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 .
Response to Amendment and Arguments
Applicant’s amendment filed on March 30, 2026 has been entered and made of record. Claims 1-7, 9-14, 16-19, and 21-23 are pending and are being examined in this application.
In light of Applicant’s remarks and amendments to the claims, the claim objection and the 101 rejection are withdrawn.
There are no specific arguments provided with respect to the 103 rejections aside from a mere conclusory statement that the cited references do not teach or disclose the limitation “wherein performing the one or more automated actions comprises automatically routing at least one message to at least one of the one or more recommended alternate messaging topologies upon occurrence of at least one event related to the one or more predicted anomalies, thereby altering an active message routing state of the at least one message-oriented middleware” as recited in the independent claims [Remarks, pg. 12].
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-3, 6, 11, 12, 14, 16, 17, 19, and 22 are rejected under 35 U.S.C. 103 as being unpatentable over Qin et al. (NPL “Open Sourcing Kafka Cruise Control”) in view of Belacel et al. (“Online Anomaly Detection for Streaming Data Implemented on Top of Kafka, Scikit-Multiflow and River”).
Regarding claim 1, QIN teaches “A computer-implemented method comprising: obtaining data pertaining to at least one messaging topology associated with at least one message-oriented middleware”:
([Page 1, paragraphs 1-2] “Apache Kafka's (a popular streaming platform that supports various messaging patterns/ messaging topology) popularity has grown tremendously over the past few years… In fact, LinkedIn's deployment recently surpassed 2 trillion messages per day, with over 1,800 Kafka servers (i.e., brokers) (showing it to be well-known)…
…Intelligent automation is critical under these circumstances, which is why we developed Cruise Control (a middleware): a general-purpose system that continually monitors our clusters and automatically adjusts the resources allocated to them to meet pre-defined performance goals (continuous monitoring of the clusters is equivalent to “obtaining data pertaining to at least one messaging topology associated with at least one message-oriented middleware”). In essence, users specify goals, Cruise Control monitors for violations of these goals, analyzes the existing workload on the cluster, and automatically executes administrative operations to satisfy those goals...”) Further, the “Cruise Control Architecture” figure on page 4 explicitly shows the Kafka Cluster/the data pertaining to the messaging topology being received in association with Cruise Control, the middleware.
Further, QIN teaches “predicting one or more anomalies associated with the at least one messaging topology by processing at least a portion of the obtained data using a first set of one or more… techniques”:
([Page 1, paragraph 2] “Intelligent automation is critical under these circumstances, which is why we developed Cruise Control: a general-purpose system that continually monitors our clusters and automatically adjusts the resources allocated to them to meet pre-defined performance goals. In essence, users specify goals, Cruise Control monitors for violations of these goals, analyzes the existing workload on the cluster, and automatically executes administrative operations to satisfy those goals...”) Here, we see that cruise control monitors for violations to goals / anomalies.
And further:
([Page 4, Anomaly Detector] “The anomaly detector identifies two types of anomalies (one or more techniques):
1. Broker failures: i.e., a non-empty broker leaves a cluster, which results in under-replicated partitions. Since this can happen during normal cluster bounces as well, the anomaly detector provides a configurable grace period before it triggers the notifier and fixes the cluster (meaning this issue is identified before being confirmed as an issue, which qualifies it as a “prediction”).
2. Goal violations: i.e., an optimization goal is violated. If self-healing is enabled, Cruise Control will proactively attempt to address the goal violation (proactively attempting to address an anomaly means it was “predicted” before it became an error.) by automatically analyzing the workload, and executing optimization proposals.”)
Further, QIN teaches “recommending one or more alternate messaging topologies associated with the at least one message-oriented middleware by processing at least a portion of the one or more predicted anomalies and at least a portion of the obtained data using a second set of one or more… techniques”:
([Page 3, Analyzer, paragraph 1] “The Analyzer is the "brain" of Cruise Control. It uses a heuristic method (one or more techniques…) to generate optimization proposals based on the user-provided optimization goals and the cluster workload model from the Load Monitor. (generating “optimization proposals” equivalates to “recommending alternate messaging topologies” which are associated with the message-oriented middleware (“Cruise Control”) by processing the predicted anomalies)”)
Further, QIN teaches “performing one or more automated actions based at least in part on one or more of the one or more predicted anomalies and the one or more recommended alternate messaging topologies”:
([Page 4, Executor] “The executor is responsible for carrying out the optimization proposals from the analyzer…”) Here, we see that the executor carries out the optimization proposals/recommendations provided before, which means is performs the recommended actions based on the predicted anomalies and the alternate messaging topologies.
Further, QIN teaches “wherein performing one or more automated actions comprises automatically routing at least one message to at least one of the one or more recommended alternate messaging topologies upon occurrence of at least one event related to the one or more predicted anomalies”:
([Page 2, Cruise Control at LinkedIn] “…2. When a broker fails, we need to automatically reassign replicas that were on that broker to other brokers in the cluster and restore the original replication factor…”) In other words, when a broker/messaging topology fails (an event related to an anomaly), we automatically reassign the messages (routing at least one message), or replicas, to other brokers / alternate messaging topologies.
And further:
([Page 4, Anomaly Detector] “The anomaly detector identifies two types of anomalies:
1. Broker failures: i.e., a non-empty broker leaves a cluster, which results in under-replicated partitions. Since this can happen during normal cluster bounces as well, the anomaly detector provides a configurable grace period before it triggers the notifier and fixes the cluster.
2. Goal violations: i.e., an optimization goal is violated. If self-healing is enabled, Cruise Control will proactively attempt to address the goal violation. by automatically analyzing the workload, and executing optimization proposals.”)
Further, QIN teaches “thereby altering an active message routing state of the at least one message-oriented middleware”:
([Page 4, Anomaly Detector] “The anomaly detector identifies two types of anomalies (identifies an active message routing state of abnormal):
1. Broker failures: i.e., a non-empty broker leaves a cluster, which results in under-replicated partitions. Since this can happen during normal cluster bounces as well, the anomaly detector provides a configurable grace period before it triggers the notifier and fixes the cluster (alters the active message routing state to normal).
2. Goal violations: i.e., an optimization goal is violated. If self-healing is enabled, Cruise Control will proactively attempt to address the goal violation (alter the active message routing state to normal).
Further, QIN teaches “wherein the method is performed by at least one processing device comprising a processor coupled to a memory”:
([Page 5, Memory or speed?] “Cruise Control is very memory-intensive... It is also a CPU-intensive application...”) This section clearly explains that the application is both memory-intensive and CPU-intensive, meaning it is performed by at least one processing device that must comprise at least one processor coupled to a memory.
QIN fails to explicitly teach the first and second sets of techniques to be “machine learning techniques.” However, analogous art, BELACEL, does teach this:
([Abstract] “Identification of rare events is an important component in intrusion and fraud detection, system monitoring and event detection. Anomalies can represent problematic situations where early, accurate and actionable insights are critical to make situational assessments in the event of unexpected conditions. For many problems, the state of the art in machine learning (the use of machine learning) is batch learning… Online anomaly detection algorithms offer rapid access to useful insights with fewer computing capacity requirements and often need to be integrated with existing or legacy data streams in the enterprise. This work introduces a proof of concept integration of two online streaming anomaly detection algorithms (one or more sets of machine learning techniques) available in the scikit multiflow and River frameworks with the popular Kafka event streaming platform.”)
It would have been obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention, to combine the base reference of QIN with the teachings of BELACEL because both references explore anomaly detection methods in relation to streaming, and both even do so with the same “Kafka” event streaming platform.
One of ordinary skill in the art would be motivated to do so because, as pointed out by BELACEL in its’ introduction, “The ability to monitor processes in near real-time that online anomaly detection algorithms afford promises rapid access to useful insights with fewer computing capacity requirements”, “Integrating online anomaly detection algorithms with existing message oriented systems is key to leveraging existing technology investments”, and “Although there exists multiple machine learning frameworks for streaming data, these novel systems tend to be stand-alone and often need to be integrated with existing or legacy data streams in the enterprise. Our work presents a proof of concept integration of two online streaming anomaly detection algorithms available in the scikit-multiflow and River frameworks with the popular Kafka event streaming platform.”
Regarding claim 2, QIN in view of BELACEL teaches the limitations of claim 1. Further, BELACEL teaches “wherein predicting one or more anomalies comprises processing the at least a portion of the obtained data using one or more unsupervised learning techniques”:
([Page 828, 3. Method] “A simple online unsupervised learning problem was selected for the purposes of introducing our streaming anomaly detection solution. We focus on applying two popular online anomaly detection algorithms for raw data stream annotation and enrichment (both of which process at least a portion of the data to detect anomalies in an unsupervised learning environment). The first algorithm is the Half-Space Trees (HS-Trees) algorithm [7] and the second is the Isolation Forest (IForestASD) algorithm [8].”)
Regarding claim 3, QIN in view of BELACEL teaches the limitations of claim 2. Further, BELACEL teaches “wherein processing the at least a portion of the obtained data using one or more unsupervised learning techniques comprises processing the at least a portion of the obtained data using at least one of using at least one isolation forest algorithm”:
([Page 828, 3. Method] “A simple online unsupervised learning problem was selected for the purposes of introducing our streaming anomaly detection solution. We focus on applying two popular online anomaly detection algorithms for raw data stream annotation and enrichment). The first algorithm is the Half-Space Trees (HS-Trees) algorithm [7] and the second is the Isolation Forest (IForestASD) algorithm [8].”)
Regarding claim 6, QIN in view of BELACEL teaches the limitations of claim 1. Further, BELACEL teaches “wherein recommending one or more alternate messaging topologies comprises processing at least a portion of the one or more predicted anomalies and at least a portion of the obtained data using at least one machine learning-based classification algorithm”:
([Page 828, 3. Method] “A simple online unsupervised learning problem was selected for the purposes of introducing our streaming anomaly detection solution. We focus on applying two popular online anomaly detection algorithms for raw data stream annotation and enrichment). The first algorithm is the Half-Space Trees (HS-Trees) algorithm [7] and the second is the Isolation Forest (IForestASD) algorithm [8].”) Both of the cited algorithms are machine learning-based classification algorithms.
Regarding claim 11, see at the rejection for claim 1. QIN further teaches “A non-transitory processor-readable storage medium having stored therein program code of one or more software programs, wherein the program code when executed by at least one processing device causes the at least one processing device” to perform the claimed steps:
([Page 5, Memory or speed?] “Cruise Control (a software program made up of program code) is very memory-intensive... It is also a CPU-intensive application because of the involved computation that is necessary to generate optimization proposals.”) Being both CPU-intensive and memory-intensive means that a CPU (a processing device which is a non-transitory medium itself), is connected to some form of memory which must store the program code, indicating a computer, which in and of itself, is a non-transitory processor-readable storage medium, which executes the program code, causing the processing device to perform the remaining limitations.
Regarding claim 12, see the rejection for claim 2.
Regarding claim 14, see the rejection for claim 6.
Regarding claim 16, see at least the rejection for claim 1. QIN further teaches “An apparatus comprising: at least one processing device comprising a processor coupled to a memory; the at least one processing device being configured” to perform the claimed steps:
([Page 5, Memory or speed?] “Cruise Control (a software program made up of program code) is very memory-intensive... It is also a CPU-intensive application because of the involved computation that is necessary to generate optimization proposals.”) Being both CPU-intensive and memory-intensive means that a CPU, which is a processing device, is connected to some form of memory which must store the program code, indicating a computer, which in and of itself, is an apparatus, which executes the program code, causing the processing device to perform the remaining limitations.
Regarding claim 17, see the rejection for claim 2.
Regarding claim 19, see the rejection for claim 6.
Regarding claim 22, QIN in view of BELACEL teaches the limitations of claim 1. Further, QIN teaches “The apparatus of claim 16, wherein performing one or more automated actions comprises automatically training at least a portion of the second set of one or more machine learning techniques using feedback related to the one or more recommended alternate messaging topologies.
([Page 1, Design goals, Extensibility]: “Kafka users will have different requirements for the availability and performance of their Kafka deployments, and will use various deployment tools, administrative endpoints, and metrics-collection mechanisms. Cruise Control must be able to satisfy arbitrary user-defined goals and execute user-defined actions.”)
([Page 3, Cruise Control Architecture): Metrics reporter feeds Standard Metrics to the Load Monitor, which loads the [cluster workload] model to the Analyzer.)
([Page 3, Load Monitor, paragraph 1] “The Load Monitor collects standard Kafka metrics from the cluster and derives per partition resource metrics that are not directly available (e.g., it estimates CPU utilization on a per-partition basis). It then generates a cluster workload model that accurately captures cluster resource utilization, which includes disk utilization, CPU utilization, bytes-in rate, and bytes-out rate, at replica-granularity. It then feeds the cluster model into the detector and analyzer.”)
([Page 3, Analyzer, paragraph 1] “The Analyzer is the "brain" of Cruise Control. It uses a heuristic method (one or more techniques…) to generate optimization proposals based on the user-provided optimization goals and the cluster workload model from the Load Monitor.” Generating “optimization proposals” equivalates to “recommended alternate messaging topologies,” which are generated based on the cluster workload model generated by the Load Monitor based on the standard metrics.)
Claims 4, 5, 13, and 18 are rejected under 35 U.S.C. 103 as being unpatentable over QIN and BELACEL in view of Sung et al. (NPL “Identifying Important Features for Intrusion Detection Using Support Vector Machines and Neural Networks”).
Regarding claim 4, QIN in view of BELACEL teaches the limitations of claim 1. Further, QIN in view of BELACEL fails to explicitly teach “wherein predicting one or more anomalies comprises processing the at least a portion of the obtained data using one or more supervised learning techniques”. However, analogous art, SUNG, does teach this:
([Abstract] “Intrusion detection (anomaly detection) is a critical component of secure information systems. This paper addresses the issue of identifying important input features in building an intrusion detection system (IDS)…
…In this paper we apply the technique of deleting one feature at a time to perform experiments on SVMs (support vector machines) and neural networks to rank the importance of input features for the DARPA collected intrusion data. Important features for each of the 5 classes of intrusion patterns in the DARPA data are identified. (identifying “important features” for “intrusion patterns” is equivalent to “predicting anomalies”, and the use of support vector machines indicates a type of supervised learning technique.)”)
It would have been obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention, to combine the base reference of QIN in view of BELACEL with the teachings of SUNG because both references pertain to anomaly detection using machine learning techniques.
One of ordinary skill in the art would be motivated to do so because, as pointed out in the abstract of SUNG, “It is shown that SVM-based and neural network based IDSs using a reduced number of features can deliver enhanced or comparable performance.”
Regarding claim 5, QIN in view of BELACEL & SUNG teaches the limitations of claim 4. Further, SUNG teaches “wherein processing the at least a portion of the obtained data using one or more supervised learning techniques comprises processing the at least a portion of the obtained data using at least one of one or more support vector machines and one or more artificial neural networks”:
([Abstract] “Intrusion detection is a critical component of secure information systems. This paper addresses the issue of identifying important input features in building an intrusion detection system (IDS)…
…In this paper we apply the technique of deleting one feature at a time to perform experiments on SVMs (support vector machines) and neural networks (artificial neural networks) to rank the importance of input features (at least a portion of the obtained data) for the DARPA collected intrusion data. Important features for each of the 5 classes of intrusion patterns in the DARPA data are identified.”)
Regarding claim 13, see the rejection for claim 4.
Regarding claim 18, see the rejection for claim 4.
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over QIN and BELACEL in view of Farnaaz et al. (NPL “Random Forest Modeling for Network Intrusion Detection System”).
Regarding claim 7, QIN in view of BELACEL teaches the limitations of claim 6. Further, QIN in view of BELACEL fails to explicitly teach “wherein processing at least a portion of the one or more predicted anomalies and at least a portion of the obtained data using at least one machine learning-based classification algorithm comprises processing at least a portion of the one or more predicted anomalies and at least a portion of the obtained data using at least one random forest classifier, wherein using the at least one random forest classifier comprises implementing one or more bootstrap aggregating techniques in connection with multiple individual classifiers each trained on different data”.
However, analogous art, FARNAAZ, does teach “wherein processing at least a portion of the one or more predicted anomalies and at least a portion of the obtained data using at least one machine learning-based classification algorithm comprises processing at least a portion of the one or more predicted anomalies and at least a portion of the obtained data using at least one random forest classifier”:
([Abstract] “With the growing usage of technology, intrusion detection (anomaly detection) became an emerging area of research. Intrusion Detection System (IDS) attempts to identify and notify the activities of users as normal (or) anomaly. IDS is a nonlinear and complicated problem and deals with network traffic data… In this paper, we have built a model for intrusion detection system using random forest classifier...”) Further, figure 1 on page 215 explicitly shows “Load dataset” (obtained data), “Classify features according to attacks -> Apply feature subset selection” (at least a portion of the one or more detected anomalies), followed by “Apply random forest algorithm”, which means that the random forest algorithm is explicitly used to process/classify obtained data and predicted anomalies.
Further, FARNAAZ teaches “wherein using the at least one random forest classifier comprises implementing one or more bootstrap aggregating techniques in connection with multiple individual classifiers each trained on different data”:
As disclosed above, FARNAAZ teaches the use of a random forest classifier, and this limitation simply describes how random forest classifiers traditionally work, as is defined at https://en.wikipedia.org/wiki/Random_forest and quoted below for convenience.
([Algorithm, Preliminaries: decision tree learning] “In particular, trees that are grown very deep tend to learn highly irregular patterns: they overfit their training sets, i.e. have low bias, but very high variance. Random forests are a way of averaging multiple deep decision trees, trained on different parts of the same training set, with the goal of reducing the variance.”)
And further:
([Algorithm: Bagging] “The training algorithm for random forests applies the general technique of bootstrap aggregating, or bagging, to tree learners. Given a training set X = x1, ..., xn with responses Y = y1, ..., yn, bagging repeatedly (B times) selects a random sample with replacement of the training set and fits trees to these samples…”)
It would have been obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention, to combine the base reference of QIN in view of BELACEL with the teachings of FARNAAZ because both references explore detection of anomalies using machine learning methods.
One of ordinary skill in the art would be motivated to do so because, as pointed out in the abstract of FARNAAZ, “Random Forest (RF) is an ensemble classifier and performs well compared to other traditional classifiers for effective classification of attacks” and “Empirical result show that proposed model is efficient with low false alarm rate and high detection rate.”
Claims 9, 21, and 23 are rejected under 35 U.S.C. 103 as being unpatentable over QIN and BELACEL in view of Du et al. (NPL “DeepLog: Anomaly Detection and Diagnosis from System Logs through Deep Learning”).
Regarding claim 9, QIN in view of BELACEL teaches the limitations of claim 1. However, QIN view of BELACEL fails to explicitly teach “wherein performing one or more automated actions comprises automatically training at least a portion of the first set of one or more machine learning techniques using feedback related to the one or more predicted anomalies”. However, analogous art, DU, does teach this:
([Abstract] “Anomaly detection is a critical step towards building a secure and trustworthy system… Log data is an important and valuable resource for understanding system status and performance issues… We propose DeepLog, a deep neural network model utilizing Long Short-Term Memory (LSTM), to model a system log as a natural language sequence. This allows DeepLog to automatically learn log patterns from normal execution, and detect anomalies when log patterns deviate from the model trained from log data under normal execution. In addition, we demonstrate how to incrementally update the DeepLog model in an online fashion so that it can adapt to new log patterns over time...”) Further, Figure 1 of page 1287 explicitly shows a diagnosis step that leads to “update model if false positive” which directly correlates to a automatically training the ML techniques using feedback related to the predicted anomalies.
It would have been obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention, to combine the base reference of QIN in view of BELACEL with the teachings of DU because both references explore anomaly detection using ML techniques.
One of ordinary skill in the art would be motivated to do so because, as pointed out in the abstract of DU, “Extensive experimental evaluations over large log data have shown that DeepLog has outperformed other existing log-based anomaly detection methods based on traditional data mining methodologies.”
Regarding claim 21, see the rejection for claim 9.
Regarding claim 23, see the rejection for claim 9.
Claim 10 is rejected under 35 U.S.C. 103 as being unpatentable over QIN and BELACEL in view of Zhang et al. (NPL “CDBTune+: An efficient deep reinforcement learning-based automatic cloud database tuning system”).
Regarding claim 10, QIN in view of BELACEL teaches the limitations of claim 1. However, QIN in view of BELACEL fails to explicitly teach “wherein performing one or more automated actions comprises automatically training at least a portion of the second set of one or more machine learning techniques using feedback related to the one or more recommended alternate messaging topologies”. However, analogous art, ZHANG, does teach “wherein performing one or more automated actions comprises automatically training at least a portion of the second set of one or more machine learning techniques using feedback related to the one or more recommended alternate…” configuration parameter settings:
([Abstract] “…we design an end-to-end automatic CDB tuning system, CDBTune+, using deep reinforcement learning (RL). CDBTune+ utilizes the deep deterministic policy gradient method to find the optimal configurations in a high-dimensional continuous space. CDBTune+ adopts a trial-and-error strategy to learn knob settings with a limited number of samples to accomplish the initial training, which alleviates the necessity of collecting a massive amount of high-quality samples. CDBTune+ adopts the reward-feedback mechanism in RL instead of traditional regression, which enables end-to-end learning and accelerates the convergence speed of our model and improves the efficiency of online tuning...”) “Reinforcement learning”, which uses a “reward-feedback” loop to learn appropriate configurations is a standard feature in the art. In this citation, it is used for determining appropriate configurations of knob settings, but when combined with QIN in view of BELACEL, it learns “alternate messaging topologies” to recommend, thus resulting in the configuration as claimed.
It would have been obvious to one of ordinary skill in the art, prior to the effective filing date of the claimed invention, to combine the base reference of QIN in view of BELACEL with the teachings of ZHANG because both references use ML techniques to optimize a system based on recommended configurations.
One of ordinary skill in the art would be motivated to do so because, as pointed out in the abstract of ZHANG, adopting “the reward-feedback mechanism in RL instead of traditional regression… enables end-to-end learning and accelerates the convergence speed of our model and improves the efficiency of online tuning.”
Conclusion
THIS ACTION IS MADE FINAL. 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.
Contact Information
Any inquiry concerning this communication or earlier communications from the examiner should be directed to GRACE PARK whose telephone number is (571)270-7727. The examiner can normally be reached M-F 8AM-5PM.
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, TAMARA KYLE can be reached at (571)272-4241. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/Grace Park/Primary Examiner, Art Unit 2144
