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 Claims
This is a Final Rejection office action in response to application Serial No. 18/763,842. Claim(s) 1-20 have been examined and fully considered, claims 1, 7, 10, and 13 have been amended.
Claim(s) 1-20 are pending in Instant Application.
Response to Arguments/Rejections
Applicant's arguments filed 12/15/2025 have been fully considered but they are not persuasive.
Per remarks, Summary of Examiner Interview applicant states “The Examiner agreed that amendments discussed during interview, and included herein, overcome the current 101, 102, and 103 rejections, subject to further search and consideration.”.
Examiner respectfully disagrees, per interview summary, Examiner states “A proposed amendment to claims 1,was discussed. Proposed amendments likely act to overcome the 103 rejection of record. Examiner recommends the claims be further amended to incorporate elements of paragraph [0079]. Upon receiving a formal response, further search and consideration will be performed.”. However, it appears that the applicant do not implement the recommended subject matter of applicant’s specification.
Regarding, “Discussion of the Claim Rejections under 35 U.S.C. 4 101” ,applicant states “As discussed above, the Examiner agreed during interview that the amendments included herein overcome the current 101 rejections. Applicant therefore respectfully requests withdrawal of this rejection. ”.
Examiner respectfully disagrees. Examiner notes that the step directed to step “triggering…” is determined to be an additional element, which merely amounts to post-solution
activity. The amendment is not claiming the use the generic components sever, processor and
vehicle contributes to “trigger…” data/information utilized for implementing said abstract
idea (i.e. does not improve the computing technology or the transportation technology, e.g., the
operation of automobile, , etc.) , at best the data is being “displayed”, however, this does not
represent an integration of said abstract idea into a practical application.
Regarding “Discussion of the Rejections Under Double Patenting”. applicant states “Reconsideration is respectfully requested in view of the remarks and amendments submitted herewith. Further, Applicant respectfully submits that the nonstatutory double patenting rejections is not ripe, for example, as no allowable claims have been indicated in the previous Office Action and without allowed claims there can be no double patenting by definition. ”.
Examiner respectfully disagrees. Once the claim features in rendered for allowance, and a “Terminal Disclaimer”, is filed, the application would be in condition for allowance.
Therefore, Examiner maintains rejections of prior action of recorded.
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 USC § 101 based on the following analysis because the claimed invention is directed to an abstract idea without being significantly more.
Step 1 of the Subject Matter Eligibility Test entails considering whether the claimed subject matter falls within the four statutory categories of patentable subject matter identified by 35 U.S.C. 101: Process, machine, manufacture, or composition of matter.
Claim(s) 1-20 are directed to methods. As such, the claims are directed to statutory categories of invention.
If the claim recites a statutory category of invention, the claim requires further analysis
in Step 2A. Step 2A of the Subject Matter Eligibility Test is a two-prong inquiry. In Prong One,
examiners evaluate whether the claim recites a judicial exception.
Claim 1 recites abstract limitations, including those bolded below:
A method, performed by a processing system comprising one or more processors and one or more memories coupled with the one or more processors, comprising:
detecting a trip associated with vehicular transportation of a mobile device;
determining a location dataset representing locations of the mobile device;
determining a plurality of features based on a first comparison between the location dataset and a bus route dataset, wherein determining the plurality of features comprises:
generating a plurality of candidate route segments based on the location dataset;
in response to generating the plurality of candidate route segments:
generating a set of candidate bus routes based on the plurality of candidate route segments;
selecting a subset of the set of candidate bus routes based on contextual information; and
determining the plurality of features based on the subset of the set of candidate bus routes;
classifying the trip based on the plurality of features; and
triggering, based on the classification of the trip, an action at the mobile device, wherein the action facilitates at least one of provision of content or a mobile device service.
Claim 13 recites abstract limitations, including those bolded below:
A method for classification of vehicle trip transportation modality,
performed by a processing system comprising one or more processors and one or more memories coupled with the one or more processors, comprising:
receiving a trip dataset comprising location data collected with a location sensor of a mobile user device;
determining a set of features by comparing the trip dataset to a transit dataset, comprising:
determining a candidate transit route comprising a series of route segments using the transit dataset; and
in response to determining the candidate transit route:
determining a dynamic time warping [DTW] similarity score for the candidate transit route and the trip dataset, wherein the set of features comprises the DTW similarity score;
based on the set of features, and the DTW similarity score, classifying the vehicle trip as a transit trip; and
triggering, based on the classification of the vehicle trip as a transit trip, an action at the mobile user device, wherein the action facilitates at least one of provision of content or a mobile device service.
These limitations, as drafted, are a process that, under its broadest reasonable
interpretation, cover performance of the limitations in the mind, or by a human using pen and
paper, and therefore recite mental processes. More specifically, as there is no recitation of a processing structure (i.e. processor, etc.) for executing the method steps, nothing in the claim element precludes the aforementioned steps from practically being performed in the human mind, or by a human using pen and paper. (i.e. determining...,, generating,…, and classifying,… etc.). Examiner notes that the mere recitation of a generic computer would not take the claim out of the mental process grouping. Thus, the claim recites an abstract idea.
If the claim recites a judicial exception in step 2A Prong One , the claim requires further analysis in step 2A Prong Two. In step 2A Prong Two, examiners evaluate whether the claim recites additional elements that integrate the exception into a practical application of that exception.
Claim 1 recites the additional elements of:
A method, performed by a processing system comprising one or more processors and one or more memories coupled with the one or more processors, comprising:
detecting a trip associated with vehicular transportation of a mobile device;
determining a location dataset representing locations of the mobile device;
determining a plurality of features based on a first comparison between the location dataset and a bus route dataset, wherein determining the plurality of features comprises:
generating a plurality of candidate route segments based on the location dataset;
in response to generating the plurality of candidate route segments:
generating a set of candidate bus routes based on the plurality of candidate route segments;
selecting a subset of the set of candidate bus routes based on contextual information; and
determining the plurality of features based on the subset of the set of candidate bus routes;
classifying the trip based on the plurality of features; and
triggering, based on the classification of the trip, an action at the mobile device, wherein the action facilitates at least one of provision of content or a mobile device service.
Claim 13 recites the additional elements of:
A method for classification of vehicle trip transportation modality,
performed by a processing system comprising one or more processors and one or more memories coupled with the one or more processors, comprising:
receiving a trip dataset comprising location data collected with a location sensor of a mobile user device;
determining a set of features by comparing the trip dataset to a transit dataset, comprising:
determining a candidate transit route comprising a series of route segments using the transit dataset; and
in response to determining the candidate transit route:
determining a dynamic time warping [DTW] similarity score for the candidate transit route and the trip dataset, wherein the set of features comprises the DTW similarity score;
based on the set of features, and the DTW similarity score, classifying the vehicle trip as a transit trip; and
triggering, based on the classification of the vehicle trip as a transit trip, .
For the following reason(s), the examiner submits that the above identified additional limitations do not integrate the above-noted abstract idea into a practical application.
Regarding the additional limitations of “detecting a trip associated with vehicular transportation of a mobile device” and “receiving a trip dataset comprising location data collected with a location sensor of a mobile user device” these limitations amount to pre-solution data gathering activities (extra-solution activity – data collection for use in the abstract idea).
Regarding the additional limitations of “triggering, based on the classification of the vehicle trip as a transit trip, triggering an action at the mobile user device, wherein the action facilitates at least one of provision of content or a mobile device service,” these limitations mount to post-solution activity (e.g., the provision of information on a display, as described in paragraph 0017 of the specification).
The functions of the “memories”, “a processing system”, “processors”, “a mobile device” and “a location sensor” are recited at a high level of generality and are merely invoked as tools to perform the abstract idea. In addition, each of these aforementioned additional elements indicate a field of use or technological environment in which to apply a judicial exception and cannot integrate the judicial exception into a practical application (see MPEP 2106.05(h)).
If the additional elements do not integrate the exception into a practical application in step 2A Prong Two, then the claim is directed to the recited judicial exception, and requires further analysis under Step 2B to determine whether they provide an inventive concept (i.e., whether the additional elements amount to significantly more than the exception itself). As discussed above, the “memories”, “a processing system”, “processors”, “a mobile device” and “a location sensor” amounts to act to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. Use of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea does not provide significantly more. See Affinity Labs v. DirecTV, 838 F.3d 1253, 1262, 120 USPQ2d 1201, 1207 (Fed. Cir. 2016) (cellular telephone); TLI Communications LLC v. AV Auto, LLC, 823 F.3d 607, 613, 118 USPQ2d 1744, 1748 (Fed. Cir. 2016) (computer server and telephone unit).
As discussed above, each of these aforementioned additional limitations also amounts to merely indicating a field of use or technological environment in which to apply a judicial exception, which does not amount to significantly more than the exception itself. (see MPEP 2106.05(h)).
As discussed above, the input/output functions of “memories”, “a processing system”, “processors”, “a mobile device” and “a location sensor” are considered as insignificant extra-solution activity. MPEP 2106.05(d)(II), and the cases cited therein, including in Trading Techs. Int’l v. IBG LLC, 921 F.3d 1084, 1093 (Fed. Cir. 2019), and Intellectual Ventures I LLC v. Erie Indemnity Co., 850 F.3d 1315, 1331 (Fed. Cir. 2017), for example, indicated that the mere displaying of data is a well understood, routine, and conventional function. The Versata and OIP Techs court decisions cited in MPEP 2106.05(d)(II) indicate that storing and retrieving data in memory is a well‐understood, routine, conventional function when it is claimed in a merely generic manner (as it is here). The Symantec, TLI, OIP Techs. and buySAFE court decisions cited in MPEP 2106.05(d)(II) indicate that mere receiving or transmitting data over a network is a well‐understood, routine, conventional function when it is claimed in a merely generic manner (as it is here). The Internet Patent Corp. v. Active Network, Inc decision indicate that a browser’s button functionality is a well‐understood, routine, conventional function when it is claimed in a merely generic manner (as it is here).
In addition, the specification demonstrates the well-understood, routine, conventional nature of additional elements as it describes the additional elements as well-understood or routine or conventional (or an equivalent term), as a commercially available product, or in a manner that indicates that the additional elements are sufficiently well-known that the specification does not need to describe the particulars of such additional elements to satisfy 35 U.S.C. §112(a).
The various metrics/limitations of claims 2-4, 7-12, 14 and 17-18 and 20 merely narrow the previously recited abstract idea limitations (e.g., further characterizing the timing of data collection (real-time), the set of candidate routes, trip classification, the plurality of features, determination of stop locations, contextual information, etc.) . For the reasons described above, this judicial exception is not meaningfully integrated into a practical application, or significantly more than the abstract idea. Examiner notes that claim 8 does not recite the active collection of data by an inertial sensor of a phone, and the analysis of the data itself is abstract.
Claim(s) 5-6 , 15-16 and 19 recite a machine-learning-based classification model, a heuristic tree-based process, a pre-trained HMM, and a Bayesian network, which are tools used in their ordinary capacity to perform the abstract idea, and therefore amount to “apply it.”
Therefore, claim(s) 1-20 is/are ineligible under 35 USC §101.
Claim Rejections - 35 USC § 102
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 the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claim(s) 1-15, and 17-20 is/are rejected under 35 U.S.C. 102 (a)(1) as being anticipated by Park (US 2018/0338223).
Regarding [claim 1], Park discloses a method, performed by a processing~ system comprising~ one or more processors and one or more memories coupled with the one or more processors (see, Paragraph [0003]: “In some contexts , telematics applications involve collecting , storing , and processing data from sensors in vehicles. In particular, mobile telematics (sometimes known as “smartphone telematics” ) uses mobile sensing technologies to collect , store , and process data from built-in or external sensors of a mobile device, such as a smartphone” and Paragraph [0075]: “At the end of the training, the final classification model is stored in a permanent storage and deployed to production servers. When a recording for a trip is presented from a mobile device for trip mode classification , a production server loads and applies the stored final classification model to the trip , producing segment- by-segment trip mode identifications. In some implementations, the final classification model can be transferred to the mobile device, which loads and applies it to identify a transportation mode for the trip. In some implementations, the computational work of applying the model to the data of the recordings can be split In between the server and the mobile device”), comprising
detecting a trip associated with vehicular transportation of a mobile device (see, Paragraph [0004] “A mobile telematics application may employ various schemes to decide when it should collect sensor data. A common mode of operation is to collect sensor data (for example, involving one or more of acceleration, gyroscope, magnetometer, position, velocity, or barometer, among others) whenever the mobile device associated with the sensor is thought to be moving, or thought to be associated with driving (for vehicular applications), or has undergone a significant change in location from a previous known position. The application collects data from the sensors until the mobile device has come to rest for a substantial amount of time.”;
determining a location dataset representing locations of the mobile device (see, Paragraphs [0035] “In some implementations, the technology identifies one or more transportation modes of one or more trips using recordings captured by one or more mobile devices. In some instances, the technology can use a variety of mobile sensor data included in the recordings to extract features useful for identifying transportation modes, participant roles, or other trip types, and non-trip phantom recordings.”
[0038]: “In some implementations of the process shown in FIG. 1, the transportation mode segment classifier and the role classifier determine the sequence of transportation modes from data collected by a mobile device 13 typically carried by the participant in a trip. Most modern mobile devices, such as smartphones, are equipped with a variety of sensors 32 that measure physical properties relevant to movements of a participant and to a transportation mode being used by the participant and other aspects of the environment. The data 34 (which can be considered part of the recordings 23) that result from measurements of these physical properties by the sensors of the mobile devices can be combined with higher-level situational descriptions 36 (from the mobile device and in some cases other data sources 38) about a trip, such as location, time, trajectory, or user, resulting in a rich set of features 20 used by the transportation mode segment classifier and role classifier. In some instances, some or all of the situational descriptions can be considered part of the recordings 23”);
determining a plurality of features based on a first comparison between the location dataset and a bus route dataset, wherein determining the plurality of features comprises:
(see, Paragraph [0051] “As shown in FIG. 1, during trip mode classification, each segment of data of a recording is classified into one of the transportation modes. From the sensors and other data sources described earlier, various features 156 are computed for each segment, which are used as inputs by the transportation mode classifier. The features can include one or a combination or two or more of the following.”);
in response to generating the plurality of candidate route segments (see at least Paragraph [0070]: “Public transit agencies publish transportation schedules and routes to help transit application developers, often in a common data format such as General Transit Feed Specification. The data can contain routes, stop locations, and timetables, among others, which can be used to identify if a close match exists with the data of the segment”);
generating a plurality of candidate route segments based on the location dataset (see, (e.g. map matching for determining road segments using location and sensor measurements, see [0064], [0066] and [0070] excerpts below); generating a set of candidate bus routes based on the plurality of candidate route segments (see, (e.g., public transit routes and identification of close matches within the segments, see [0064], [0066] and [0070] excerpts below); selecting a subset of the set of candidate bus routes based on contextual information (see, (e.g., matching of time series with routes, see [0064], [0066] and [0070] excerpts below); and determining the plurality of features based on the subset of the set of candidate bus routes (e.g., The top matches within the predefined match threshold are selected as features, see [0064], [0066] and [0070] excerpts below); and
[0066] “Map-match information: Telematics applications often involve a task of “map matching”, a process for determining the trajectory of a vehicle on route networks given location and sensor measurements provided by a mobile device. If the map matching information is available, the classifier can collect two categories of information: 1. the types of the matched route segments, including the category of a route (e.g. road, train track, or ferry route) and the class of a route within the specific route category (e.g. highway vs local street for category road). For each segment, the most dominant category and class, and their relative fractions, are used as features. 2. how much the map matched output deviates from input location measurements (i.e., GPS or network location). A high deviation makes the classifier prefer certain transportation modes that do not have to travel on a route network (e.g., bike, foot, boat, off-road) than others (e.g., car on a road network).”;
[0070] “Public transit information: Public transit agencies publish transportation schedules and routes to help transit application developers, often in a common data format such as General Transit Feed Specification. The data can contain routes, stop locations, and timetables, among others, which can be used to identify if a close match exists with the data of the segment, and what the closest matches are. If one or more close matches are found, then the associated transportation modes can be directly used as features for the classifier. To perform matching, the technology encodes the trajectory of the trip to be classified by extracting stop locations (from speed and accelerometer and gyroscope measurements) and corresponding timestamps, and represent them as a series of timestamped stop events. The series is matched against routes in public transit data (with stop times and locations embedded in it). Dynamic time warping can be used to tolerate minor differences and omissions of stop times and locations. The top matches within the predefined match threshold are selected as features.”
See also: [0064]: “Proximity to transport stations: The distance from locations associated with a recording to nearby transport stations, including airports, train stations, and bus stops can be measured as a feature. For each of the first and last GPS locations of a recording, we find the closest station of each type (e.g., airports, bus stations, or train stations). Then, the closest distance is taken as a feature representing proximity to that type of transport station. For example, when the first GPS location is considered, the distance from it to the closest airport is taken as a feature representing proximity of the trip start to the nearest airport. This is repeated for every combination of the first or the last GPS coordinates and different types of transport stations. Finally, among these proximity distances, the smallest one can be used to form another feature representing which type of transport station is closest among different choices of transport modes.”);
classifying the trip based on the plurality of features (see, Paragraph [0072] “Given the features acquired or computed as described above, the transportation mode segment classifier predicts the most likely transportation mode of the given segment and a confidence measure ranging from 0 to 1. The classifier uses gradient boosting with decision trees as base learners (“gradient boosted trees”). A classification model is built by fitting a large number of decision trees on training data, gradually one by one, until a chosen loss function (“training error”) is minimized, or the number of trees being fitted reaches a predefined number of rounds.”;
[0075] “When a recording for a trip is presented from a mobile device for trip mode classification, a production server loads and applies the stored final classification model to the trip, producing segment-by-segment trip mode identifications. In some implementations, the final classification model can be transferred to the mobile device, which loads and applies it to identify a transportation mode for the trip. In some implementations, the computational work of applying the model to the data of the recordings can be split between the server and the mobile device.”); and
triggering, based on the classification of the trip, an action at the mobile device, wherein the action facilitates at least one of provision of content or a mobile device service (see, Paragraphs [0044]: “The sequence of transportation mode labels or identifications of roles or both can be provided as feeds to a wide variety of consuming applications and users, including individual users and enterprises and their applications.”; and [0099]: “As mentioned earlier, part or all of the transportation mode and role classification steps may be done on the phone instead of or in conjunction with the server”).
As to [claim 2]. Park discloses the method of Claim 1. Park discloses
wherein the location dataset is determined at a first time, wherein the action is triggered in substantially real time relative to the first time (see, least Paragraph [0038]: “the transportation mode segment classifier and the role classifier determine the sequence of transportation modes from data collected by a mobile device 13 typically carried by the participant in a trip. Most modern mobile devices, such as smartphones, are equipped with a variety of sensors 32 that measure physical properties relevant to movements of a participant and to a transportation mode being used by the participant and other aspects of the environment. The data 34 (which can be considered part of the recordings 23) that result from measurements of these physical properties by the sensors of the mobile devices can be combined with higher-level situational descriptions 36 (from the mobile device and in some cases other data sources 38) about a trip, such as location, time”; and [0041]: “One or more of the technology stages of FIG. 1 may be applied live (in real time) during the collection of sensor data (e.g., by running every few minutes using the data collected after the most recent application of the data, even before a recording or a trip has been completed). Or it may be applied after an entire recording has been made. Each recording (a sequence of recordings may be thought of as a single larger recording) can include metadata (e.g., part of or in addition to the situational descriptions 36 in FIG . 1) describing the recording”).
As to [claim 3], Park discloses the method of Claim 1.
Park discloses
wherein the set of candidate bus routes comprises a candidate route comprising a transfer between public transit lines (see, Paragraph [0070]: “Public transit information : Public transit agencies publish transportation schedules and routes to help transit application developers , often in a common data format such as General Transit Feed Specification . The data can contain routes , stop locations , and timetables , among others , which can be used to identify if a close match exists with the data of the segment , and what the closest matches are . If one or more close matches are found , then the associated transportation modes can be directly used as features for the classifier . To perform matching , the technology encodes the trajectory of the trip to be classified by extracting stop locations (from speed and accelerometer and gyroscope measurements ) and corresponding timestamps, and represent them as a series of timestamped stop events . The series is matched against routes in public transit data ( with stop times and locations embedded in it).”; 0064]: “Proximity to transport stations: The distance from locations associated with a recording to nearby transport stations, including airports, train stations, and bus stops can be measured as a feature. For each of the first and last GPS locations of a recording, we find the closest station of each type (e.g., airports, bus stations, or train stations). Then, the closest distance is taken as a feature representing proximity to that type of transport station. For example, when the first GPS location is considered, the distance from it to the closest airport is taken as a feature representing proximity of the trip start to the nearest airport. This is repeated for every combination of the first or the last GPS coordinates and different types of transport stations. Finally, among these proximity distances, the smallest one can be used to form another feature representing which type of transport station is closest among different choices of transport modes.”);).
As to [claim 4], Park discloses the method of Claim 1.
Park discloses
wherein classifying the trip comprises:
classifying the trip based on a satisfaction of a trip length condition and satisfaction of a respective probability condition for the plurality of features (see, Paragraphs [0034]: “The technology is capable of identifying a sequence of one or more transportation modes, participant roles, or other trip types from a recording representing a single trip or multiple trips. The modes of transportation that can be identified include: car, airplane, bus, train, bike, boat, motorcycle, off-road, and foot, for example. In off-road mode, trips are on unsurfaced roads, often using an all-terrain vehicle (ATV) or skis” and [0066]: “If the map matching information is available, the classifier can collect two categories of information: 1. The types of the matched route segments, including the category of a route ( e.g. road, train track, or ferry route) and the class of a route within the specific route category ( e.g. highway vs local street for category road). For each segment, the most dominant category and class, and their relative fractions, are used as features. 2. how much the map matched output deviates from input location measurements (i.e., GPS or network location). A high deviation makes the classifier prefer certain transportation modes that do not have to travel on a route network (e.g., bike, foot, boat, off-road) than others (e.g., car on a road network)”; [0070] Public transit information: Public transit agencies publish transportation schedules and routes to help transit application developers, often in a common data format such as General Transit Feed Specification. The data can contain routes, stop locations, and timetables, among others, which can be used to identify if a close match exists with the data of the segment, and what the closest matches are. If one or more close matches are found, then the associated transportation modes can be directly used as features for the classifier. To perform matching, the technology encodes the trajectory of the trip to be classified by extracting stop locations (from speed and accelerometer and gyroscope measurements) and corresponding timestamps, and represent them as a series of timestamped stop events. The series is matched against routes in public transit data (with stop times and locations embedded in it). Dynamic time warping can be used to tolerate minor differences and omissions of stop times and locations. The top matches within the predefined match threshold are selected as features.; [0072] Given the features acquired or computed as described above, the transportation mode segment classifier predicts the most likely transportation mode of the given segment and a confidence measure ranging from 0 to 1. [0080] In this HMM, the state space comprises the possible transportation modes. The observation space is based on the segment transportation mode identification label and its associated confidence (probability). The probability value is converted to a discrete value by quantization and augmented using the transportation mode prediction label to produce a new observation symbol used as an input for HMM. ).
As to [claim 5], Park discloses the method of Claim 1.
Park discloses
wherein the trip is classified with a machine-learning- based classification model (see at least Paragraph [0073]: “The technology preprocesses the training data including sensor data, on-device activity predictions, map match results, and user history data. The preprocess step includes data cleansing, outlier removal, timestamp correction, and other steps to bring input training data into a form ready for processing. The preprocessed training data for each trip is segmented as described earlier. Each resulting segment becomes an individual data point in the training set for the segment classifier. Features are extracted for each segment as described earlier. The features computed from all of the data points (segments) constitute a feature matrix, which the classifier training algorithm takes as input. The training algorithm trains the classifier using the feature matrix and evaluates the classifier using cross validation, by checking the expected performance and varying training parameters as necessary. Once a good classification model is found, a final classifier is trained using the best model found using all training data available. The final model is saved in persistent storage so that it can be loaded later and used for prediction of transportation modes for recordings of new trips”).
As to [claim 6], Park discloses the method of Claim 1.
Park discloses
wherein classifying the trip (see, Abstract; and Paragraph [0012]: “Identifying a type of a trip that is captured in a recording can be very useful for a variety of purposes. Here we describe technology that uses telematics data (e.g., recordings of sensor data captured on a mobile device, and other information) to identify (for example, classify) such a type, for example, a mode of transportation or a role of a person being transported or other types.”) comprises a multi-class classification using a heuristic, tree-based selection process (see, Paragraphs [0072]: “A classification model is built by fitting a large number of decision trees on training data, gradually one by one, until a chosen loss function ("training error") is minimized, or the number of trees being fitted reaches a predefined number of rounds.”; and [0074]: “The strategies include one or a combination of two or more of: limiting the depth of the individual decision trees, limiting the maximum number of trees in the ensemble , controlling the learning rate of sub sequent base learners, imposing a minimum number of samples required for the decision tree to split further, and random subsampling of training examples and features during decision tree fitting process”).
As to [claim 7], Park discloses the method of Claim 1.
Park discloses
wherein determining a plurality of features comprises:
determining a set of stop locations based on the location dataset (see, Paragraph [0054]: “Buses often make frequent stops the durations and intervals of which are fairly uniform compared to other transportation modes. The distribution of acceleration and jerk give complementary evidences in addition to acceleration and gyroscope based features for distinguishing heterogeneous types of motorized vehicles of which structure, size and weight differ.”); and
comparing the set of stop locations to bus stops of the bus route dataset (see at least Paragraph [0070]: “Public transit information: Public transit agencies publish transportation schedules and routes to help transit application developers, often in a common data format such as General Transit Feed Specification. The data can contain routes, stop locations, and timetables, among others, which can be used to identify if a close match exists with the data of the segment, and what the closest matches are. If one or more close matches are found, then the associated transportation modes can be directly used as features for the classifier. To perform matching, the technology encodes the trajectory of the trip to be classified by extracting stop locations (from speed and accelerometer and gyroscope measurements) and corresponding timestamps, and represent them as a series of timestamp ed stop events. The series is matched against routes in public transit data (with stop times and locations embedded in it). Dynamic time warping can be used to tolerate minor differences and omissions of stop times and locations. The top matches within the predefined match threshold are selected as features”).
As to [claim 8], Park discloses the method of Claim 7.
Park discloses
wherein the set of stop locations is determined based on inertial data captured by an inertial sensor on the mobile device (see, Paragraphs [0038]-[0039]--- Examiner notes that Park does not explicitly discloses what type of sensors 32 on the mobile device, however, the sensors are performing the same capabilities an inertial sensor capturing inertial data---; see also [0053] for disclosure of inertial sensors).
As to [claim 9], Park discloses the method of Claim 7.
Park discloses
wherein, the plurality of features comprises a score determined based on a proximity of a trip end point to a bus stop (see, Paragraph [0064]: “The distance from locations associated with a recording to nearby transport stations, including airports, train stations, and bus stops can be measured as a feature. For each of the first and last GPS locations of a recording, we find the closest station of each type ( e.g., airports, bus stations, or train stations). Then, the closest distance is taken as a feature representing proximity to that type of transport station. For example, when the first GPS location is considered, the distance from it to the closest airport is taken as a feature representing proximity of the trip start to the nearest airport. This is repeated for every combination of the first or the last GPS coordinates and different types of transport stations. Finally, among these proximity distances, the smallest one can be used to form another feature representing which type of transport station is closest among different choices of transport modes”; [0072] Given the features acquired or computed as described above, the transportation mode segment classifier predicts the most likely transportation mode of the given segment and a confidence measure ranging from 0 to 1.).
As to [claim 10]. Park discloses the method of Claim 1.
Park discloses
additionally comprising determining a second plurality of features based on a comparison of the location dataset and a railway dataset, and wherein classifying the (see, Paragraph [0065]: “On-device activity recognition output: Modem smartphones (and other mobile devices) and mobile operating systems offer an on-device activity recognition facility to aid contextual applications to collect user activities or react to a change in user's physical activity. They often provide coarse-grained information on what the current physical activity of the user is, for example, whether the user is walking, in a vehicle (without necessarily identifying the transportation mode), or still. These on-device activity predictions, although often coarse-grained and not highly accurate, can be used as supporting evidence for transportation mode classification, if available. The raw activity recognition output can be used as it is, or summarized to produce a distribution of distinct activity labels”; and [0070]: “If one or more close matches are found, then the associated transportation modes can be directly used as features for the classifier. To perform matching, the technology encodes the trajectory of the trip to be classified by extracting stop locations (from speed and accelerometer and gyroscope measurements) and corresponding timestamps, and represent them as a series of timestamped stop events.”; see also [0034], [0042], [0064], [0066], [0068]).
As to [claim 11], Park discloses the method of Claim 1.
Park discloses
wherein the contextual information comprises: a direction of traversal on a roadway (see, Paragraph [0066] Map-match information: Telematics applications often involve a task of “map matching”, a process for determining the trajectory of a vehicle on route networks given location and sensor measurements provided by a mobile device. If the map matching information is available, the classifier can collect two categories of information: 1. the types of the matched route segments, including the category of a route (e.g. road, train track, or ferry route) and the class of a route within the specific route category (e.g. highway vs local street for category road). [0070]: “The data can contain routes, stop locations, and timetables, among others, which can be used to identify if a close match exists with the data of the segment, and what the closest matches are. If one or more close matches are found, then the associated transportation modes can be directly used as features for the classifier. To perform matching, the technology encodes the trajectory of the trip to be classified by extracting stop locations (from speed and accelerometer and gyroscope measurements)”); and a bus route schedule (see, Paragraph [0070]: “Public transit information: Public transit agencies publish transportation schedules and routes to help transit application developers, often in a common data format such as General Transit Feed Specification. The data can contain routes, stop locations, and timetables, among others, which can be used to identify if a close match exists with the data of the segment, and what the closest matches are. If one or more close matches are found, then the associated transportation modes can be directly used as features for the classifier.”).
As to [claim 12], Park discloses the method of Claim 1.
Park discloses
wherein the plurality of features comprises a dynamic time warping (DTW) similarity score, wherein the DTW similarity score is determined by: generating a candidate bus route comprising a series of route segments within the bus route dataset; and determining the DTW similarity score for the candidate bus route and the location dataset (see, Paragraphs [0070]: “Dynamic time warping can be used to tolerate minor differences and omissions of stop times and locations. The top matches within the predefined match threshold are selected as features.”; and [0088]: “Trip similarity: This feature examines if the current trip has spatial or patio-temporal similarity to trips in one of the trip groups. For each trip group, every trip within the group is compared to the current trip by origin and destination, travel path and time of day, in which the current trip and the trip in the group are considered similar if the matches are successful within predefined thresholds. A similarity of each trip group is computed as a fraction of similar trips in the group and is provided to the role classifier as a feature representing trip similarity to the group”).
Regarding [claim 13], Park discloses a method for classification of vehicle trip transportation modality (see, Abstract; and Paragraph [0051]: “As shown in FIG. 1, during trip mode classification, each segment of data of a recording is classified into one of the transportation modes. From the sensors and other data sources described earlier, various features 156 are computed for each segment, which are used as inputs by the transportation mode classifier.”) performed by a processing system comprising one or more processors and one or more memories coupled with the one or more processors (see, Paragraph [0003]: “In some contexts , telematics applications involve collecting , storing , and processing data from sensors in vehicles. In particular, mobile telematics (sometimes known as “smartphone telematics” ) uses mobile sensing technologies to collect , store , and process data from built-in or external sensors of a mobile device, such as a smartphone” and Paragraph [0075]: “At the end of the training, the final classification model is stored in a permanent storage and deployed to production servers. When a recording for a trip is presented from a mobile device for trip mode classification , a production server loads and applies the stored final classification model to the trip , producing segment- by-segment trip mode identifications. In some implementations, the final classification model can be transferred to the mobile device, which loads and applies it to identify a transportation mode for the trip. In some implementations, the computational work of applying the model to the data of the recordings can be split In between the server and the mobile device”), comprising:
receiving a trip dataset comprising location data collected with a location sensor of a mobile user device (see, Paragraphs [0035]: “the technology can use a variety of mobile sensor data included in the recordings to extract features useful for identifying transportation modes, participant roles, or other trip types, and non-trip phantom recordings”; and [0052]: “Descriptive statistics: For each sensor type, a set of descriptive statistics is computed for the data that belongs to a given segment, including mean, standard deviation, coefficient of variation… For multi-axial sensors, such as tri-axial accelerometer or tri-axial gyroscope, the statistics are computed from the magnitude of instantaneous vector measurements.”);
determining a set of features by comparing the trip dataset to a transit dataset (see, Paragraph [0070]: “Public transit information: Public transit agencies publish transportation schedules and routes to help transit application developers, often in a common data format such as General Transit Feed Specification. The data can contain routes, stop locations, and timetables, among others, which can be used to identify if a close match exists with the data of the segment, and what the closest matches are. If one or more close matches are found, then the associated transportation modes can be directly used as features for the classifier. To perform matching, the technology encodes the trajectory of the trip to be classified by extracting stop locations (from speed and accelerometer and gyroscope measurements) and corresponding timestamps, and represent them as a series of timestamp ed stop events. The series is matched against routes in public transit data (with stop times and locations embedded in it). Dynamic time warping can be used to tolerate minor differences and omissions of stop times and locations. The top matches within the predefined match threshold are selected as features”),comprising:
determining a candidate transit route comprising a series of route segments using the transit dataset (see, Paragraph [0070]: “Public transit agencies publish transportation schedules and routes to help transit application developers, often in a common data format such as General Transit Feed Specification. The data can contain routes, stop locations, and timetables, among others, which can be used to identify if a close match exists with the data of the segment”)); and
in response to determining the candidate route (see at least Paragraph [0070]: “Public transit agencies publish transportation schedules and routes to help transit application developers, often in a common data format such as General Transit Feed Specification. The data can contain routes, stop locations, and timetables, among others, which can be used to identify if a close match exists with the data of the segment”);
determining a dynamic time warping [DTW] similarity score for the candidate transit route and the trip dataset, wherein the set of features comprises the DTW similarity score (see, Paragraphs [0070]: “Dynamic time warping can be used to tolerate minor differences and omissions of stop times and locations. The top matches within the predefined match threshold are selected as features.”; and [0088]: “Trip similarity: This feature examines if the current trip has spatial or spatio-temporal similarity to trips in one of the trip groups. For each trip group, every trip within the group is compared to the current trip by origin and destination, travel path and time of day, in which the current trip and the trip in the group are considered similar if the matches are successful within predefined thresholds. A similarity of each trip group is computed as a fraction of similar trips in the group and is provided to the role classifier as a feature representing trip similarity to the group”);
based on the set of features, and the DTW similarity score, classifying the vehicle trip as a transit trip (see, Paragraphs [0072] “Given the features acquired or computed as described above, the transportation mode segment classifier predicts the most likely transportation mode of the given segment and a confidence measure ranging from 0 to 1. The classifier uses gradient boosting with decision trees as base learners (“gradient boosted trees”). A classification model is built by fitting a large number of decision trees on training data, gradually one by one, until a chosen loss function (“training error”) is minimized, or the number of trees being fitted reaches a predefined number of rounds.”; and [0075] “When a recording for a trip is presented from a mobile device for trip mode classification, a production server loads and applies the stored final classification model to the trip, producing segment-by-segment trip mode identifications. In some implementations, the final classification model can be transferred to the mobile device, which loads and applies it to identify a transportation mode for the trip. In some implementations, the computational work of applying the model to the data of the recordings can be split between the server and the mobile device.”); and
triggering, based on the classification of the vehicle trip as a transit trip, wherein the action facilitates at least one of provision of content or a mobile device service (see, Paragraphs [0044]: “The sequence of transportation mode labels or identifications of roles or both can be provided as feeds to a wide variety of consuming applications and users, including individual users and enterprises and their applications.”; and [0099]: “As mentioned earlier, part or all of the transportation mode and role classification steps may be done on the phone instead of or in conjunction with the server.”).
As to [claim 14], Park discloses the method of Claim 13.
Park discloses
wherein the vehicle trip is classified based on satisfaction of a trip length condition associated with the trip dataset (see, Paragraphs [0034]: “The technology is capable of identifying a sequence of one or more transportation modes, participant roles, or other trip types from a recording representing a single trip or multiple trips. The modes of transportation that can be identified include: car, airplane, bus, train, bike, boat, motorcycle, off-road, and foot, for example. In off-road mode, trips are on unsurfaced roads, often using an all-terrain vehicle (ATV) or skis” and [0066]: “If the map matching information is available, the classifier can collect two categories of information: 1. The types of the matched route segments, including the category of a route ( e.g. road, train track, or ferry route) and the class of a route within the specific route category ( e.g. highway vs local street for category road). For each segment, the most dominant category and class, and their relative fractions, are used as features. 2. how much the map matched output deviates from input location measurements (i.e., GPS or network location). A high deviation makes the classifier prefer certain transportation modes that do not have to travel on a route network (e.g., bike, foot, boat, off-road) than others (e.g., car on a road network)”; [0070] Public transit information: Public transit agencies publish transportation schedules and routes to help transit application developers, often in a common data format such as General Transit Feed Specification. The data can contain routes, stop locations, and timetables, among others, which can be used to identify if a close match exists with the data of the segment, and what the closest matches are. If one or more close matches are found, then the associated transportation modes can be directly used as features for the classifier. To perform matching, the technology encodes the trajectory of the trip to be classified by extracting stop locations (from speed and accelerometer and gyroscope measurements) and corresponding timestamps, and represent them as a series of timestamped stop events. The series is matched against routes in public transit data (with stop times and locations embedded in it). Dynamic time warping can be used to tolerate minor differences and omissions of stop times and locations. The top matches within the predefined match threshold are selected as features.; [0072] Given the features acquired or computed as described above, the transportation mode segment classifier predicts the most likely transportation mode of the given segment and a confidence measure ranging from 0 to 1. [0080] In this HMM, the state space comprises the possible transportation modes. The observation space is based on the segment transportation mode identification label and its associated confidence (probability). The probability value is converted to a discrete value by quantization and augmented using the transportation mode prediction label to produce a new observation symbol used as an input for HMM.).
As to [claim 15]. Park discloses the method of Claim 13.
Park discloses
wherein the set of features is determined with a pretrained Hidden Markov Model (HMM) (see, Paragraph [0037]: “The output of the transportation mode segment classifier, a sequence 22 of transportation mode prediction labels 24 (each label identifying a transportation mode) and probabilities 26 (each representing the probability that a given label is correct), is smoothed using a temporal smoothing model 28 based on a Hidden Markov Model (HMM)”.
As to [claim 17], Park discloses the method of Claim 13, wherein determining the candidate transit route comprises:
generating the series of route segments based on the trip dataset (see, Paragraph [0051] “As shown in FIG. 1, during trip mode classification, each segment of data of a recording is classified into one of the transportation modes. From the sensors and other data sources described earlier, various features 156 are computed for each segment, which are used as inputs by the transportation mode classifier. The features can include one or a combination or two or more of the following ”);
generating a set of candidate transit routes based on the series of route segments (see, Paragraph [0064]: “The distance from locations associated with a recording to nearby transport stations, including airports, train stations, and bus stops can be measured as a feature. For each of the first and last GPS locations of a recording, we find the closest station of each type (e.g., airports, bus stations, or train stations). Then, the closest distance is taken as a feature representing proximity to that type of transport station”; [0066] “Map-match information: Telematics applications often involve a task of “map matching”, a process for determining the trajectory of a vehicle on route networks given location and sensor measurements provided by a mobile device. If the map matching information is available, the classifier can collect two categories of information: 1. the types of the matched route segments, including the category of a route (e.g. road, train track, or ferry route) and the class of a route within the specific route category (e.g. highway vs local street for category road). For each segment, the most dominant category and class, and their relative fractions, are used as features. 2. how much the map matched output deviates from input location measurements (i.e., GPS or network location). A high deviation makes the classifier prefer certain transportation modes that do not have to travel on a route network (e.g., bike, foot, boat, off-road) than others (e.g., car on a road network).”; Paragraph [0070]: “Public transit agencies publish transportation schedules and routes to help transit application developers, often in a common data format such as General Transit Feed Specification. The data can contain routes, stop locations, and timetables, among others, which can be used to identify if a close match exists with the data of the segment” and [0078]: “The segment classifier described above identifies a transportation mode for each segment in a recording of a trip and generates a corresponding probability measure. An individual identification may not be accurate, for example, due to lack of information within the segment data or due to imperfect segment transportation mode classification or a combination of them. Such identifications, taken one by one for the segments, do not benefit from temporal regularities and dynamics that span successive segments”); and
selecting the candidate transit route from the set of candidate transit routes (see, Paragraph [0072] “Given the features acquired or computed as described above, the transportation mode segment classifier predicts the most likely transportation mode of the given segment and a confidence measure ranging from 0 to 1.; [0075] “When a recording for a trip is presented from a mobile device for trip mode classification, a production server loads and applies the stored final classification model to the trip, producing segment-by-segment trip mode identifications).
As to [claim 18], Park discloses the method of Claim 17.
Park discloses
wherein the candidate transit route is selected based on contextual information (see, Paragraphs 0051] “As shown in FIG. 1, during trip mode classification, each segment of data of a recording is classified into one of the transportation modes. From the sensors and other data sources described earlier, various features 156 are computed for each segment, which are used as inputs by the transportation mode classifier. The features can include one or a combination or two or more of the following ; [0065]: “On-device activity recognition output: Modem smartphones (and other mobile devices) and mobile operating systems offer an on-device activity recognition facility to aid contextual applications to collect user activities or react to a change in user's physical activity. They often provide coarse-grained information on what the current physical activity of the user is, for example, whether the user is walking, in a vehicle (without necessarily identifying the transportation mode), or still. These on-device activity predictions, although often coarse-grained and not highly accurate, can be used as supporting evidence for transportation mode classification, if available. The raw activity recognition output can be used as it is, or summarized to produce a distribution of distinct activity labels”; [0070]; and [0087]: “For each type of identified key location, the frequencies of role as driver and passenger are tracked. For the classification of a current trip, the role classifier first determines if the current trip matches any of the user's identified key locations. If there is a match, the relative frequency of driver role over passenger role enables user-specific location-based contextual bias”).
As to [claim 19], Park discloses the method of Claim 13.
Park discloses
wherein the vehicle trip is classified using a heuristic, tree- based classification process (see, Paragraphs [0072]: “A classification model is built by fitting a large number of decision trees on training data, gradually one by one, until a chosen loss function ("training error") is minimized, or the number of trees being fitted reaches a predefined number of rounds.”; and [0074]: “The strategies include one or a combination of two or more of: limiting the depth of the individual decision trees, limiting the maximum number of trees in the ensemble , controlling the learning rate of sub sequent base learners, imposing a minimum number of samples required for the decision tree to split further, and random subsampling of training examples and features during decision tree fitting process”).
As to [claim 20], Park discloses the method of Claim 19.
wherein classifying the vehicle trip (see Abstract; and Paragraph [0012]: “Identifying a type of a trip that is captured in a recording can be very useful for a variety of purposes. Here we describe technology that uses telematics data (e.g., recordings of sensor data captured on a mobile device, and other information) to identify (for example, classify) such a type, for example, a mode of transportation or a role of a person being transported or other types.”) comprises determining a decision parameter associated with a transit transportation class based on a joint probability associated with the set of features and a differential comparison feature of a second set of features, wherein the second set of features is determined by comparing the trip dataset to a roadway driving dataset (see at least Paragraph [0078]: “The segment classifier described above identifies a transportation mode for each segment in a recording of a trip and generates a corresponding probability measure. An individual identification may not be accurate, for example, due to lack of information within the segment data or due to imperfect segment transportation mode classification or a combination of them. Such identifications, taken one by one for the segments, do not benefit from temporal regularities and dynamics that span successive segments. For example, people may not switch back and forth frequently between different transportation modes, and certain mode transitions may occur more frequently in general than the others. For example, it is not likely to switch from airplane to car without foot in between; and the probability of a transition from airplane to bike is much less than to car” and [0080]: “In this HMM, the state space comprises the possible transportation modes. The observation space is based on the segment transportation mode identification label and its associated confidence (probability). The probability value is converted to a discrete value by quantization and augmented using the transportation mode prediction label to produce a new observation symbol used as an input for HMM.” ***Examiner interprets that reference discloses the feature “a railway transportation class” but is not limited and possess other plurality of features that classify each mode that is associated with a decision parameter probability***; see also e.g. map matching for determining road segments using location and sensor measurements, see [0064], [0066] and [0070]).
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claim(s) 16 is/are rejected under 35 U.S.C. 103 as being unpatentable over Park in view of Bleasdale-Shepherd et al. (US 2021/0038979), hereinafter, referred to as “Bleasdale-Shepherd”.
As to [claim 16], Park discloses the method of Claim 13.
Park discloses
wherein the vehicle trip is classified using a multi-class (see, Paragraph [0042]: “an output of the transportation mode classifier 18 is a series of transportation modes along with classification confidences 26. In some implementations, for each segment in a sequence, the classifier assigns one of the following transportation mode labels to it along with a confidence value: car, airplane, train, bus, boat, bike, motorcycle, foot (walking or running), off-road”), (see at least Paragraph [0072]: “A classification model is built by fitting a large number of decision trees on training data, gradually one by one, until a chosen loss function ("training error") is minimized, or the number of trees being fitted reaches a predefined number of rounds.”; and [0075]: “At the end of the training, the final classification model is stored in a permanent storage and deployed to production servers. When a recording for a trip is presented from a mobile device for trip mode classification, a production server loads and applies the stored final classification model to the trip, producing segment-by-segment trip mode identifications. In some implementations, the final classification model can be transferred to the mobile device, which loads and applies it to identify a transportation mode for the trip.”), however, Park does not explicitly disclose …tree-based classification model comprising a Bayesian network.
However, Bleasdale-Shepherd teaches
…tree-based classification model comprising a Bayesian network (see at least Paragraph [0038]: “The trained machine learning model(s) 216 may represent a single model or an ensemble of base-level machine learning models, and may be implemented as any type of machine learning model 216. For example, suitable machine learning models 216 for use with the techniques and systems described herein include, without limitation, neural networks, tree-based models, support vector machines (SVMs), kernel methods, random forests, splines (e.g., multivariate adaptive regression splines), hidden Markov model (HMMs), Kalman filters (or enhanced Kalman filters), Bayesian networks (or Bayesian belief networks), expectation maximization, genetic algorithms, linear regression algorithms, nonlinear regression algorithms, logistic regression-based classification models”).
Accordingly, it would have been obvious to one of ordinary skill in the art before the filing of the invention to further modify Park by combining ….tree-based classification model comprising a Bayesian network as taught by Bleasdale-Shepherd. One would be motivated to make this modification to achieve an improved classification system because an ensemble of models can operate as a committee of individual machine learning models that is collectively “smarter” than any individual machine learning model of the ensemble (see Bleasdale-Shepherd [0038]).
Double Patenting
The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on nonstatutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. See MPEP § 2146 et seq. for applications not subject to examination under the first inventor to file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be accompanied by a reply requesting reconsideration of the prior Office action. Even where the NSDP rejection is provisional the reply must be complete. See MPEP § 804, subsection I.B.1. For a reply to a non-final Office action, see 37 CFR 1.111(a). For a reply to final Office action, see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c) may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13.
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The actual filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/apply/applying-online/eterminal-disclaimer.
Claim(s) 1-2, 4-17, and 19-20 are rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1-18 of U.S. Patent No. US 12,056,633 B2. Although the claims at issue are not identical, they are not patentably distinct from each other because the scope of the inventions for both applications are similar as claim(s) 1-2, 4-17, and 19-20 of the instant application and claim(s) 1-18 are directed to similar limitations as presented below.
Regarding claim(s) 1 of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches:
A method, performed by a processing system comprising one or more processors and one or more memories coupled with the one or more processors, comprising:
detecting a trip associated with vehicular transportation of a mobile device;
determining a location dataset representing locations of the mobile device;
determining a plurality of features based on a first comparison between the location dataset and a bus route dataset, wherein determining the plurality of features comprises:
generating a plurality of candidate route segments based on the location dataset;
in response to generating the plurality of candidate route segments:
generating a set of candidate bus routes based on the plurality of candidate route segments;
selecting a subset of the set of candidate bus routes based on contextual information; and
determining the plurality of features based on the subset of the set of candidate bus routes; classifying the trip based on the plurality of features; and
triggering, based on the classification of the trip, an action at the mobile device, wherein the action facilitates at least one of provision of content or a mobile device service.
As to claim(s) 2 of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: wherein the location dataset is determined at a first time, wherein the action is triggered in substantially real time relative to the first time.
As to claim 4 of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: wherein classifying the trip comprises: classifying the trip based on a satisfaction of a trip length condition and satisfaction of a respective probability condition for the plurality of features.
As to claim 5 of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: wherein the trip is classified with a machine-learning- based classification model.
As to claim 6 of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: wherein classifying the trip comprises a multi-class classification using a heuristic, tree-based selection process.
As to claim 7, of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: wherein determining a plurality of features comprises:" determining a set of stop locations based on the location dataset; and " comparing set of set of the stop locations to bus stops of the bus route dataset.
As to claim 8, of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: wherein the set of stop locations is determined based on inertial data captured by an inertial sensor on the mobile device.
As to claim 9, of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: wherein, the plurality of features comprises a score determined based on a proximity of a trip end point to a bus stop.
As to claim 10, of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: additionally comprising determining a second plurality of features based on a comparison of the location dataset and a railway dataset, and wherein classifying the
As to claim 11, of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: wherein the contextual information comprises: a direction of traversal on a roadway; and a bus route schedule.
As to claim 12, of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: wherein the plurality of features comprises a dynamic time warping (DTW) similarity score, wherein the DTW similarity score is determined by:" generating a candidate bus route comprising a series of route segments within the bus route dataset; and " determining the DTW similarity score for the candidate bus route and the location dataset.
Regarding claim 13, of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches:
A method for classification of vehicle trip transportation modality comprising:" receiving a trip dataset comprising location data collected with a location sensor of a mobile user device; " determining a set of features by comparing the trip dataset to a transit dataset, comprising: " determining a candidate transit route comprising a series of route segments using the transit dataset; and" determining a dynamic time warping [DTW] similarity score for the candidate transit route and the trip dataset, wherein the set of features comprises the DTW similarity score; " based on the set of features, classifying the vehicle trip as a transit trip; and " based on the classification of the vehicle trip as a transit trip, triggering an action at the mobile user device.
As to claim 14, of the instant application, claim(s) 1-20 of U.S. Patent No. US 12,056,633 B2 teaches: wherein the vehicle trip is classified based on satisfaction of a trip length condition associated with the trip dataset.
As to claim 15, of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: wherein the set of features is determined with a pretrained Hidden Markov Model (HMM).
As to claim 16, of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: wherein the vehicle trip is classified using a multi-class, tree-based classification model comprising a Bayesian network.
As to claim 17, of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: wherein determining the candidate transit route comprises:" generating the series of route segments based on the trip dataset; " generating a set of candidate transit routes based on the series of route segments; and " selecting the candidate transit route from the set of candidate transit routes.
As to claim 19, of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: wherein the vehicle trip is classified using a heuristic, tree- based classification process.
As to claim 20, of the instant application, claim(s) 1-18 of U.S. Patent No. US 12,056,633 B2 teaches: wherein classifying the vehicle trip comprises determining a decision parameter associated with a transit transportation class based on a joint probability associated with the set of features and a differential comparison feature of a second set of features, wherein the second set of features is determined by comparing the trip dataset to a roadway driving dataset.
Terminal Disclaimer
A terminal disclaimer may be effective to overcome a nonstatutory double patenting rejection over a reference patent (37 CFR 1.321(b) and (c)). A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be accompanied by a reply requesting reconsideration of the prior Office action. Even where the NSDP rejection is provisional, the reply must be complete. MPEP § 804, subsection I.B.1. For a reply to a non-final Office action, see 37 CFR 1.111(a). For a reply to final Office action, see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c) may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13.
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/PatentForms. The filing date of the application will determine what form should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/TerminalDisclaimer.
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 BAKARI UNDERWOOD whose telephone number is (571)272-8462. The examiner can normally be reached M - F 8:00 TO 4:30.
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, Abby Flynn can be reached (571) 272-9855. 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.
/B.U./Examiner, Art Unit 3663
/ABBY J FLYNN/Supervisory Patent Examiner, Art Unit 3663