DETAILED ACTION
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Notice for all US Patent Applications filed on or after March 16, 2013
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 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.
Status of the Claims
This communication is in response to communications received on 3/27/25. Claim(s) none is/are amended, claim(s) none is/are cancelled, claim(s) none is/are new, and applicant does not provide any information on where support for the amendments can be found in the instant specification. Therefore, Claims 1-20 is/are pending and have been addressed below.
Priority
Acknowledgment is made of applicant's claim for foreign priority based on an application(s) filed in China on 12/4/24. Should applicant desire to obtain the benefit of foreign priority under 35 U.S.C. 119(a)-(d) prior to declaration of an interference, a certified English translation of the foreign application must be submitted in reply to this action. 37 CFR 41.154(b) and 41.202(e).
Failure to provide a certified translation may result in no benefit being accorded for the non-English application.
Response to Arguments
There are no arguments.
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 is/are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claim(s) does/do not fall within at least one of the four categories of patent eligible subject matter as noted below.
The limitation(s) below for representative claim(s) 1, 8, and 15 that, under its broadest reasonable interpretation, is directed to job matching.
Step 1: The claim(s) as drafted, is/are a process (claim(s) 1-7 recites a series of steps) and system (claim(s) 8-20 recites a series of components).
Step 2A – Prong 1: The claimed invention is directed to an abstract idea without significantly more. The claim(s) recite(s) (emphasis added):
Claim 1: determining a service processing indicator score of each service work order of a first set of service work orders;
selecting, from the first set of service work orders, a second set of service work orders with service processing indicator scores that satisfy a preset value;
determining first data of each of the second set of service work orders based on service work order information in each of the second set of service work orders;
determining second data of each of the second set of service work orders based on a service processing result in each of the second set of service work orders;
generating a training set based on the first data and the second data;
training an initial information recommendation model based on the training set, to obtain an information recommendation model;
obtaining a processing sequence of each of a plurality of to-be-processed service work orders based on the information recommendation model; and
displaying the processing sequence of the plurality of to-be-processed service work orders.
Claim(s) 8 and 15: same analysis as claim(s) 1.
Dependent claims 2-7, 9-14, and 16-20 recite the same or similar abstract idea(s) as independent claim(s) 1, 8, and 15 with merely a further narrowing of the abstract idea(s): .
The identified limitations of the independent and dependent claims above fall well-within the groupings of subject matter identified by the courts as being abstract concepts of:
a method of organizing human activity (commercial or legal interactions including advertising, marketing or sales activities or behaviors, or business relations) because the invention is directed to economic and/or business relationships as they are associated with job matching.
Step 2A – Prong 2: This judicial exception is not integrated into a practical application because:
The additional elements unencompassed by the abstract idea include an information recommendation apparatus, processing circuitry (claim(s) 8), a non-transitory computer-readable storage medium, processor (claim(s) 15).
The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements as described above with respect to Step 2A Prong 2 fails to describe:
Improvements to the functioning of a computer, or to any other technology or technical field - see MPEP 2106.05(a)
Applying or using a judicial exception to effect a particular treatment or prophylaxis for a disease or medical condition – see Vanda Memo
Applying the judicial exception with, or by use of, a particular machine – see MPEP 2106.05(b)
Effecting a transformation or reduction of a particular article to a different state or thing - see MPEP 2106.05(c)
Applying or using the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is more than a drafting effort designed to monopolize the exception - see MPEP 2106.05(e) and Vanda Memo.
Thus the additional elements as described above with respect to Step 2A Prong 2 are merely (as additionally noted by instant specification [0158]) invoked as a tool and/or general purpose computer to apply instructions of an abstract idea in a particular technological environment, and/or mere application of an abstract idea in a particular technological environment and merely limiting the use of an abstract idea to a particular technological field do not integrate an abstract idea into a practical application (MPEP 2106.05(f)&(h)).
Step 2B: The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. Thus the additional elements as described above with respect to Step 2A Prong 2 are merely (as additionally noted by instant specification [0158]) invoked as a tool and/or a general purpose computer to apply instructions of an abstract idea in a particular technological environment, and/or mere application of an abstract idea in a particular technological environment and merely limiting the use of an abstract idea to a particular technological field do not integrate an abstract idea into a practical application and thus similarly the combination and arrangement of the above identified additional elements when analyzed under Step 2B also fails to necessitate a conclusion that the claims amount to significantly more than the abstract idea for the same reasons as set forth above (MPEP 2106.05(f)&(h)).
Claim Rejections - 35 USC § 102
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 public use, on sale or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claim(s) 1, 3-4, 7, 8, 10-11, 14, 15, and 17-18 is/are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Nath et al. (US 2015/0317582 A1).
Regarding claim 1, 8, and 15, Nath teaches an information recommendation method, the method comprising:
determining a service processing indicator score of each service work order of a first set of service work orders;
selecting, from the first set of service work orders, a second set of service work orders with service processing indicator scores that satisfy a preset value [for the limitations above, see at least [0052] a) initially determine probability of task(s) being completed by a worker(s) and b) selecting a worker(s) based on a higher probability (threshold) of a subset ( all (or most tasks) ) being completed “In other words, suppose a new task received from a task publisher, and that the Context-Aware Crowdsourced Task Optimizer knows how complicated the task is, what the payment for the task is, how far the task is from particular workers or the expected travel routes of those workers. Given such information, the Context-Aware Crowdsourced Task Optimizer will evaluate the learned worker models to return a probabilistic likelihood that particular workers will perform the particular task or bundle (optionally within some period of time). However, given a potentially very large worker pool, it is likely that multiple workers will be identified as having relatively high probabilities of completing particular tasks or task bundles. So, the question becomes which workers will receive the recommendation where they have the same or similar probability of completing the task or task bundle. As noted above, the recommendation question is answered by assigning or recommending tasks in a way that optimizes the task completion rates over all (or most tasks).”];
determining first data of each of the second set of service work orders based on service work order information in each of the second set of service work orders;
determining second data of each of the second set of service work orders based on a service processing result in each of the second set of service work orders [for the limitations above, see at least [0026] receive task data “In general, as illustrated by FIG. 1, the processes enabled by the Context-Aware Crowdsourced Task Optimizer begin operation by using a task input module 100 to receive one or more tasks 105 from human or virtual task publishers (110, 115, 120). In addition, the task input module 100 also receives one or more optional task contexts, e.g., prices, location, deadlines, number of instances, etc.”;
[0027] evaluate task data “In various embodiments, a task feedback module 125 compute completion rates for various contexts such as prices, deadlines, etc., and uses these completion rates to provide guidance to task publishers (110, 115, 120) for specifying various task contexts prior to publishing those tasks 105.”];
generating a training set based on the first data and the second data;
training an initial information recommendation model based on the training set, to obtain an information recommendation model [see at least [0016] machine learning trained on data “In other words, instead of allowing the workers to choose any task they want, the Context-Aware Crowdsourced Task Optimizer uses machine learning to train predictive models that are used to determine which workers are most likely to complete particular tasks (or bundles of two or more tasks) successfully, and then to assign or recommend tasks and task bundles to workers most likely to complete those tasks.”;
[0090] training data includes tasks and workers “The learned worker models are updated over time as more data is collected for each worker. Periodic, continuous, or real-time updates to these models over time ensures that the Context-Aware Crowdsourced Task Optimizer has the ability to provide accurate and up to date estimations of workers' predicted behaviors regarding recommended tasks or task bundles.”];
obtaining a processing sequence of each of a plurality of to-be-processed service work orders based on the information recommendation model; and
displaying the processing sequence of the plurality of to-be-processed service work orders [for the limitations above, see at least [0051] order flow is 4-6 “For example, referring back to FIG. 2, the idea of bundling tasks can be illustrated by assuming that worker ω2 (240) accepts a recommendation of task bundle TA presented by the Context-Aware Crowdsourced Task Optimizer. Assume that in this example, to do task τ4, worker ω2 (240) demands extra payment to cover the trip overhead (e.g., cost of gas, time, etc.) of detouring to go the task location. However, once the worker is at task τ4's location, he might as well do tasks τ5 and τ6 that are located nearby, without any significant trip overhead. Consequently, by bundling the tasks together and recommending bundle TA to the worker, the Context-Aware Crowdsourced Task Optimizer can increase the completion rate of all the three tasks in this bundle, possibly with a dynamically adjusted price that is smaller than the sum of all individual task payments.”;
[0032, 0095] priority based on bonus “Further, a bonus may be provided upon completion of an entire bundle, or completion of particular tasks or bundles prior to some deadline.” where jobs are separated into 2 item bundles thus a plurality (many bundles) are still given processing order “More specifically, FIG. 3 shows task bundle TC (350) paired or recommended to human worker ω3 (310). Single task τ4 (360) is shown as being paired or recommended to human worker ω4 (320). Task bundle TD (370) is shown as being paired or recommended to human worker ω6 (370). Finally, task bundle TE (380) is shown as being paired or recommended to virtual worker ω5 (330).”;
[0055] provide tasks to worker as recommended “Another example of actively notifying workers is that in various embodiments, the Context-Aware Crowdsourced Task Optimizer pushes recommendations of tasks to workers that are near tasks that can be accepted … For example, the Context-Aware Crowdsourced Task Optimizer can alert the worker or send a message to a computing device (e.g., a cell phone) of the worker, such as, for example, “Based on your typical daily commute (or current travel route), you will be passing near Restaurant X. If you stop in and take a picture of the menu we will pay you a $5 for completion of that task.” In other words, the Context-Aware Crowdsourced Task Optimizer tries to learn known present locations, travel routes, anticipated future locations, etc., for workers, and then to use this and other information to recommend bundles of one or more tasks to the workers based on the location of those tasks relative to the location of the worker.”].
Regarding claim 3, 10, and 17, Nath teaches the method according to claim 1, wherein the service work order information comprises:
execution subject information, execution object information, execution process information, and service basic information [Examiner notes applicant has not acted as his or her own lexicographer to specifically define or redefine any of execution subject information, execution object information, execution process information, or service basic information in the specification [0038, 0067] thus the broadest reasonable interpretation is being used,
then see at least [0026] “In addition, the task input module 100 also receives one or more optional task contexts, e.g., prices, location, deadlines, number of instances, etc.”;
[0133] “However, it should be clear that, as noted above, the Context-Aware Crowdsourced Task Optimizer can support an arbitrary number of contexts (i.e., specifications or requirements for particular tasks) and can support deadlines of any time, by simply adjusting the utility or probability Pτω (associated with assigning task τ to worker ω).”;
[0098] “In this example, the task bundling may take into account various worker contexts such as drone range, lifting capabilities, and task contexts such as delivery addresses, package sizes, package weights, etc.”;
[0021] “payment or rewards associated with tasks or task bundles, task completion rates, tasks prices or costs, task deadlines, etc.”;
[0044] “transportation modes (e.g., walking, bicycling, driving, train travel, air travel, etc.), travel routes, … For example, a traffic monitoring or reporting task can be recommended to workers walking or driving at, near, or towards a target location, while a speech data collection task can be recommended to workers at quiet locations, noisy locations, locations where particular groups of people are speaking or where particular languages are being spoken, etc., depending on any task requirements or constraints specified by the task publisher.”;
[0134] “For example, consider a traffic monitoring task that requests workers in a “driving” context (i.e., one of the task parameters, xi, is a requirement that the worker is driving a vehicle) and on a specific freeway.”;
[0070] “if those tasks are simpler and require very short trips”;
[0110] “Finding an accurate minimum traveling distance required for each task, given the set Tω, can be very challenging.”;
[0016] more task requirements “The machine learning techniques consider each workers history and behavior with respect to task types, task completion rates, contexts such as locations, travel direction, schedule, capabilities or skills of each worker (e.g., professional photographer, foreign language skills, etc.), capabilities of the workers mobile computing devices or tools (e.g., high resolution still or video cameras, laser rangefinders, microphones, etc.).”;
[0088] more task requirements “Given the above parameters, and any of a wide range of additional considerations or parameters (e.g., age, gender, fitness level, education, skills, worker's computing devices, tools, equipment, travel capabilities, quality reviews of worker or task result from task publisher, etc.)”;
[0004] more task requirements “multiple tasks compatible with workers' contexts (e.g., worker history, present or expected worker locations, travel paths, working hours, skill set, capabilities of worker's mobile computing devices, etc.)”].
Regarding claim 4, 11, and 18, Nath teaches the method according to claim 3, wherein the first data includes input data for training the initial information recommendation model, and the second data includes label data for training the initial information recommendation model [see at least [0026] receive task data “In general, as illustrated by FIG. 1, the processes enabled by the Context-Aware Crowdsourced Task Optimizer begin operation by using a task input module 100 to receive one or more tasks 105 from human or virtual task publishers (110, 115, 120). In addition, the task input module 100 also receives one or more optional task contexts, e.g., prices, location, deadlines, number of instances, etc.”;
[0027] evaluate task data “In various embodiments, a task feedback module 125 compute completion rates for various contexts such as prices, deadlines, etc., and uses these completion rates to provide guidance to task publishers (110, 115, 120) for specifying various task contexts prior to publishing those tasks 105.”].
Regarding claim 7 and 14, Nath teaches the method according to claim 1, wherein the method further comprises: for each of the plurality of to-be-processed service work orders,
performing first processing on the respective to-be-processed service work order based on the information recommendation model, to obtain first feature data;
performing second processing on the respective to-be-processed service work order based on the information recommendation model, to obtain second feature data;
processing the first feature data and the second feature data based on the information recommendation model, to obtain a probability that a service processing result for the respective to-be-processed service work order indicates success [for the limitations above, see at least [0052] a) initially determine probability of task(s) being completed by a worker(s) and b) selecting a worker(s) based on a higher probability (threshold) of a subset ( all (or most tasks) ) being completed “In other words, suppose a new task received from a task publisher, and that the Context-Aware Crowdsourced Task Optimizer knows how complicated the task is, what the payment for the task is, how far the task is from particular workers or the expected travel routes of those workers. Given such information, the Context-Aware Crowdsourced Task Optimizer will evaluate the learned worker models to return a probabilistic likelihood that particular workers will perform the particular task or bundle (optionally within some period of time). However, given a potentially very large worker pool, it is likely that multiple workers will be identified as having relatively high probabilities of completing particular tasks or task bundles. So, the question becomes which workers will receive the recommendation where they have the same or similar probability of completing the task or task bundle. As noted above, the recommendation question is answered by assigning or recommending tasks in a way that optimizes the task completion rates over all (or most tasks).”;
[0140] determining how sub values (feature values) affect overall score (probability which is task completion rate) and use best combination of sub values “This enables the task publisher to work with an interactive user interface component of the Context-Aware Crowdsourced Task Optimizer to adjust or modify one or more task contexts (e.g., task pricing, deadlines, etc.) to see what effect those adjustments will have on the predicted task completion rate. As such, the task publisher can specify one or more tasks, and then adjust or otherwise modify any associated contexts before the task is actually offered or recommended to workers.”]; and
determining the processing sequence of the plurality of to-be-processed service work orders based on the probability that each of the plurality of service processing results indicates success [see at least [0051] order flow is 4-6 “For example, referring back to FIG. 2, the idea of bundling tasks can be illustrated by assuming that worker ω2 (240) accepts a recommendation of task bundle TA presented by the Context-Aware Crowdsourced Task Optimizer. Assume that in this example, to do task τ4, worker ω2 (240) demands extra payment to cover the trip overhead (e.g., cost of gas, time, etc.) of detouring to go the task location. However, once the worker is at task τ4's location, he might as well do tasks τ5 and τ6 that are located nearby, without any significant trip overhead. Consequently, by bundling the tasks together and recommending bundle TA to the worker, the Context-Aware Crowdsourced Task Optimizer can increase the completion rate of all the three tasks in this bundle, possibly with a dynamically adjusted price that is smaller than the sum of all individual task payments.”;
[0032, 0095] priority based on bonus “Further, a bonus may be provided upon completion of an entire bundle, or completion of particular tasks or bundles prior to some deadline.” where jobs are separated into 2 item bundles thus a plurality (many bundles) are still given processing order “More specifically, FIG. 3 shows task bundle TC (350) paired or recommended to human worker ω3 (310). Single task τ4 (360) is shown as being paired or recommended to human worker ω4 (320). Task bundle TD (370) is shown as being paired or recommended to human worker ω6 (370). Finally, task bundle TE (380) is shown as being paired or recommended to virtual worker ω5 (330).”].
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries set forth in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966), that are applied 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.
It has been held that a prior art reference must either be in the field of applicant’s endeavor or, if not, then be reasonably pertinent to the particular problem with which the applicant was concerned, in order to be relied upon as a basis for rejection of the claimed invention. See In re Oetiker, 977 F.2d 1443, 24 USPQ2d 1443 (Fed. Cir. 1992).
Claim(s) 2, 9, and 16 is/are rejected under 35 U.S.C. 103 as being unpatentable over Nath et al. (US 2015/0317582 A1) in view of Stack Exchange published August 4, 2011 (reference U on the Notice of References Cited).
Regarding claim 2, 9, and 16, Nath teaches the method according to claim 1, wherein the determining the service processing indicator score further comprises:
(original) determining a first processing indicator score of each service work order of the first set of service work orders based on first processing conversion data, first communication data, and a first weight, the first processing conversion data and the first communication data corresponding to each of the first set of service work orders;
Nath teaches (original vs citation vs plain lettering, underlined, and bolded citation that doesn’t fit clearly in original format beginning with and additionally) determining a first processing indicator score of each service work order of the first set of service work orders based on first processing conversion data, first communication data, and a first weight, the first processing conversion data and the first communication data corresponding to each of the first set of service work orders and additionally first processing conversion data, first communication data, the first processing conversion data and the first communication data corresponding to each of the first set of service work orders;
(original) determining a second processing indicator score of each service work order of the first set of service work orders based on first violation data, first complaint data, and a second weight, the first violation data and the first complaint data corresponding to each of the first set of service work orders; and
Nath teaches (original vs citation vs plain lettering, underlined, and bolded citation that doesn’t fit clearly in original format beginning with and additionally) determining a second processing indicator score of each service work order of the first set of service work orders based on first violation data, first complaint data, and a second weight, the first violation data and the first complaint data corresponding to each of the first set of service work orders and additionally first violation data, first complaint data, the first violation data and the first complaint data corresponding to each of the first set of service work orders; and
determining the service processing indicator score of each service work order based on the first processing indicator score of the respective service work order and the second processing indicator score of the respective service work order [for the limitations above, see at least [0052] a) initially determine probability of task(s) being completed by a worker(s) and b) selecting a worker(s) based on a higher probability (threshold) of a subset ( all (or most tasks) ) being completed “In other words, suppose a new task received from a task publisher, and that the Context-Aware Crowdsourced Task Optimizer knows how complicated the task is, what the payment for the task is, how far the task is from particular workers or the expected travel routes of those workers. Given such information, the Context-Aware Crowdsourced Task Optimizer will evaluate the learned worker models to return a probabilistic likelihood that particular workers will perform the particular task or bundle (optionally within some period of time). However, given a potentially very large worker pool, it is likely that multiple workers will be identified as having relatively high probabilities of completing particular tasks or task bundles. So, the question becomes which workers will receive the recommendation where they have the same or similar probability of completing the task or task bundle. As noted above, the recommendation question is answered by assigning or recommending tasks in a way that optimizes the task completion rates over all (or most tasks).”;
[0140] determining how sub values (scores for individual contexts) affect overall score (probability which is task completion rate) “This enables the task publisher to work with an interactive user interface component of the Context-Aware Crowdsourced Task Optimizer to adjust or modify one or more task contexts (e.g., task pricing, deadlines, etc.) to see what effect those adjustments will have on the predicted task completion rate. As such, the task publisher can specify one or more tasks, and then adjust or otherwise modify any associated contexts before the task is actually offered or recommended to workers.”;
[0021] non limiting method determine sub values including weights “Advantageously the Context-Aware Crowdsourced Task Optimizer can use any of a large number of optimization algorithms or processes to solve this optimization problem, e.g., greedy algorithms, expectation-maximization algorithms, etc. … Constraints are generally either hard constraints which set conditions that are to be satisfied for particular variables, or soft constraints which are used to penalize or weight variables in the objective function when corresponding conditions on the variables are not satisfied. Some of the constraints considered by the Context-Aware Crowdsourced Task Optimizer include, but are not limited to, available workers, present and future contexts of those available workers, available tasks, payment or rewards associated with tasks or task bundles, task completion rates, tasks prices or costs, task deadlines, etc.”;
the following citations include various items that could be various first data
then see at least [0026] “In addition, the task input module 100 also receives one or more optional task contexts, e.g., prices, location, deadlines, number of instances, etc.”;
[0133] “However, it should be clear that, as noted above, the Context-Aware Crowdsourced Task Optimizer can support an arbitrary number of contexts (i.e., specifications or requirements for particular tasks) and can support deadlines of any time, by simply adjusting the utility or probability Pτω (associated with assigning task τ to worker ω).”;
[0098] “In this example, the task bundling may take into account various worker contexts such as drone range, lifting capabilities, and task contexts such as delivery addresses, package sizes, package weights, etc.”;
[0021] “payment or rewards associated with tasks or task bundles, task completion rates, tasks prices or costs, task deadlines, etc.”;
[0044] “transportation modes (e.g., walking, bicycling, driving, train travel, air travel, etc.), travel routes, … For example, a traffic monitoring or reporting task can be recommended to workers walking or driving at, near, or towards a target location, while a speech data collection task can be recommended to workers at quiet locations, noisy locations, locations where particular groups of people are speaking or where particular languages are being spoken, etc., depending on any task requirements or constraints specified by the task publisher.”;
[0134] “For example, consider a traffic monitoring task that requests workers in a “driving” context (i.e., one of the task parameters, xi, is a requirement that the worker is driving a vehicle) and on a specific freeway.”;
[0070] “if those tasks are simpler and require very short trips”;
[0110] “Finding an accurate minimum traveling distance required for each task, given the set Tω, can be very challenging.”;
[0016] more task requirements “The machine learning techniques consider each workers history and behavior with respect to task types, task completion rates, contexts such as locations, travel direction, schedule, capabilities or skills of each worker (e.g., professional photographer, foreign language skills, etc.), capabilities of the workers mobile computing devices or tools (e.g., high resolution still or video cameras, laser rangefinders, microphones, etc.).”;
[0088] more task requirements “Given the above parameters, and any of a wide range of additional considerations or parameters (e.g., age, gender, fitness level, education, skills, worker's computing devices, tools, equipment, travel capabilities, quality reviews of worker or task result from task publisher, etc.)”;
[0004] more task requirements “multiple tasks compatible with workers' contexts (e.g., worker history, present or expected worker locations, travel paths, working hours, skill set, capabilities of worker's mobile computing devices, etc.)”].
Nath doesn’t/don’t explicitly teach however Stack Exchange discloses
(original vs citation) determining a first processing indicator score of each service work order of the first set of service work orders based on first processing conversion data, first communication data, and a first weight, the first processing conversion data and the first communication data corresponding to each of the first set of service work orders;
(original vs citation) determining a second processing indicator score of each service work order of the first set of service work orders based on first violation data, first complaint data, and a second weight, the first violation data and the first complaint data corresponding to each of the first set of service work orders [see at least [pg 1-2] combining variables into sub scores and the sub scores into a final score].
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Nath with Stack Exchange to include the limitation(s) above as disclosed by Stack Exchange. Doing so would improve Nath’s (Nath) job matching via a more specific sub score methodology to provide a more in depth job match [see at least Stack Exchange [pg 1-2] ].
Furthermore, all of the claimed elements were known in the prior arts of a) Nath and b) Stack Exchange and c) one skilled in the art could have combined the elements as claimed by known methods with no change in their respective functions, and the combination would have yielded predictable results to one of ordinary skill in the art before the effective filing date of the claimed invention.
Claim(s) 5, 12, and 19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Nath et al. (US 2015/0317582 A1) in view of Jakkam Reddi et al. (US 2023/0113984 A1).
Regarding claim 5, 12, and 19, Nath teaches the method according to claim 1, wherein the training the initial information recommendation model further comprises:
inputting the first data into the initial information recommendation model, to obtain a service processing result indicating success of a service work order;
(original vs citation) determining a first loss function based on the service processing result in each of the second set of service work orders; and
(original vs citation) performing reverse iteration processing on the initial information recommendation model based on the first loss function until the initial information recommendation model converges, to obtain the information recommendation model [for the limitations above, see at least [0016] machine learning trained on data “In other words, instead of allowing the workers to choose any task they want, the Context-Aware Crowdsourced Task Optimizer uses machine learning to train predictive models that are used to determine which workers are most likely to complete particular tasks (or bundles of two or more tasks) successfully, and then to assign or recommend tasks and task bundles to workers most likely to complete those tasks.”;
[0090] training data includes tasks and workers “The learned worker models are updated over time as more data is collected for each worker. Periodic, continuous, or real-time updates to these models over time ensures that the Context-Aware Crowdsourced Task Optimizer has the ability to provide accurate and up to date estimations of workers' predicted behaviors regarding recommended tasks or task bundles.”].
Nate doesn’t/don’t explicitly teach however Jakkam Reddi discloses
(original vs citation vs clarification: strikethrough of italicized original followed by plain lettering clarification in bold) determining a first loss function based on the a difference between the estimated outputs and the one or more values; and
(original vs citation) performing reverse iteration processing on the model based on the first loss function until the model converges, to obtain the information recommendation model [see at least [0009] “One example aspect of the present disclosure is directed to a computer-implemented method for optimizing machine-learned models that provides improved convergence properties. The method includes determining, by one or more computing devices, a gradient of a loss function that evaluates a performance of a machine-learned model that includes a plurality of parameters. The method includes determining, by the one or more computing devices, a candidate learning rate control value based at least in part on the gradient of the loss function. The method includes comparing, by the one or more computing devices, the candidate learning rate control value to a maximum previously observed learning rate control value. The method includes, when the candidate learning rate control value is greater than the maximum previously observed learning rate control value: setting a current learning rate control value equal to the candidate learning rate control value; and setting the maximum previously observed learning rate control value equal to the candidate learning rate control value. The method includes, when the candidate learning rate control value is less than the maximum previously observed learning rate control value: setting the current learning rate control value equal to the maximum previously observed learning rate control value.”;
[0023] “For example, the learning rate control value can be an exponential moving average of squared past and current gradients of a loss function that evaluates performance of the machine-learned model on training data. The learning rate can be a function of and inversely correlated to the learning rate control value. To avoid situations where the learning rate increases iteration-over-iteration, the system can select, for use in determining the current learning rate for the current iteration, a maximum of a candidate learning rate control value determined for the current iteration and a maximum previously observed learning rate control value seen in past iterations. By selecting the maximum of the candidate learning rate control value and the maximum previously observed control value, the system can ensure that the current learning rate (which may be inversely correlated to the selected control value) does not increase during the iterative optimization. In such fashion, the optimization system can be endowed with “long-term memory” of past gradients. As a result, the optimization techniques described herein can provide the benefits of use of an adaptive learning rate, while avoiding certain scenarios in which existing adaptive optimization techniques fail to converge (e.g., scenarios which result in learning rates that are not monotonically non-increasing). The systems and methods of the present disclosure provide guaranteed convergence, while also reducing the number of hyperparameters, converging faster than certain existing techniques, and providing superior generalization capacity.”].
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Nate with Jakkam Reddi to include the limitation(s) above as disclosed by Tan. Doing so would improve Nate’s (Nate) machine learning fine tuning to explicitly use a loss function instead of the implied use of loss function [see at least Jakkam Reddi [0004] ].
Furthermore, all of the claimed elements were known in the prior arts of a) Nate and b) Jakkam Reddi and c) one skilled in the art could have combined the elements as claimed by known methods with no change in their respective functions, and the combination would have yielded predictable results to one of ordinary skill in the art before the effective filing date of the claimed invention.
Claim(s) 6, 13, and 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Nath in view of Jakkam Reddi as applied to claim(s) 5, 12, and 19 above and further in view of Ha et al. published February 25, 2019 (reference V on the Notice of References Cited).
Regarding claim 6, 13, and 20, modified Nath teaches the method according to claim 5,
and Nath teaches wherein the method further comprises:
determining first timing information when the service processing result indicates success;
determining second timing information when the service processing result indicates unsuccess;
(original vs citation) determining a second loss function based on the first timing information and the second timing information; and
(original vs citation) performing reverse iteration processing on the initial information recommendation model by using the first loss function and the second loss function until the initial information recommendation model converges, to obtain the information recommendation model [for the limitations above, see at least [0052] a) initially determine probability of task(s) being completed by a worker(s) and b) selecting a worker(s) based on a higher probability (threshold) of a subset ( all (or most tasks) ) being completed “In other words, suppose a new task received from a task publisher, and that the Context-Aware Crowdsourced Task Optimizer knows how complicated the task is, what the payment for the task is, how far the task is from particular workers or the expected travel routes of those workers. Given such information, the Context-Aware Crowdsourced Task Optimizer will evaluate the learned worker models to return a probabilistic likelihood that particular workers will perform the particular task or bundle (optionally within some period of time). However, given a potentially very large worker pool, it is likely that multiple workers will be identified as having relatively high probabilities of completing particular tasks or task bundles. So, the question becomes which workers will receive the recommendation where they have the same or similar probability of completing the task or task bundle. As noted above, the recommendation question is answered by assigning or recommending tasks in a way that optimizes the task completion rates over all (or most tasks).”;
[0140] determining how sub values (scores for individual contexts) affect overall score (probability which is task completion rate) “This enables the task publisher to work with an interactive user interface component of the Context-Aware Crowdsourced Task Optimizer to adjust or modify one or more task contexts (e.g., task pricing, deadlines, etc.) to see what effect those adjustments will have on the predicted task completion rate. As such, the task publisher can specify one or more tasks, and then adjust or otherwise modify any associated contexts before the task is actually offered or recommended to workers.”;
[0060] “A task is considered to be successfully completed if it is performed by at least one worker within the deadline of the task.”].
Modified Nate doesn’t/don’t explicitly teach however Ha discloses
(original vs citation vs clarification: strikethrough of italicized original followed by plain lettering clarification in bold) determining a second loss function based on input; and
(original vs citation) performing reverse iteration processing on the initial information recommendation model by using the first loss function and the second loss function until the initial information recommendation model converges, to obtain the information recommendation model [see at least [pg 1] “analyze the performance of alternating minimization for loss functions optimized over two variables … Our results further reveal important distinctions between alternating and non-alternating methods. Since computing the alternating minimization steps may not be tractable for some problems, we also consider an inexact version of the algorithm and provide a set of sufficient conditions to ensure fast convergence of the inexact algorithms.”;
[pg 1-2] “Our aim is to study the convergence behavior of two methods for this problem. First, we consider the alternating minimization method, where we iterate the steps
This type of method can be practical in scenarios where the loss function is relatively simple to minimize when viewed as a function of either x or y only—for instance, in multivariate regression, where x represents the coefficients and y represents the covariance structure. In other settings, even the marginal minimization steps are expensive to calculate, but we can instead consider approximating each one with gradient descent. The resulting alternating gradient descent algorithm iterates the following steps:” and equations].
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify modified Nate with Ha to include the limitation(s) above as disclosed by Ha. Doing so would improve modified Nate’s (Nate) machine learning fine tuning to explicitly use a loss function instead of the implied use of loss function [see at least Ha [pg 1-2] ].
Furthermore, all of the claimed elements were known in the prior arts of a) modified Nate and b) Ha and c) one skilled in the art could have combined the elements as claimed by known methods with no change in their respective functions, and the combination would have yielded predictable results to one of ordinary skill in the art before the effective filing date of the claimed invention.
Conclusion
When responding to the office action, any new claims and/or limitations should be accompanied by a reference as to where the new claims and/or limitations are supported in the original disclosure.
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Qian et al. – CN 111242752 A (relevant because it teaches task assignment)
Amiri – The bundled task assignment problem in mobile crowdsensing: a lagrangean relaxation-based solution approach (relevant because it teaches task assignment)
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JAMES WEBB whose telephone number is (313)446-6615. The examiner can normally be reached on M-F 10-3.
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, Jerry O’Connor can be reached on (571) 272-6787. 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.
/JAMES WEBB/Examiner, Art Unit 3624
Prosecution Insights