DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Status of the Claims
Claims 1-20 were previously pending and subject to a non-final rejection dated April 8, 2026. In the Response, submitted on September 3, 2026, claims 1, 3, 15, 17 and 19-20 were amended. Therefore, claims 1-20 are currently pending and subject to the following final rejection.
Response to Arguments
Applicant’s Remarks on Pages 19-30 of the Response, regarding the previous rejections of the claims under 35 U.S.C. 101 have been fully considered but are not found persuasive.
On Pages 20-21 of the Response, in discussing Step 2A, Prong One, Applicant states “The amended claim does not require entering a contract, conducting a sale, processing payment, setting a price, advertising a product, or carrying out a commercial negotiation. Nor do the newly added limitations describe an interaction between people. Instead, they define internal computer operations involving: generation of electronic records having a specified schema; initialization of a computer control value from specified record fields; iterative training and updating of a machine learning model; feedback-based modification of an operative rule set; conditional insertion and deletion of records in an in-memory data structure; composite key-based record retention; and self-joining pre-validated records while avoiding repeated validation operations. The fact that these operations ultimately contribute to airline demand prediction does not convert every intervening computer operation into a commercial interaction. A claim is not properly reduced to its field of use or ultimate business utility while disregarding the particular technical implementation recited in the body of the claim.”
Examiner respectfully disagrees and notes that the test for subject matter eligibility requires determining whether the identified limitations fall within at least one of the enumerated groupings (mathematical concepts; certain methods of organizing human activity; and mental processes). The “phrase ‘methods of organizing human activity’ is used to describe concepts relating to:… commercial or legal interactions…” and “determination should be based on whether the activity itself falls within one of the sub-groupings.” (See MPEP 2106.04(a)(2)(II)). The sub-grouping of commercial or legal interactions includes “… sales activities or behaviors”. Therefore, the claims need not “require entering a contract, conducting a sale, processing payment, setting a price, advertising a product, or carrying out a commercial negotiation” to fall under a commercial interaction. Here, predicting demand at an airline level for future flight path for procurement of an appropriate number and type of aircraft for: determining flight share information for a target segment of location pair, determining agency gap values at the airline level, and predicting at least one of a number of seats expected for the target segment of location pair for a target airline and a corresponding aircraft type for the target segment of location pair, recites a commercial interaction, such as a sales behavior. Furthermore, Examiner notes the specification explains that the “disclosure generally relates to…predicted fare market share per airline per airline….to meet expected demand.” Therefore, under broadest reasonable interpretation, and in light of the specification, the claim limitations of predicting demand at an airline level for a future flight path for procurement of an appropriate number and type of aircraft (as discussed in further detail in the below amended rejection) recite a certain method of organizing human activity, such as a commercial interaction. See In re Maucorps, 609 F.2d 481, 485, 203 USPQ 812, 816 (CCPA 1979) (using an algorithm for determining the optimal number of visits by a business representative to a client) for reciting a commercial interaction.
Examiner further notes that the “internal computer operations” listed merely include instructions to implement an abstract idea on a computer, or merely use a computer as a tool to perform the abstract idea. See MPEP 2106.05(f). Thus, Applicant’s “technical implementation recited in the body of the claim” does not take the claims out of the grouping of a certain method of organizing human activity.
On Page 21 of the Response, in arguing that the claim integrates any alleged exception into a practical application, Applicant states “[t]he amended claim recites a particular machine learning-driven connection-building and record-management architecture. The processor and machine-learning model do not merely calculate demand more quickly. They generate and modify operative controls that determine the state and contents of a working data structure and the subsequent processing performed on its records.”
Examiner respectfully disagrees and notes that the “machine learning-driven connection-building and record-management architecture” is recited at a high-level of generality, such that when viewed individually and as a whole, the machine-learning additional elements merely include instructions to implement an abstract idea on a computer, or merely use a computer as a tool to perform the abstract idea. The generated “operative controls that determine the state and contents of a working data structure and the subsequent processing performed on its records” merely reflects including instructions to implement an abstract idea on a computer, or merely use a computer as a tool to perform the abstract idea.
On Page 21 of the Response, Applicant further argues “Each connection record includes a first domestic or international indicator and a second domestic-or-international indicator. The processor uses those two indicators to initialize the variable upper limit value to: a first predetermined value for a domestic/domestic connection; a greater second predetermined value for a mixed domestic/international connection; or a still greater third predetermined value for an international/international connection. The claim therefore identifies the input fields, the categories represented by those fields, the processing relationship, and the resulting ordered values. It does not merely state that an unspecified limit is selected or adjusted. These limitations reflect the maximum connect time architecture described in Spec…”
While Examiner agrees that the amended claims provide more details regarding ordered value limits, the amendments, however, merely further describe the abstract idea. Specifically, initializing the variable upper limit values to: a first value when the flights are domestic flights, a second value when one flight is domestic and another flight is international, and a third value when the flights are international - merely narrows the previously discussed abstract idea.
On Page 22 of the Response, Applicant further argues “requires training the model over multiple iterations using historical flight schedule data and then updating the trained model based on additional flight schedule data. The model update includes performing data pattern detection, determining an adjustment to the variable upper limit value based on the detected pattern and identified connection record fields, and applying that adjustment. The update also includes applying exception rules to identify invalid connections, recognizing the particular exception rule that produces an invalid connection, and autonomously removing that rule. Those limitations define a feedback mechanism. An invalid output is not merely reported. It is used to identify and remove an operative rule, thereby changing the rule set used in subsequent processing.”
Examiner respectfully disagrees and notes “training the model using…data and then updating the trained model based on additional…data” merely reflects including instructions to implement an abstract idea on a computer (i.e., machine learning model), or merely use a computer (i.e., a machine learning model) as a tool to perform the abstract idea.
“[P]erforming data pattern detection, determining an adjustment to the variable upper limit value based on the detected pattern and identified connection record fields, and applying that adjustment. The update also includes applying exception rules to identify invalid connections, recognizing the particular exception rule that produces an invalid connection, and… removing that rule. Those limitations define a feedback mechanism. An invalid output is … used to identify and remove an operative rule, thereby changing the rule set used in subsequent processing” merely reflect the abstract idea. That the rule is removed “autonomously”, merely reflects including instructions to implement the abstract idea on a computer.
On Page 22 of the Response, Applicant further argues “The machine learning output is not merely displayed or used as advisory information. The trained and updated model generates an updated upper limit value and an updated exception rule set. Those model outputs directly determine whether a connection record is: inserted into the connection table; excluded from the connection table; retained as the shortest-dwell record among records sharing the same composite key; or deleted after the model is further updated using additional flight schedule data. This establishes the nexus requested during the interview. The model training and updating are functionally connected to the remaining claim limitations because the resulting model outputs change the actual contents of the working data structure and the records made available for subsequent processor operations. The claim recites the following closed-loop sequence:
PNG
media_image1.png
100
635
media_image1.png
Greyscale
PNG
media_image2.png
60
608
media_image2.png
Greyscale
That is not a generic instruction to apply machine learning to demand forecasting. It is a defined control loop governing the state of a particular data structure.”
Examiner disagrees and notes “generat[ing] an updated upper limit value and an updated exception rule set” such that “the resulting model outputs change the actual contents of the working data structure and the records made available for subsequent processor operations” reflects the abstract idea. That the data structure is “a defined control loop” does not provide a technical improvement to the machine learning model itself. See Recentive Analytics, Inc. v. Fox Corp., 134 F.4th 1205 (Fed. Cir. 2025). In that case, similar to here, “[t]he requirements that the machine learning model be ‘iteratively trained’ or dynamically adjusted in the Machine Learning Training patents do not represent a technological improvement” because “[i|terative training using selected training material and dynamic adjustments based on real-time changes are incident to the very nature of machine learning.” Id. at 1212. Thus, Applicant’s arguments are not found persuasive.
On Page 23 of the Response, Applicant further argues “The amended claim further requires building double connections by self-joining stored single connection records. The second leg of a first stored single connection record must overlap the first leg of a second stored single connection record. Importantly, this self-join is performed without repeating air service restriction, exception rule, or total connecting time validations previously performed for the stored single connection records. The resulting improvement is not merely that a general purpose computer executes an existing process faster. The claim changes the processing architecture itself. Validation results obtained during construction of the single connections are preserved through the stored records and reused during construction of the double connections. The processor therefore does not reconstruct double connections from unvalidated source data and rerun the same validation operations. Spec…describe constructing double connections by joining single connection results with themselves and avoiding replication of validations already performed during the single connection process. The amended claim now incorporates that disclosed validation reuse architecture.”
Examiner respectfully disagrees and notes that reusing validation results during construction of the double connections, is not a technical improvement to the connection table data structure itself. That the “process runs faster than the single connection process” is a result of using stored records (previously validated), and not a result of a technical improvement in the data structure. In contrast, See Enfish, LLC v. Microsoft Corp., 822 F.3d 1327, 118 USPQ2d 1684 (Fed. Cir. 2016), where the claims recited a self-referential database having two key features: all entity types can be stored in a single table; and the table rows can contain information defining the table columns. Enfish, 822 F.3d at 1332, 118 USPQ2d at 1687. The features enabled the claimed table to achieve benefits over conventional databases, such as increased flexibility, faster search times, and smaller memory requirements. Enfish, 822 F.3d at 1337, 118 USPQ2d at 1690.
Here, the claims’ connection table data structure recites no technical improvements in the structural features (resulting in increased flexibility, faster search times, and smaller memory requirements). Rather, the “process runs faster” merely because of the data being processed (self-joining of stored single connection records). Thus, Applicant’s arguments are not found persuasive.
On Pages 23-24 Applicant further argues “The Office Action found the prior arguments unpersuasive because a reduction in memory and processing load resulting merely from processing less data does not necessarily establish an improvement in computer functionality. The amended claim does not rest solely on the proposition that fewer records are processed. It recites the architecture that changes how records are generated, validated, selected, maintained, and reused: the connection records have defined fields and composite keys; the initial limit is derived through a specified categorical mapping; model updating uses detected patterns and invalid connection feedback; the model modifies the operative limit and exception rule set; those controls determine record admission, exclusion, retention, and deletion; and pre-validated records are self-joined without repeated validation. Thus, any reduction in storage or processor load is produced by the claimed record structure, feedback control mechanism, and validation reuse process-not solely by the inherent speed or capacity of a general purpose computer.”
Examiner respectfully disagrees and notes “chang[ing] how records are generated, validated, selected, maintained, and reused: the connection records have defined fields and composite keys; the initial limit is derived through a specified categorical mapping; model updating uses detected patterns and invalid connection feedback; the model modifies the operative limit and exception rule set; those controls determine record admission, exclusion, retention, and deletion; and pre-validated records are self-joined without repeated validation” reflect the abstract idea. As discussed above, that the “process runs faster than the single connection process” is a result of using stored records (previously validated), and not a result of a technical improvement in the data structure.
On Page 24 of the Response, in discussing Example 39, Applicant argues “Example 39 is…’instructive because it confirms that a specific machine learning feedback process should not be abstracted into the generalized phrase ‘training model.’ Amended claim 1 likewise recites more than generic training. It requires identifying invalid connection results, recognizing the particular exception rule that produces an invalid connection, autonomously removing that rule, and producing an updated rule set that governs subsequent processing. The eligibility-relevant parallel is: in Example 39, erroneous classification results are incorporated into a subsequent training stage; and in amended claim 1, invalid connection results are used to identify and remove an operative exception rule before the updated rule set controls subsequent connection table operations. The claimed feedback and update limitations therefore should be evaluated at their recited level of specificity rather than collapsed into the generalized concept of machine learning.”
Examiner respectfully disagrees and notes Example 39 is eligible because the claim does not recite any of the enumerated judicial exceptions. Unlike Example 39, the current claims do recite an abstract idea. Therefore, Applicant’s arguments are not analogous and therefore not persuasive.
On Pages 25-26 of the Response, in discussing Example 40, Applicant argues “The Office Action distinguished the previously pending claim from Example 40 because the prior claim did not recite a comparably specific condition-responsive mechanism. Amended claim 1 now does. In particular, amended claim 1: initializes different upper limit values based on specified record indicators; adaptively adjusts the upper limit value and exception rule set based on additional flight schedule data and invalid connection feedback; inserts records only when the records comply with the updated controls; excludes or deletes records that fail those controls; retains only one shortest dwell record for a shared composite key; and repeats the model update and table modification process when further schedule data are received. As in Example 40, defined and adaptively updated conditions control whether data are retained in the working computer system. The conditional operations are not ancillary to the claimed demand prediction. They determine the structure and content of the connection dataset on which the subsequent processor operations are performed. Amended claim 1 additionally recites a validation reuse self-join architecture, further distinguishing the amended claim from generic filtering or generic data reduction.”
Examiner respectfully disagrees and notes eligible claim 1 of Example 40, limits collection of additional Netflow protocol data to when the initially collected data reflects an abnormal condition, which avoids excess traffic volume on the network and hindrance of network performance. The collected data can then be used to analyze the cause of the abnormal condition. This provides a specific improvement over prior systems, resulting in improved network monitoring. Here, nothing in the claims or specification recites a similar technical improvement similar to the improvement to “networking monitoring”. That is, “initializes different upper limit values based on specified record indicators; adaptively adjusts the upper limit value and exception rule set based on additional flight schedule data and invalid connection feedback; inserts records only when the records comply with the updated controls; excludes or deletes records that fail those controls; retains only one shortest dwell record for a shared composite key; and repeats the model update and table modification process when further schedule data are received…claim 1 additionally recites a validation reuse self-join architecture” reflects the abstract idea, and not an improvement in the machine learning model itself or in the claimed table data structure. Thus, Applicant’s analogy to Example 40 is not found persuasive.
On Pages 26 of the Response, in discussing Example 47, Applicant argues “Amended claim 1 is analogous to eligible claim 3 rather than ineligible claim 2. The claimed model output does not merely describe a connection, identify a limit, or present a recommendation. It causes particular operations within the computerized connection building subsystem: a compliant record is inserted; a noncompliant record is excluded; only the shortest dwell record for a composite key is retained; a previously stored record is deleted following a further model update; and double connections are constructed from the resulting validated record set without repeating prior validations. These operations change the state of the connection table and control the subsequent path of processor execution. They are concrete model-responsive computer actions, not merely the communication of analytical information”
Examiner respectfully disagrees and notes Claim 3 of Example 47 recites “(c) determining at least one detected anomaly is associated with one or more malicious network packets; (d) detecting a source address associated with the one or more malicious network packets in real time; (e) dropping the one or more malicious network packets in real time; and (f) blocking future traffic from the source address.” The background explains that existing systems use various detection techniques for detecting potentially malicious network packets and can alert a network administrator to potential problems. The disclosed system detects network intrusions and takes real-time remedial actions, including dropping suspicious packets and blocking traffic from suspicious source addresses. The background section further explains that the disclosed system enhances security by acting in real time to proactively prevent network intrusions. The claimed invention reflects this improvement in the technical field of network intrusion detection. Steps (d)-(f) provide for improved network security using the information from the detection to enhance security by taking proactive measures to remediate the danger by detecting the source address associated with the potentially malicious packets. Specifically, the claim reflects the improvement in step (d), dropping potentially malicious packets in step (e), and blocking future traffic from the source address in step (f). These steps reflect the improvement described in the background. Thus, the claim as a whole integrates the judicial exception into a practical application such that the claim is not directed to the judicial exception.
Here, no technology similar to the technical field of network intrusion detection is being improved. Applicant’s claimed “particular operations” of inserting a compliant record; excluding a noncompliant record; retaining only the shortest dwell record for a composite key is retained; deleting a previously stored record is deleted following a further model update; and constructing double connections from the resulting validated record set without repeating prior validations, merely reflect the abstract idea and are not similar to the technical improvement of “dropping the one or more malicious network packets in real time; and…blocking future traffic from the source address” of Example 47, claim 3. Thus, Applicant’s analogy to Example 47, claim 3 is not found persuasive.
On Pages 26-37 of the Response, in discussing Ex Parte Desjardins, Applicant states “The relevance of Desjardins is broader and directly applicable: the eligibility analysis must evaluate the claim as a whole, must account for the improvement reflected in the actual claim limitations, and must recognize that computer-technology improvements may be implemented through logical structures and processes rather than through newly recited physical components…claim 1 now reflects its asserted improvement through particular logical structures and processes: the defined connection record schema, categorical limit initialization, iterative model update, invalid result feedback, autonomous rule removal, adaptive data structure modification, composite key retention, and validation reuse self-join. The amended claim therefore should not be evaluated at the high level of ‘using machine learning to predict airline demand.’ Its character as a whole is a specific machine learning controlled connection data processing architecture used within the broader prediction method.”
Examiner respectfully disagrees and notes in Desjardins, the specification identified the improvement to machine learning technology by explaining how the machine learning model is trained to learn new tasks while protecting knowledge about previous tasks to overcome the technical problem of “catastrophic forgetting,” and that the claims reflected the improvement identified in the specification. Here, unlike Desjardins, the claims do not recite a technical improvement to the “particular logical structures” (beyond the high level recitation of the machine learning model and table data structure) and the alleged improved “processes” (which merely describe the abstract idea). Furthermore, unlike Desjardins, Applicant’s specification (e.g., Paras. 140-147) merely discloses improvements in the data (“limit output in the connection table”, a data “process runs faster”, and “allow[ing] the code to be generated in a cleaner manner” reflects an improvement in the data itself, but not any claimed technology. Thus, Applicant’s comparison to the technical improvements in Desjardins is not found persuasive.
On Pages 27-28, in discussing Recentive Analytics, Applicant argues “Applicant does not rely on iterative training or dynamic adjustment standing alone. Nor does amended claim 1 merely apply an unspecified machine learning technique to a new category of commercial data to generate a result. The claim expressly delineates: the fields of the records processed by the model; the mapping used to initialize the operative upper limit value; the historical and additional data used for training and updating; the pattern detection and adjustment operations; the feedback condition identifying an invalid result; the identification and autonomous removal of the exception rule producing that result; the updated control outputs; the direct use of those outputs to modify a connection table; and the reuse of validated records without repeating specified validations. Accordingly, the eligibility argument is not that machine learning itself is automatically eligible or that iterative model training necessarily improves technology. The argument is that the particular claimed combination applies the machine learning update through a defined data structure and processing architecture that changes the manner in which the computer constructs, validates, stores, modifies, and reuses connection records.”
Examiner disagrees for the reasons discussed above. Delineating “the fields of the records processed by the model; the mapping used to initialize the operative upper limit value; the historical and additional data used for training and updating; the pattern detection and adjustment operations; the feedback condition identifying an invalid result; the identification and autonomous removal of the exception rule producing that result; the updated control outputs; the direct use of those outputs to modify a connection table; and the reuse of validated records without repeating specified validations” merely reflects the category of data of the abstract idea; but does not recite an improvement in the technology of the machine learning updating. Thus, Applicant’s arguments are not found persuasive.
On Page 28 of the Response, Applicant further argues “The Office Action analyzes generic processors, memory, servers, a computer model, a machine learning model, and real-time updating largely as separate components. The amended claim must instead be evaluated as a whole and in its claimed order. The operative combination is:
PNG
media_image3.png
307
610
media_image3.png
Greyscale
The practical application arises from the interaction of those limitations. The model output governs the state of the working data structure; the contents of that data structure govern the records available for subsequent processing; and the retained validation state permits double connections to be constructed without rerunning specified validation operations. This ordered combination imposes meaningful limits on any alleged abstract idea. It does not preempt airline demand prediction generally, use of QSI values generally, machine learning generally, or database filtering generally. It is confined to a particular connection record schema, parameter initialization process, feedback-based model update, table control mechanism, composite key retention rule, and validation reuse self-join architecture.”
Examiner respectfully disagrees for the reasons discussed above. The defined connection records, categorical initialization of an upper-limit value, historical-data model training, additional-data model updating, invalid-connection feedback, removal of a causal rule generation of updated model controls, controlled record insertion, exclusion, and retention, further-data model updating, dynamic deletion of stored records, self-joining of prevalidated records without repeated validation, downstream QSI, circuity, share, gap, and demand processing, reflect the abstract idea. That the records are electronic and that the removal of a causal rule is autonomous, does not integrate the abstract idea into a practical application. Thus, when viewed individually and as a whole/ordered combination, the additional elements still merely include instructions to implement an abstract idea on a computer, or merely use a computer as a tool to perform the abstract idea. See MPEP 2106.05(f). Thus, Applicant’s arguments are not found persuasive.
On Page 29 of the Response, Applicant further argues “To the extent the Office Action characterizes individual components or operations as generic or conventional, that characterization does not resolve Step 2A, Prong Two. The Prong Two inquiry asks whether the claim as a whole integrates the alleged judicial exception into a practical application. It does not ask whether each additional element, viewed separately, was well-understood, routine, and conventional. Examples 40 and 47 illustrate this distinction…Likewise, the eligibility of amended claim 1 does not depend on asserting that processors, memories, databases, self-joins, or machine learning models were individually unknown. The relevant question is whether the claimed combination applies any alleged exception through the particular feedback controlled, data structure-based, and validation reuse architecture recited in the claim.”
Examiner agrees that the Prong Two inquiry “does not ask whether each additional element, viewed separately, was well-understood, routine, and conventional”. Rather, the Step 2A, Prong Two inquiry asks “whether the claim as a whole integrates the exception into a practical application of that exception.” As discussed above, the “particular feedback” “data structure” and “validation reuse” merely reflect the abstract idea, and the claimed technology is merely includes instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP § 2106.05(f). Thus, Applicant’s arguments are not found persuasive.
On Pages 29-30 of the Response, in discussing Step 2B, Applicant argues “Even if the Office maintains that amended claim 1 is directed to an abstract idea under Step 2A, the claim recites significantly more under Step 2B when its limitations are considered individually and as an ordered combination. The relevant additional elements are not merely a generic processor, memory, network, or machine learning model. They include: a connection record having a specifically recited schema; a composite key formed from specified date, window, carrier, and airport fields; categorical initialization of three ordered upper limit values; iterative training and updating using historical and additional schedule data; invalid result feedback; recognition and autonomous removal of a causal exception rule; generation of updated model controls; model-responsive insertion, exclusion, retention, and deletion of records; further adaptive modification upon receipt of additional data; and construction of double connections through a self-join without repeating specified validations. The Office Action's conclusion that generic computer components were known does not establish that this particular ordered combination was well-understood, routine, and conventional. A conclusion of conventionality under Step 2B must be supported in accordance with MPEP §§ 2106.05(d) and 2106.07(a). See Berkheimer V. HP Inc., 881 F.3d 1360, 1368 (Fed. Cir. 2018). The Office Action does not identify evidence establishing that the newly claimed combination of adaptive model feedback, composite key record control, dynamic table modification, and validation reuse self-joining was widely prevalent or in common use. Applicant does not rely on novelty or non-obviousness as the legal test for eligibility. Nevertheless, the Office's express determination that the prior art fails to disclose, teach, or suggest all limitations of the independent claims reinforces that the present record does not establish the newly claimed ordered combination to be routine and conventional. Accordingly, amended claim 1 integrates any alleged judicial exception into a practical application under Step 2A, Prong Two and, alternatively, recites significantly more under Step 2B.”
Examiner respectfully disagrees. For the reasons discussed above, Examiner notes that “a connection record having a specifically recited schema; a composite key formed from specified date, window, carrier, and airport fields; categorical initialization of three ordered upper limit values; iterative training and updating using historical and additional schedule data; invalid result feedback; recognition and…removal of a causal exception rule; generation of updated model controls; model-responsive insertion, exclusion, retention, and deletion of records; further adaptive modification upon receipt of additional data; and construction of double connections through a self-join without repeating specified validations” merely reflects the abstract idea (with additional element of “autonomous” removal feature discussed above).
Examiner further notes it is unclear where “The Office Action's conclusion that generic computer components were known” is recited. Absent such evidence that Examiner categorized additional elements as “known”, Applicant’s arguments regarding any required Berkheimer evidence are moot. Lastly Examiner notes “lack of novelty under 35 U.S.C. 102 or obviousness under 35 U.S.C. 103 of a claimed invention does not necessarily indicate that additional elements are well-understood, routine, conventional elements. Because they are separate and distinct requirements from eligibility, patentability of the claimed invention under 35 U.S.C. 102 and 103 with respect to the prior art is neither required for, nor a guarantee of, patent eligibility under 35 U.S.C. 101.” See MPEP 2106.05 Eligibility Step 2B(I). Thus, Applicant’s arguments are not found persuasive.
On Page 30 of the Response, Applicant argues “Claims 19 and 20 are therefore eligible for at least the same reasons discussed above with respect claim 1.” As such, Examiner notes the above response to the arguments of claim 1 apply to claims 19 and 20 as well.
35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1
Claims 1-18 are directed to a method (i.e., a process), claim 19 is directed to a system comprising a processor, and claim 20 is directed to a non-transitory computer readable medium (i.e., a machine). Therefore, the claims all fall within one of the four statutory categories of invention.
Step 2A, Prong One
Claims 1, 19 and 20 recite a series of steps/functions of predicting a demand at an airline level for a future flight path for procurement of an appropriate number and type of aircraft by: acquiring and aggregating raw data; parsing the acquired raw data and identifying and aggregating a plurality of segments of location pairs; generating, for each candidate connecting itinerary including a first flight and a second flight derived from the plurality of segments of location pairs, a connection record including a first domestic or international indicator for a first flight, a second domestic or international indicator for a second flight, a total connecting time, and a composite key including a departure date, a departure window, one or more carrier identifiers, an origin airport, a connection airport, and a destination airport; generating a connection table configured to store connection records; initializing a variable upper limit value for total connecting time by processing the first domestic or international indicator and the second domestic or international indicator, wherein the variable upper limit value is initialized to: a first predetermined upper limit value when the first and second flights are domestic flights; a second predetermined upper limit value greater than the first predetermined upper limit value when one of the first and second flights is a domestic flight and the other of the first and second flights is an international flight; and a third predetermined upper limit value greater than the second predetermined upper limit value when the first and second flights are international flights; training a model over a plurality of iterations based on historical flight schedule data; updating the model based on additional flight schedule data wherein updating the model includes: performing data pattern detection on the additional flight schedule data; determining, based on a data pattern detected in at least the additional flight schedule data and based on the first domestic or international indicator, the second domestic or international indicator, and the total connecting time of the connection records, an adjustment to the variable upper limit value; adjusting the variable upper limit value according to the determined adjustment; applying a set of exception rules to the connection records to identify invalid connections; recognizing, based on the identified invalid connections, an exception rule that results in creation of an invalid connection; and removing the recognized exception rule from the set of exception rules; generating a model output including an updated variable upper limit value and an updated set of exception rules, wherein the updated set of exception rules includes a rule that, for connection records having the same composite key, retains only one connection record having a shortest dwell time; building, via a model, a plurality of single connections based on the aggregated plurality of segments of location pairs, wherein each of the plurality of segments of location pairs is serviced by an aircraft of at least one of a plurality of airlines, wherein building the plurality of single connections includes: inserting a connection record into the connection table data structure only when the connection record complies with one or more air service restriction limits and the updated set of exception rules and has a total connecting time of no more than the updated variable upper limit value; excluding the connection record from the connection table data structure when the connection record fails an air service restriction limit, fails an exception rule of the updated set of exception rules, or has a total connecting time greater than the updated variable upper limit value; and for connection records having the same composite key, retaining in the connection table data structure only the connection record having the shortest dwell time; in response to receiving further flight schedule data, further updating the model based on the further flight schedule data to generate a further updated variable upper limit value and a further updated set of exception rules, and modifying the connection table data structure by deleting any stored connection record having a total connecting time greater than the further updated variable upper limit value or failing an exception rule of the further updated set of exception rules; building a plurality of double connections by self-joining stored single connection records in the connection table data structure such that a second leg of a first stored single connection record overlaps a first leg of a second stored single connection record, without repeating air service restriction, exception rule, or total connecting time validations previously performed for the first and second stored single connection records; for each of the plurality of segments of location pairs, generating a quality of service index (QSI) coefficient for each of the plurality of airlines; determining a connection window for one or more connection flights based on the aggregated plurality of segments of location pairs, wherein the connection window includes minimum and maximum connect times with categorized dwell time penalties within a predetermined range applied based on total connection time ranges; generating for each of the plurality of segments of location pairs, a circuity curve based on a distance of a segment of a location pair, wherein the circuitry curve implements a three-tier penalty system with differing factors applied based on distance-based circuity thresholds wherein itineraries above a highest circuitry curve are removed from the connection table data structure when stored in the connection table data structure and and excluded from further storage and downstream evaluation; determining flight share information at an airline level for a target segment of location pair based on a corresponding QSI coefficient and a corresponding circuity curve; determining agency gap values at the airline level for the target segment of location pair based on the flight share information; and updating the model based on the flight share information and the agency gap values for predicting at least one of a number of seats expected for the target segment of location pair for a target airline and a corresponding aircraft type for the target segment of location pair.”
The claims as a whole (predicting demand at an airline level for a future flight path for procurement of an appropriate number and type of an aircraft) recite a certain method of organizing human activity. The limitations recited above– (under broadest reasonable interpretation and in light of the specification, e.g., Para. 1) recite the abstract idea of a certain method of organizing human activity, (e.g., commercial interaction such as a sales activity or behaviors). Therefore, the claims recite an abstract idea.
The mere recitation of generic computer components ((i) a communication network (claims 1, 19, and 20), (ii) one or more servers (claims 1, 19, and 20), (iii) a processor operably coupled to a memory; memory (claims 1 and 19), (iv) a computer model (claims 1, 19, and 20), (v) a non-transitory computer readable storage medium that stores a computer program that is executed by a processor (claim 20), (vi) machine learning model (that is trained and updated) (claims 1, 19, and 20), (vii) processor evaluation (claims 1, 19, and 20), and (viii) autonomously removing; automatically removing; and dynamically modifying functions (claims 1, 19, and 20)) recited at a high-level of generality, does not take the claims out of the certain methods of organizing human activity grouping. Thus, the claims recite an abstract idea.
Step 2A, Prong Two
The judicial exception is not integrated into a practical application. Claims 1, 19 and 20 as a whole amount to: merely including instructions to implement an abstract idea on a computer, or merely using a computer as a tool to perform an abstract idea. See MPEP 2106.05(f).
The claims recite the additional elements of: (i) a communication network (claims 1, 19, and 20), (ii) one or more servers (claims 1, 19, and 20), (iii) a processor operably coupled to a memory; memory (claims 1 and 19), (iv) a computer model (claims 1, 19, and 20), (v) a non-transitory computer readable storage medium that stores a computer program that is executed by a processor (claim 20), (vi) machine learning model (that is trained and updated) (claims 1, 19, and 20), (vii) processor evaluation (claims 1, 19, and 20), and (viii) autonomously removing; automatically removing; and dynamically modifying functions (claims 1, 19, and 20))
The above additional elements are recited at a high-level of generality such that, when viewed as whole/ordered combination, they amount to no more than merely including instructions to implement an abstract idea on a computer, or merely using a computer as a tool to perform an abstract idea. (See MPEP 2106.05(f)).
Accordingly, these additional elements, when viewed individually and as a whole/ordered combination (See Fig. 1), do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. Thus, the claims are directed to an abstract idea.
Step 2B
As discussed above with respect to Step 2A Prong Two, the additional elements amount to no more than merely including instructions to implement an abstract idea on a computer, or merely using a computer as a tool to perform an abstract idea. See MPEP 2106.05(f). The same analysis applies here in 2B, i.e., merely including instructions to implement an abstract idea on a computer, or merely using a computer as a tool to perform an abstract idea, do not integrate the abstract idea into a practical application at Step 2A or provide an inventive concept at Step 2B.
Therefore, the additional elements discussed above do not integrate the abstract idea into a practical application at Step 2A or provide an inventive concept at Step 2B. Thus, even when viewed individually and as a whole/ordered combination, nothing in the claims add significantly more (i.e., an inventive concept) to the abstract idea. Thus, the claims are ineligible.
Claims 3-18 recite details in the claim limitations which merely narrow the previously recited abstract idea limitiaitions. For these reasons, described above with respect to claim 1, these judicial exceptions, when viewed as a whole/ordered combination, are not meaningfully integrated into a practical application or significantly more than the abstract idea. Thus, claims 3-18 are ineligible.
Dependent claim 2 describes details of transmitting, to a target airline, at least one of the number of seats expected for the target segment of location pair and the corresponding type of aircraft for the target segment of location pair; and assigning the corresponding type of aircraft for the target segment of location pair, and thus further describe the abstract idea. The claim adds the additional element of a computer of the target airline, which, when viewed as a whole/ordered combination, fails to integrate the abstract idea into a practical application or amount to significantly more because the computer merely includes instructions to implement an abstract idea on a computer, or merely uses the computer as a tool to perform an abstract idea. See MPEP 2106.05(f).
Allowable over the Prior Art
Claims 1-20 are allowable over the prior art, because the prior art fails to disclose, teach, or suggest, all the limitations of the independent claims in their entirety. However, the claims are subject to the above rejections under 35 U.S.C. 101.
Prior Art
The closest prior art includes:
“The Route Network Development Problem based on QSI Models” by Idrissi et al., dated 2017, (hereinafter “Idrissi”). Idrissi discloses that airlines have to choose flight schedules by considering demand, passengers preferences and competitors. The problem of allocating a new flight involves the route network development, and consists to determine a set of (OriginDestination) pairs to serve and then choose flight schedules with respect to the Quality of Service Index (QSI).
U.S. Patent Application Publication No. 2015/0371245 to Bental et al. (hereinafter “Bental”). Bental discloses QSI is a metric that quantifies the value of travel itineraries to passengers for a given carrier in a given market and is readily known in the airline industry. QSI may be derived from various factors including but not limited to number of stops by the carrier between an origin and a destination, the type of aircraft (e.g., widebody jet, narrowbody jet, turboprop, etc.), flight frequency (how many different flights in a day), travel time, and time-of-day (during business hours v. outside of business hours).
U.S. Patent No. 12,248,897 to Deng et al. (hereinafter “Deng”). Deng discloses a deep learning-based demand forecasting system for revenue management systems (flights, airlines).
U.S. Patent Application Publication No. 2015/0149234 to Rasheed et al. (hereinafter “Rasheed”). Rasheed discloses obtaining market information for the airline that defines at least one market for the airline, determining a plurality of aircraft types (priority groupings) for the airline within the at least one market to define an airline fleet model, and determining deployment priorities for the plurality of aircraft types within the at least market.
U.S. Patent Application Publication No. 2022/0230108 to Bollapragada et al. (hereinafter “Bollapragada”). Bollapragada discloses providing an airline recovery scheduling solution that reduces airline cost and improves crew and passenger satisfaction that would benefit airlines and passengers alike when such disruptions occur.
“The airline seat capacity allocation problem: An expected marginal profit approach” by Abdelghany et al., dated 2023 (hereinafter “Abdelghany”). Abdelghany discloses establishing optimal flight capacity allocation per fleet type, with the ultimate goal of maximizing the overall profitability of the airline's daily schedule.
“Connections Analysis FAQs” by OAG Analytics, dated September, 2017 (hereinafter “OAG Analytics”). OAG Analytics discloses that circuity is a term which reflects how far from the Great Circle Distance between the origin and destination is the actual routing when two or more flights connect. A large circuity value applies where the hub airport at which two flights connect is further from the great circle route between the origin and destination of the route.
U.S. Patent Application Publication No. 2020/0167324 to Sun et al. (hereinafter “Sun”). Sun discloses analyzing the relationship of the same data stored as different data types (i.e., as redundant data) and de-clutters tables of redundant data (e.g., removes “USA” from Table 1 and Table 3).
The following is prior art not cited but considered relevant:
"Quality of Service Index (QSI)" by Airport Council International, dated January, 2018 (hereinafter “Airport Council International”). Airport Council International discloses developing logical assumptions to account for share “premium” or “gap” variances, and the make adjustments to QSI weighting factors.
U.S. Patent No. 11,948,111 to Deng et al. (hereinafter “Deng”). Deng discloses a method of training a neural network to approximate a forecasting error of a passenger-demand forecasting model.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Rupangini Singh whose telephone number is (571)270-0192. The examiner can normally be reached on Monday - Friday 9:30 AM - 6:30 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, Shannon Campbell can be reached on 571-272-5587. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/RUPANGINI SINGH/
Examiner, Art Unit 3628
