 9,
Claim 9 recites a method for performing the function of the medium of claim 2. All other limitations in claim 9 are substantially the same as those in claim 2, therefore the same rationale for rejection applies.
Regarding claim 16,
Claim 16 recites a system for performing the function of the medium of claim 2. All other limitations in claim 16 are substantially the same as those in claim 2, therefore the same rationale for rejection applies.
Claims 3, 6, 10, 13, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Li in view of Onishi, further in view of Balasubramanian, further in view of Putrevu, further in view of Dong.
Regarding claim 3,
Li, Onishi, Balasubramanian, and Putrevu jointly teach The non-transitory machine-readable medium of claim 2,
Li further teaches:
wherein retrieving the set of machine learning models further comprises, upon determining that the subcategory of the defined category associated with the data object is a second subcategory in the plurality of defined subcategories, ((Li [0042]) “The machine-learning model may be generated using one or more machine-learning algorithms. For example, the data set can include one or more events (e.g., an incurred expense, flight reservation, booked hotel, paid meal, etc.) that have previously occurred. Each event of the one or more events can include or correspond to one or more event parameters that identify a characteristic of the event. Examples of an event parameter can include a location of the event (e.g., destination city of a flight), a vendor name associated with the event (e.g., restaurant name), a date and time of the event (e.g., time of the expense), distance of an event (e.g., travel distance when travelling by train or bus), and other suitable parameters of an event”, event parameters of events with types correspond to subcategories of defined categories, several event parameters are defined such as destination city, time, and distance, some varying by means of transportation such as by plane, train, or bus)
Putrevu further teaches:
a third machine learning model configured to predict distance values based on the third set of data of the data object in the set of machine learning models, ((Putrevu [0010]) “Alternatively, the online concierge system generates the trained distance prediction model as a tree based ensemble model combining multiple decision trees”, a tree based ensemble model combining multiple decision trees includes several machine learning models)
and a fourth machine learning model configured to predict distance values based on the third set of data of the data object ((Putrevu [0010]) “Alternatively, the online concierge system generates the trained distance prediction model as a tree based ensemble model combining multiple decision trees”, a tree based ensemble model combining multiple decision trees includes several machine learning models)
Dong teaches the following further limitation that neither Li, nor Onishi, nor Balasubramanian, nor Putrevu teaches:
including in the set of machine learning models a second machine learning model configured to predict a type of the second subcategory associated with the data object based on a subset of the third set of data associated with the data object, ((Dong Abstract) “we propose a vehicle type classification method using a semisupervised convolutional neural network from vehicle frontal-view images…For a given vehicle image, the network can provide the probability of each type to which the vehicle belongs. Unlike traditional methods by using handcrafted visual features, our method is able to automatically learn good features for the classification task”, an image of a vehicle is a data object, features of an image correspond to a subset of a set of data associated with a data object)
At the time of filing, one of ordinary skill in the art would have motivation to combine Li, Onishi, Balasubramanian, Putrevu, and Dong by taking the medium of claim 2, including a data object with associated data, taught jointly by Li, Onishi, Balasubramanian, and Putrevu, and using multiple machine learning models, including a second machine learning model that predicts a second vehicle subcategory associated with a data object (paragraph [0025] of the specification states “Transport data objects can be associated with an airfare subcategory, a ground transport subcategory, or a car rental subcategory”) based on a subset of data associated with the data object, taught by Dong, as machine learning classifiers are very well-known in the art for recognition of types of main subjects within both textual documents and images, conferring the predictable benefit of a more time and cost-effective solution than the use of human labor for equivalent tasks. Such a combination would be obvious.
Regarding claim 6,
Li, Onishi, Balasubramanian, Putrevu, and Dong jointly teach The non-transitory machine-readable medium of claim 3,
Balasubramanian further teaches:
wherein generating the first set of data values comprises: determining whether the at least one portion of the first set of data values can be determined based on the regular expression; ((Balasubramanian [0308]) “The input company name goes through the first module 606 to detect any static profanity. The module 606 marks a field IsFrivolousCompany to true or false based on regex 310 profanity comparison”)
and upon determining that the at least one portion of the first set of data values can be determined based on the regular expression, using the regular expression to determine the at least one portion of the first set of data values instead of using the third machine learning model ((Balasubramanian [0308]-[0309]) “If IsFrivolousCompany is true, the other two modules are skipped by the orchestration engine 614 and that input company name is marked as frivolous when returned by the service 604… If the field IsFrivolousCompany is false, the company name goes to the second module 610 to detect any other forms of profanity or keyboard gibberish…More details about the machine learning module 610 can be found elsewhere herein”)
At the time of filing, one of ordinary skill in the art would have motivation to combine the medium jointly taught by Li, Onishi, Balasubramanian, Putrevu, and Dong, for the parent claim of claim 6, claim 3. No new embodiments are introduced, so the reason to combine is the same as for the parent claim.
Regarding claim 10,
Claim 10 recites a method for performing the function of the medium of claim 3. All other limitations in claim 10 are substantially the same as those in claim 3, therefore the same rationale for rejection applies.
Regarding claim 17,
Claim 17 recites a system for performing the function of the medium of claim 3. All other limitations in claim 17 are substantially the same as those in claim 3, therefore the same rationale for rejection applies.
Claims 4, 5, 11, 12, 18, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Li in view of Onishi, further in view of Balasubramanian, further in view of Putrevu, further in view of Dong, further in view of Aslandere.
Regarding claim 4,
Li, Onishi, Balasubramanian, Putrevu, and Dong jointly teach The non-transitory machine-readable medium of claim 3,
Aslandere teaches the following further limitation that neither Li, nor Onishi, nor Balasubramanian, nor Putrevu, nor Dong teaches:
wherein generating the first set of data values comprises, upon determining that the predicted type of the second subcategory associated with the data object is a first type, using the third machine learning model to determine the first set of data values (Aslandere [0089] “In one example, multiple trained neural networks may exist for various vehicles including all-electric vehicles, hybrid vehicles, and combustion engine vehicles”, there are at least three types with associated neural networks listed)
At the time of filing, one of ordinary skill in the art would have motivation to combine Li, Onishi, Balasubramanian, Putrevu, Dong, and Aslandere by taking the medium of claim 3, including a data object with associated data, taught jointly by Li, Onishi, Balasubramanian, Putrevu, and Dong, and using a specific machine learning model to determine data when the data object is of a specific type, taught by Aslandere, as it is well-known within the art that, all other factors being equal, machine learning models that have been trained for specific circumstances perform better within the area they are specifically tailored for, providing the predictable benefit of greater accuracy at the expense of reduced flexibility. Such a combination would be obvious.
Regarding claim 5,
Li, Onishi, Balasubramanian, Putrevu, Dong, and Aslandere jointly teach The non-transitory machine-readable medium of claim 4,
Aslandere further teaches:
wherein generating the first set of data values further comprises: upon determining that the predicted type of the second subcategory associated with the data object is a second type, using the fourth machine learning model to determine the first set of data values (Aslandere [0089] “In one example, multiple trained neural networks may exist for various vehicles including all-electric vehicles, hybrid vehicles, and combustion engine vehicles”, there are at least three types with associated neural networks listed)
At the time of filing, one of ordinary skill in the art would have motivation to combine the medium jointly taught by Li, Onishi, Balasubramanian, Putrevu, Dong, and Aslandere for the parent claim of claim 5, claim 4. No new embodiments are introduced, so the reason to combine is the same as for the parent claim.
Regarding claims 11 and 12,
Claims 11 and 12 recite a method for performing the function of the medium of claims 4 and 5, respectively. All other limitations in claims 11 and 12 are substantially the same as those in claims 4 and 5, therefore the same rationale for rejection applies.
Regarding claim 13,
Claim 13 recites a method for performing the function of the medium of claim 6. All other limitations in claim 13 are substantially the same as those in claim 6, therefore the same rationale for rejection applies.
Regarding claims 18 and 19,
Claims 18 and 19 recite a system for performing the function of the medium of claims 4 and 5, respectively. All other limitations in claims 18 and 19 are substantially the same as those in claims 4 and 5, therefore the same rationale for rejection applies.
Claim 21 is rejected under 35 U.S.C. 103 as being unpatentable over Li in view of Onishi, further in view of Balasubramanian, further in view of Biswas.
Regarding claim 21,
Li, Onishi, and Balasubramanian jointly teach The system of claim 15, wherein the instructions further cause the at least one processing unit to:
Biswas teaches the following further limitations that neither Li, nor Onishi, nor Balasubramanian teaches:
if the at least one portion of the first set of data values is successfully extracted using the regular expression, generate a training dataset comprising the at least one portion and the data object; ((Biswas [0017]) “In an embodiment, rules engine 104 can comprise a software and/or hardware device that processes raw data and applies labels to the raw data according to one or more rules. In one embodiment, rules engine 104 can comprise a software application that matches the raw data to a plurality of regular expressions to determine labels. In such an embodiment, each regular expression may be associated with a label, and when raw text data features (e.g., a paragraph, sentence, etc.) match the regular expression, the rules engine 104 applies the corresponding label to the features and outputs a labeled example to model training unit 106”, a label corresponds to a portion of a set of data values, text corresponds to a data object)
and train at least one of the machine learning models of the set of machine learning models using the training dataset ((Biswas [0018]) “Model training unit 106 receives labeled examples from the rules engine 104 and trains an ML model using the labeled examples”)
At the time of filing, one of ordinary skill in the art would have motivation to combine Li, Onishi, Balasubramanian, and Biswas by taking the system of claim 15, including a data object a set of data values, of which a portion is attempted to be extracted using a regular expression, and a set of machine learning models, jointly taught by Li, Onishi, and Balasubramanian, and adding that, contingent on the regular expression successfully generating the portion, generating a training dataset containing the portion of data values and the data object the data values are derived from to a training dataset, and training a model from the set of models using it, taught by Biswas, as doing so imparts the predictable benefit of generating labels for training data in a cheaper and faster fashion than can be achieved by manual human labelling, allowing for cheaper training of a machine learning model. Such a combination would be obvious.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Goodsitt et al. (U.S. Patent Application Publication No. 2021/0192282) teaches a method of tagging and indexing a dataset by applying a series of nodes to generate a data structure.
De Nunzio et al. (U.S. Patent Application Publication No. 2022/0335822) teaches a method of monitoring the pollutant emissions from at least one vehicle over a section of road using a pollutant emission machine learning model.
Sperling et al. (U.S. Patent Application Publication No. 2009/0292617) teaches a method of calculating the carbon emissions from shipping a purchased product based on information such as type of transportation and distance.
Wenzel (U.S. Patent Application Publication No. 2010/0305839) teaches a method for determining emissions values for vehicles or subgroups of vehicles.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to VICTOR A NAULT whose telephone number is (703) 756-5745. The examiner can normally be reached M - F, 12 - 8.
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, Miranda Huang can be reached at (571) 270-7092. 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.
/V.A.N./Examiner, Art Unit 2124
/Kevin W Figueroa/Primary Examiner, Art Unit 2124
Prosecution Insights