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 .
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claim(s) 1 – 20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e. an abstract idea) without significantly more.
In step 1, of the 101-analysis set forth in the MPEP 2106, the examiner has determined
that the following limitations recite a process that, under the broadest reasonable interpretation, falls within one or more statutory categories (processes).
In step 2A prong 1, of the 101-analysis set forth in MPEP 2106, the examiner has determined
that the following limitations recite a process that, under broadest reasonable interpretation, covers
a mental process but for the recitation of generic computer components:
Regarding claim(s) 1 and analogous claim 11:
generating a first representation of the information in the first database;
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves observing information stored, evaluating the information to determine what aspects should be included to form a representation. See (MPEP 2106.04)).
generating a second representation of the information in the second database;
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves observing information stored, evaluating the information to determine what aspects should be included to form a representation. See (MPEP 2106.04)).
generating, [ ], a first information vector from the first representation;
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves observing the previously formed representation, evaluating which element of feature should be selected to produce a corresponding information vector. See (MPEP 2106.04)).
generating, [ ], a second information vector from the second representation;
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves observing the previously formed representation, evaluating which element of feature should be selected to produce a corresponding information vector. See (MPEP 2106.04)).
detecting a trigger event relating to one of the plurality of digital applications;
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves observing the state and behaviors of applications, evaluating whether a particular condition is has occurred and making a judgement that such a condition constitutes a trigger event. See (MPEP 2106.04)).
generating a third representation of the information relating to the trigger event;
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves forming a new representation for a received information related to a trigger event. See (MPEP 2106.04)).
generating, [ ], a third information vector from the third representation;
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves observing the previously formed representation, evaluating which element of feature should be selected to produce a corresponding information vector. See (MPEP 2106.04)).
If the claim limitations, under their broadest reasonable interpretation, covers performance of the limitations as a mental process, but for the recitation of generic computer components, then it falls within the mental process. Accordingly, the claim recites an abstract idea.
Step 2A Prong 2 of the 101-analysis, set forth in MPEP 2106, the examiner has determined that
the following additional elements do not integrate this judicial exception into a practical application:
As evaluated below:
The preamble is deemed insufficient to transform the judicial exception to a patentable invention to a patentable invention because the preamble generally links the use of a judicial exception to a particular technological environment or field of use, see MPEP 2106.05(h).
via an embedding algorithm
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation which does not amount to more than a recitation of the words "apply it" (or an equivalent), such as mere instructions to implement an abstract idea on a computer. See MPEP 2106.05(f)).
storing, in a first database, information relating to a plurality of digital applications;
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity, See MPEP (2106.05(g))).
storing, in a second database, information relating to historical performance issues associated with the plurality of digital applications;
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity, See MPEP (2106.05(g))).
training the hybrid neural network, wherein said training comprises:
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation which does not amount to more than a recitation of the words "apply it" (or an equivalent), such as mere instructions to implement an abstract idea on a computer. See MPEP 2106.05(f)).
feeding, using a masking algorithm, the first and the second information vectors to a transformer block neural network; and
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity, See MPEP (2106.05(g))).
feeding the second representation to a natural language processing (NLP) engine that is separate from the transformer block neural network;
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity, See MPEP (2106.05(g)))
receiving information relating to the trigger event
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity, See MPEP (2106.05(g))).
feeding the third information vector to the transformer block neural network;
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity, See MPEP (2106.05(g))).
receiving a first output from the transformer block neural network;
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity, See MPEP (2106.05(g))).
feeding the third representation to the NLP engine;
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity, See MPEP (2106.05(g))).
receiving a second output from the NLP engine;
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity, See MPEP (2106.05(g))).
feeding the first and the second outputs to a neural collaborative filtering engine; and
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity, See MPEP (2106.05(g))).
receiving, as output from the neural collaborative filtering engine, a set of predicted application failures.
(i.e.: deemed insufficient to transform the judicial exception to a patentable invention because the claim recites limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity, See MPEP (2106.05(g))).
In Step 2B of the 101-analysis set forth in the 2019 PEG, the examiner has determined that the
claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception:
Regarding limitation (I and IV), recite mere application of the abstract idea or mere instructions to implement an abstract idea on a computer are deemed insufficient to transform the judicial exception to a patentable invention because the limitations generally apply the use of a generic computer and/or process with the judicial exception, see MPEP
2106.05(f).
Regarding limitation (V, VI, VII, VIII, IX, X, XI, XII and XIII), additional elements considered extra/post solution activity, as analyzed above, are activity that are well-understood routine and conventional, specifically: the courts have recognized the computer functions as well‐understood, routine, and conventional functions.
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TL| Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
Regarding limitation (II and III), additional elements considered extra/post solution activity, as analyzed above, are activity that are well-understood routine and conventional, specifically: the courts have recognized the computer functions as well‐understood, routine, and conventional functions.
Storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93;
As analyzed above, the additional elements, analyzed above, do not integrate the noted judicial exception into a practical application because they do not impose any meaningful limits on practicing the abstract idea. Therefore, the claim is directed to an abstract idea.
Regarding claim 2, dependent upon claim 1, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
training the hybrid neural network to create relational associations within the plurality of digital applications, within the historical performance issues, and between the historical performance issues and the plurality of digital applications.
Deemed insufficient to transform the judicial exception to a patentable invention because the limitation is directed to mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea and are considered to adding the words “apply it” (or an equivalent) with the judicial exception, See MPEP 2106.05(f).
Limitations directed to using the computer as a tool for implementing an abstract idea cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B.
4. Claim 12, recite similar subject matter as claim 2, so is rejected under the same rationale.
Regarding claim 3, dependent upon claim 1, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
calculating, for each of the set of predicted application failures, a probability score; and
(i.e.: the broadest reasonable interpretation, the claim recites abstract idea: mental process: It involves evaluating predicted failures and assigning probability values, which is estimating likelihood, assessing risks or performing calculation. See (MPEP 2106.04)).
transmitting an alert to a system supervisor when the probability score of a predicted application failure exceeds a predetermined threshold probability score.
The recitation in the additional limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity and well-understood routine and conventional (2106.05(d)).
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TL| Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
The additional limitations as analyze failed to integrate a judicial exception into a practical application at Step 2A and provide an inventive concept in Step 2B, per the analysis above.
Claim 13, recite similar subject matter as claim 3, so is rejected under the same rationale.
Regarding claim 4, dependent upon claim 1, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
storing, in the second database, historical solution data, said historical solution data comprising information relating to actions that successfully resolved the historical performance issues;
The recitation in the additional limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity and well-understood routine and conventional (2106.05(d)).
Storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93.
The additional limitations as analyze failed to integrate a judicial exception into a practical application at Step 2A and provide an inventive concept in Step 2B, per the analysis above.
as part of the training, feeding the hybrid neural network with the historical solution data;
The recitation in the additional limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity and well-understood routine and conventional (2106.05(d)).
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TL| Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
The additional limitations as analyze failed to integrate a judicial exception into a practical application at Step 2A and provide an inventive concept in Step 2B, per the analysis above.
and receiving, as part of the output from the neural collaborative filtering engine, for each of the set of predicted application failures, a recommended action to resolve the predicted application failure.
The recitation in the additional limitation directed to mere data gathering as deemed insufficient to transform the judicial exception because claimed elements are considered insignificant extra-solution activity and well-understood routine and conventional (2106.05(d)).
Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TL| Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network). See MPEP 2106.05(d)(II).
The additional limitations as analyze failed to integrate a judicial exception into a practical application at Step 2A and provide an inventive concept in Step 2B, per the analysis above.
Claim 14, recite similar subject matter as claim 4, so is rejected under the same rationale.
Regarding claim 5, dependent upon claim 1, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
wherein: the output from the neural collaborative filtering engine further comprises a probability score for each of the predicted application failures; and the method further comprises providing the set of predicted application failures as rows in a table, wherein the rows are ordered according to the probability scores.
The recitation in the additional limitation simply links the judicial exception to a field of use and/or technology environment, see MPEP 2106.05(h).
Limitations directed to field of use cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B.
Claim 15, recite similar subject matter as claim 5, so is rejected under the same rationale.
Regarding claim 6, dependent upon claim 5, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
wherein: the output from the neural collaborative filtering engine further comprises a plurality of parameters for each of the predicted application failures, the parameters comprising: an application identifier; an issue identifier; an indicator representing a source of the failure; and a predicted time of the failure; and the table comprises a plurality of columns for representing the plurality of parameters.
The recitation in the additional limitation simply links the judicial exception to a field of use and/or technology environment, see MPEP 2106.05(h).
Limitations directed to field of use cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B.
Claim 16, recite similar subject matter as claim 6, so is rejected under the same rationale.
Regarding claim 7, dependent upon claim 6, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
wherein the indicator also represents whether the source of the failure is a human or a non-human source.
The recitation in the additional limitation simply links the judicial exception to a field of use and/or technology environment, see MPEP 2106.05(h).
Limitations directed to field of use cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B.
Claim 17, recite similar subject matter as claim 7, so is rejected under the same rationale.
Regarding claim 8, dependent upon claim 1, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
wherein the NLP engine comprises a SpaCy engine.
The recitation in the additional limitation simply links the judicial exception to a field of use and/or technology environment, see MPEP 2106.05(h).
Limitations directed to field of use cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B.
Claim 18, recite similar subject matter as claim 8, so is rejected under the same rationale.
Regarding claim 9, dependent upon claim 1, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
wherein the trigger event is from a list of trigger events that comprises a software modification, a hardware modification, a network change, a server change, a new application integration, a threshold data storage level, and a data provider change.
The recitation in the additional limitation simply links the judicial exception to a field of use and/or technology environment, see MPEP 2106.05(h).
Limitations directed to field of use cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B.
Claim 19, recite similar subject matter as claim 9, so is rejected under the same rationale.
Regarding claim 10, dependent upon claim 1, and fail to resolve the deficiencies identified above by
integrating the judicial exception into a practical application, or introducing significantly more than the judicial exception. The claim recites:
wherein the information relating to the trigger event comprises documentation describing the trigger event or a log describing the trigger event.
The recitation in the additional limitation simply links the judicial exception to a field of use and/or technology environment, see MPEP 2106.05(h).
Limitations directed to field of use cannot integrate a judicial exception into a practical application at Step 2A or provide an inventive concept in Step 2B.
Claim 20, recite similar subject matter as claim 10, so is rejected under the same rationale.
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.
Claim(s) 1 – 4 and 11 – 14 are rejected under 35 U.S.C. 103 as being unpatentable over MEHTA et al., Pub. No.: US11488041B2, in view of Orhan et al., Pub. No.: US11748568B1, and Zhao et al., Pub. No.: US20210133535A1.
Regarding claim 1, MEHTA teaches: A method for predicting application failures using a hybrid neural network with multi-threaded inputs, the method comprising:
storing, in a first database, information relating to a plurality of digital applications;
(MEHTA (col. 3 line [60 – 65]), “The system 100 also includes a backend database server 106 [storing, in a first database] which is connected to the network 120. The backend database server 106 includes a database which may store data used with the software applications [information relating to a plurality of digital applications] and which may be supported by the web server 102 and the application server 104.”)
storing, in a second database, information relating to historical performance issues associated with the plurality of digital applications;
(MEHTA (col. 4 line [1 – 12]), “The system 100 also includes an incident database server 108 [storing, in a second database] which is connected to the network 120. The incident database server 108 includes a database which stores data concerning historical incidents occurring in the system [information relating to historical performance issues associated with the plurality of digital applications]. For example, the incident database server 108 may store incident tickets, where the incident tickets are submitted by application support teams in the event of an incident such as an outage or system failure. The incident tickets may be stored in table format in which the table contains a description of the incident ticket, an identification or name of the affected application, a symptom of the incident, and a resolution for the incident.”)
detecting a trigger event relating to one of the plurality of digital applications;
receiving information relating to the trigger event;
generating a third representation of the information relating to the trigger event;
(MEHTA, (col. 7 line [13 – 25]), “During processing of the web server log files in steps 302, 304, and 306, the system may, either simultaneously or not, process log files generated by the application server 104. In step 312, application server log files are ingested from the application server by the log database 112. Once ingested, the application server log files [receiving information relating to the trigger event] are parsed such that the log lines, which may contain narratives of the activities which triggered the generation of the log entries [detecting a trigger event relating to one of the plurality of digital applications], and the associated metadata are separated. The application server log entries may be parsed by any suitable parsing utility such as Spark or Map Reduce [generating a third representation of the information relating to the trigger event]. The benefit of this manner of parsing is that it accounts for the possibility that different development teams will define log entry formats differently.”)
receiving, as output from the neural collaborative filtering engine, a set of predicted application failures.
(MEHTA, (col. 13 line [24 – 32]), “The counter-measures can vary depending upon a symptom and type of application being analyzed. For example, if an incident is predicted based on the identification of memory scarcity as a symptom, a memory cleanup job on the server may be triggered automatically to prevent the failure [receiving, as output from the neural collaborative filtering engine, a set of predicted application failures]. In other cases, some services may be restarted to avoid failures. The counter-measures are defined by the application owner or administrator.”)
MEHTA does not teach:
training the hybrid neural network, wherein said training comprises: generating a first representation of the information in the first database;
generating a second representation of the information in the second database;
generating, via an embedding algorithm, a first information vector from the first representation;
generating, via the embedding algorithm, a second information vector from the second representation;
generating, via the embedding algorithm, a third information vector from the third representation;
feeding, using a masking algorithm, the first and the second information vectors to a transformer block neural network; and
feeding the second representation to a natural language processing (NLP) engine that is separate from the transformer block neural network;
feeding the third information vector to the transformer block neural network;
receiving a first output from the transformer block neural network;
feeding the third representation to the NLP engine;
receiving a second output from the NLP engine;
feeding the first and the second outputs to a neural collaborative filtering engine; and
Orhan teaches:
training the hybrid neural network, wherein said training comprises: generating a first representation of the information in the first database;
(Orhan, (col. 11 line [30 – 44]), “In contrast, in scenario C, the combinations of metrics and statistics that are most likely to be useful for significant anomaly detections may be selected automatically using one or more machine learning models [training the hybrid neural network, wherein said training comprises] as indicated above. Analytics/monitoring services 320 may use an existing corpus 326 of anomaly-related information [generating a first representation of the information in the first database], including for example metric metadata records, post-mortem analyses of problem events, and the like, to pre-train a machine learning model that can interpret the anomaly-related text attributes of the corpus to generate anomaly analysis relevance scores for various metrics/statistics combinations. If an application owner 304 decides that automated anomaly detection/analysis is to be performed with respect to a particular application App1, enablement of automated anomaly analysis may be initiated as indicated in element 311.”)
generating a second representation of the information in the second database;
(Orhan, (col. 8 line [43 – 56]), “The relevance detectors 125 may analyze existing information about metrics, such as monitored metrics metadata 132, monitored metrics values 133, and problem incident post-mortem analysis reports 178 [generating a second representation of the information in the second database] (i.e.: historical performance issues), to learn about the kinds of metrics and statistics which have proved useful for anomaly analysis in the past for various existing applications 155 and 156. This knowledge, which may be represented as learned parameters of one or more machine learning models (such as deep neural network models), may then be used in various embodiments to identify metrics and statistics which should be collected and analyzed for other applications 155 and 156 going forward, without requiring the application owners to go to the trouble of specifying the metrics/statistics or even defining anomalies.”)
generating, via an embedding algorithm, a first information vector from the first representation;
generating, via the embedding algorithm, a second information vector from the second representation;
(Orhan, (col. 15 line [20 – 33]), “The pre-processing 608 may include, for example, operations such as scanning, extraction, case normalization, punctuation removal, stemming, discarding of common words like “and”, “the” and the like and the extraction of terms representing metric names and statistics names in some embodiments. In at least some embodiments in which supervised learning is used, the pre-processing stage 608 of the pipeline may also comprise obtaining labels for the records (e.g., from a selected set of annotators). The records may also be vectorized (transformed into vectors suitable for use as input to the type of machine learning model being used) [generating, via an embedding algorithm, a first information vector from the first representation; generating, via the embedding algorithm, a second information vector from the second representation]. For supervised learning, the labeled records may be subdivided into training, evaluation and test datasets 609 in various embodiments.”)
generating, via the embedding algorithm, a third information vector from the third representation;
(Orhan, (col. 15 line [20 – 33]), “The pre-processing 608 may include, for example, operations such as scanning, extraction, case normalization, punctuation removal, stemming, discarding of common words like “and”, “the” and the like and the extraction of terms representing metric names and statistics names in some embodiments. In at least some embodiments in which supervised learning is used, the pre-processing stage 608 of the pipeline may also comprise obtaining labels for the records (e.g., from a selected set of annotators). The records may also be vectorized (transformed into vectors suitable for use as input to the type of machine learning model being used) [generating, via the embedding algorithm, a third information vector from the third representation]. For supervised learning, the labeled records may be subdivided into training, evaluation and test datasets 609 in various embodiments.”)
Orhan and MEHTA are related to the same field of endeavor (i.e.: machine learning system for predicting and preventing system incidents). It would have been obvious to a person having ordinary skill in the art before the effective filing date of the invention to combine the teaching of Orhan with teachings of MEHTA to enhance the ability to focus on the most useful information for anticipating issues (Orhan, Abstract)
MEHTA in view of Orhan do not teach:
feeding, using a masking algorithm, the first and the second information vectors to a transformer block neural network; and
feeding the second representation to a natural language processing (NLP) engine that is separate from the transformer block neural network;
feeding the third information vector to the transformer block neural network;
receiving a first output from the transformer block neural network;
feeding the third representation to the NLP engine;
receiving a second output from the NLP engine;
feeding the first and the second outputs to a neural collaborative filtering engine; and
Zhao teaches:
feeding, using a masking algorithm, the first and the second information vectors to a transformer block neural network; and
(Zhao, “[0026] A transformer model is a deep learning model [to a transformer block neural network] that can take an input (typically sequential data such as natural language text) in the form of a sequence of vectors [feeding, using a masking algorithm, the first and the second information vectors], and converts the input data into a vector called an encoding, and then decodes the vector back into another sequence. The encoder includes a set of encoding layers that processes the input iteratively one layer after another. Similarly, the decoder includes a set of decoding layers generating an output sequence, using the output of the encoder as an input. The output of the encoder contains latent information from the input sequence that permits the decoder to generate the output sequence. Transformer models may be trained using semi-supervised learning, such as multi-stage training procedures involving unsupervised pre-training followed by supervised fine-tuning. Pre-training may include a much larger training dataset than what is used for fine-tuning, as when labeled training data is less readily available.”)
feeding the second representation to a natural language processing (NLP) engine that is separate from the transformer block neural network;
(Zhao, “[0041] … The subset of sequences 335 a are acquired from one or more sources (e.g., a database, a URL, an email inbox, a document registry, and the like). In some instances, the subset of sequences 335 a are acquired from a data storage structure such as a database, a data system (e.g., one or more data systems 320), or the like associated with the one or more sequence generating modalities (e.g., a natural language processor such as a chatbot system) [feeding the second representation to a natural language processing (NLP) engine that is separate from the transformer block neural network].”)
feeding the third information vector to the transformer block neural network;
(Zhao, “[0041] … The subset of sequences 335 a are acquired from one or more sources (e.g., a database, a URL, an email inbox, a document registry, and the like). In some instances, the subset of sequences 335 a are acquired from a data storage structure such as a database, a data system (e.g., one or more data systems 320), or the like associated with the one or more sequence generating modalities (e.g., a natural language processor such as a chatbot system) [feeding the third information vector to the transformer block neural network].”)
receiving a first output from the transformer block neural network;
(Zhao, “[0035] … The feed forward network 235 is similar to a hidden layer (e.g., a couple of linear layers with a ReLU activation in between) in a regular feed forward network and its parameters are updated by running backpropagation based on transformer loss (output, target) [receiving a first output from the transformer block neural network] with the target being the output sequence such as a translated sentence or a sentence that completes a composition. The first decoder 205 further comprises a second normalization layer 240 that takes as input the first normalization and the feed forward output of the feed forward network 235, normalizes the inputs across each of the features, and outputs a second normalization (e.g., compute mean and variance from all of the summed inputs to the neurons in a layer on a single training or deployment case).”)
feeding the third representation to the NLP engine;
(Zhao, “[0119] Communications subsystem 724 provides an interface to other computer systems and networks. Communications subsystem 724 serves as an interface for receiving data from and transmitting data to other systems from computer system 700. For example, communications subsystem 724 may enable computer system 700 to establish a communication channel to one or more client devices via the Internet for receiving and sending information from and to the client devices. For example, the communication subsystem may be used to receive input for one or more natural language processing models [feeding the third representation to the NLP engine].”)
receiving a second output from the NLP engine;
(Zhao, “[0027] … The output [receiving a second output] of the feed-forward neural network is passed into the subsequent encoder layer and to the corresponding decoder layer, as described below. The output of the final encoder layer of the encoder is passed to each layer of the decoder. The encodings also contain positional information describing the order of each input vector in the input sequence, such as the position of constituent words in a natural language sentence [from the NLP engine]. The positional information permits the encoder to account for ordering of the sequence components.”)
feeding the first and the second outputs to a neural collaborative filtering engine; and
(Zhao, “[0035] … The feed forward network 235 is similar to a hidden layer (e.g., a couple of linear layers with a ReLU activation in between) in a regular feed forward network and its parameters are updated by running backpropagation based on transformer loss (output, target) with the target being the output sequence such as a translated sentence or a sentence that completes a composition. The first decoder 205 further comprises a second normalization layer 240 that takes as input the first normalization and the feed forward output of the feed forward network 235, normalizes the inputs across each of the features, and outputs a second normalization (e.g., compute mean and variance from all of the summed inputs to the neurons in a layer on a single training or deployment case) [feeding the first and the second outputs to a neural collaborative filtering engine].”)
Zhao, MEHTA and Orhan are related to the same field of endeavor (i.e.: machine learning system for predicting and preventing system incidents). It would have been obvious to a person having ordinary skill in the art before the effective filing date of the invention to combine the teaching of Zhao with teachings of MEHTA and Orhan to produce predictions, alerts or summaries from logs and incident data in a more compact, clear and computationally efficient manner. (Zhao, Abstract).
Claim 11 recites analogous limitations as claim 1, so is rejected under similar rationale.
Regarding claim 2, MEHTA in view of Orhan and Zhao teach the method of claim 1.
MEHTA further teaches: further comprising training the hybrid neural network to create relational associations within the plurality of digital applications, within the historical performance issues, and between the historical performance issues and the plurality of digital applications.
(MEHTA, (col. 4 line [1 – 12]), “The system 100 also includes an incident database server 108 which is connected to the network 120. The incident database server 108 includes a database which stores data concerning historical incidents occurring in the system. For example, the incident database server 108 may store incident tickets, where the incident tickets are submitted by application support teams in the event of an incident such as an outage or system failure. The incident tickets may be stored in table format in which the table contains a description of the incident ticket, an identification or name of the affected application, a symptom of the incident, and a resolution for the incident [to create relational associations within the plurality of digital applications, within the historical performance issues, and between the historical performance issues and the plurality of digital applications].”)
Claim 12 recites analogous limitations as claim 2, so is rejected under similar rationale.
Regarding claim 3, MEHTA in view of Orhan and Zhao teach the method of claim 1.
MEHTA further teaches: further comprising: calculating, for each of the set of predicted application failures, a probability score; and transmitting an alert to a system supervisor when the probability score of a predicted application failure exceeds a predetermined threshold probability score.
(MEHTA, (col. 13 line [1 – 13]), Additionally, the incident prediction engine 360 can be configured to generate alerts and/or take counter-measures to prevent predicted outages, as shown in block 372. An alert is generated [transmitting an alert to a system supervisor when the probability score of a predicted application failure] at the dashboard and/or disseminated via the network 120 when an outage prediction score [calculating, for each of the set of predicted application failures, a probability score;] exceeds a predefined prediction threshold [exceeds a predetermined threshold probability score]. The prediction threshold may be customized by the application owner such that an alert is generated when the probability of an outage reaches an unacceptable level. The alert is intended to inform the application owner that the probability of an outage based on the input log data has reached the unacceptable level, thereby affording the application owner an opportunity to take prophylactic measures.)
Claim 13 recites analogous limitations as claim 3, so is rejected under similar rationale.
Regarding claim 4, MEHTA in view of Orhan and Zhao teach the method of claim 1.
MEHTA further teaches: further comprising: storing, in the second database, historical solution data, said historical solution data comprising information relating to actions that successfully resolved the historical performance issues;
(MEHTA, (col. 4 line [1 – 12]), “The system 100 also includes an incident database server 108 which is connected to the network 120. The incident database server 108 includes a database which stores data concerning historical incidents occurring in the system. For example, the incident database server 108 may store incident tickets, where the incident tickets are submitted by application support teams in the event of an incident such as an outage or system failure. The incident tickets may be stored in table format in which the table contains a description of the incident ticket, an identification or name of the affected application, a symptom of the incident, and a resolution for the incident [storing, in the second database, historical solution data, said historical solution data comprising information relating to actions that successfully resolved the historical performance issues].”)
as part of the training, feeding the hybrid neural network with the historical solution data; and
(MEHTA, (col. 14 line [6 – 16]), “The real time model 604 is trained with regular frequency, for example, on a weekly basis, using historical anomalous logs identified by the incident prediction model 606 [as part of the training, feeding the hybrid neural network with the historical solution data]. In this way, the real time model 604 can be effectively trained using the analysis performed by the incident prediction model 606. With this training, the real time model 604 can detect incidents based on raw log messages ingested. In contrast with the incident prediction model 606 which is primarily useful for anomaly detection while a system or application is offline, the real time incident detection model 604 is useful for incident detection while a system or application is online.”)
receiving, as part of the output from the neural collaborative filtering engine, for each of the set of predicted application failures,
(MEHTA, (col. 12 line [48 – 58]), “The incident prediction engine 360 divides a timeline into equal intervals of time using a sliding window. It then uses the combination of features occurring within the window to predict the incident which occurred historically in the same window. The window is then shifted by equal time steps in a sliding manner. The same process is repeated to train the classifier model to predict the time of the incident. The trained classifier model is then used to predict the future incidents [receiving, as part of the output from the neural collaborative filtering engine, for each of the set of predicted application failures]. The classifier model is trained by batch jobs on regular intervals, such as daily or weekly, at step 356 to keep the model up to date with behavioral changes in the servers.”)
a recommended action to resolve the predicted application failure.
(MEHTA, (col. 14 line [46 – 55]), “The system provides advanced alerts that the status of any system components has become critical, e.g., that the system component is likely to fail and further provides a probable cause of the likely failure. The system may also suggest a resolution for the predicted incident based upon similar historical incidents that have been resolved [a recommended action to resolve the predicted application failure]. This information can allow a platform owner to avoid failures or take counter actions. The platform owner may also optimize the server utilization or enable automatic scaling based upon the prediction.”)
Claim 14 recites analogous limitations as claim 4, so is rejected under similar rationale.
Claim(s) 5 – 6 and 15 – 16 are rejected under 35 U.S.C. 103 as being unpatentable over MEHTA in view of Orhan, Zhao and in further view of Dean et al., Pub. No.: US11703826B1.
Regarding claim 5, MEHTA in view of Orhan and Zhao teach the method of claim 1.
MEHTA further teaches: wherein: the output from the neural collaborative filtering engine further comprises a probability score for each of the predicted application failures; and the method further comprises providing the set of predicted application failures as rows in a table,
(MEHTA, (col. 4 line [1 – 12]), “The system 100 also includes an incident database server 108 which is connected to the network 120. The incident database server 108 includes a database which stores data concerning historical incidents occurring in the system. For example, the incident database server 108 may store incident tickets, where the incident tickets are submitted by application support teams in the event of an incident such as an outage or system failure. The incident tickets may be stored in table format [the output from the neural collaborative filtering engine further comprises a probability score for each of the predicted application failures; and the method further comprises providing the set of predicted application failures as rows in a table] in which the table contains a description of the incident ticket, an identification or name of the affected application, a symptom of the incident, and a resolution for the incident.”)
MEHTA in view of Orhan and Zhao do not teach:
wherein the rows are ordered according to the probability scores.
Dean t