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 .
DETAILED ACTION
This office action is in response to the claims received on 4/4/2024.
Claim Interpretation
Plain Meaning (MPEP 2111.01): MPEP 2111.01 states: The plain meaning of a term means the ordinary and customary meaning given to the term by those of ordinary skill in the art at the time of the invention. The ordinary and customary meaning of a term may be evidenced by a variety of sources, including the words of the claims themselves, the specification, drawings, and prior art. However, the best source for determining the meaning of a claim term is the specification. An applicant is entitled to be their own lexicographer and may rebut the presumption that claim terms are to be given their ordinary and customary meaning by clearly setting forth a definition of the term that is different from its ordinary and customary meaning(s) in the specification at the relevant time. See In re Paulsen, 30 F.3d 1475, 1480, 31 USPQ2d 1671, 1674 (Fed. Cir. 1994). In this case:
"Non-transitory machine-readable medium": Regarding claims 15-20, the specification does mention “non-transitory” in par. 83 without redefining this term. Therefore, it has the original meaning. The specification provides a definition for claimed “machine-readable medium” in par. 84, which is similar to the original meaning; and therefore, claimed “non-transitory machine-readable medium” clearly excludes all transitory embodiments.
Subject Matter Eligible under 35 USC 101
Please refer to the Subject Matter Eligibility Test for Products and Processes in MPEP 2106 and in the 2019 Revised Patent Subject Matter Eligibility Guidance:
Step 1: See MPEP 2106.03: 35 U.S.C. 101 enumerates four categories of subject matter that Congress deemed to be appropriate subject matter for a patent: processes, machines, manufactures and compositions of matter. Claims 1-7, 15-20 include claims directed to a machine, claims 8-14 are directed to a process, which are statutory categories. Therefore, the answer in step 1 is YES.
Step 2A prong 1: Yes, the independent claims recite an abstract idea of a concept performed in the human mind including an evaluation in “receiving, from the trained machine learning model, a predicted measure of user interaction for a set of words in the user communication”; therefore, the answer is YES.
Step 2A prong 2: Yes, the claim does recite additional elements that integrate the exception into a practical application of the exception. Par. 62 of the specification explains that once a predicted measure of user interaction is obtained, as the claims require, an email communication server will send messages only when they are likely to be engaged with, i.e., opened by a recipient, reducing the number of low interaction messages sent, and improving the efficiency of a communication server by reducing the amount of network bandwidth and processor cycles used to process messages that are erased upon receipt. Accordingly, use of the claimed method improves the operation of the communication server. Therefore, the claimed invention provides an improvement to the state-of-the-art advertisement servers, which is therefore a specific improvement over prior systems, and claimed “predicted measure of user interaction” reflects such practical application. Please refer to MPEP 2106.04(d): "Integration of a Judicial Exception Into A Practical Application" under the header "Relevant considerations for evaluating whether additional elements integrate a judicial exception into a practical application": "Limitations the courts have found indicative that an additional element (or combination of elements) may have integrated the exception into a practical application include: • An improvement in the functioning of a computer, or an improvement to other technology or technical field, as discussed in MPEP §§ 2106.04(d)(1) and 2106.05(a)". See for example the court decision in 118 USPQ2d 1684 Enfish, LLC v. Microsoft Corp; U.S. Court of Appeals Federal Circuit, page 1689: “much of the advancement made in computer technology consists of improvements to software that, by their very nature, may not be defined by particular physical features but rather by logical structures and processes”. Therefore, the answer to prong 2 is YES, and the claims are eligible in step 2A.
Citation of Relevant Prior Art
The prior art made of record and not relied upon is considered pertinent to applicant’s disclosure: Mathur et al (publication number 2014/0149223), hereinafter Mathur, teaches (please refer to Mathur Figs. 1B and 5) an advertisement server which embeds an ad 150 within content displayed on a user device’s display 105. If the user clicks on the ad, the user performs step 505 of process 500, where the user requests an advertisement from the ad server; this step addresses the first limitation of claims 1 and 17, because the user request is an input to the ad server. In step 520, the ad server calculates an expected value for candidate advertisements; paragraph 45 of Mathur explains how the ad server calculates the expected value of revenue from displaying the ad for the user; step 520 addresses the second limitation of claims 1 and 17. Regarding the third limitation, step 555 displays the advertisement on a user interface; however, the claims require presenting a predicted measure of user interaction, and Mathur falls short of this requirement.
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 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 of this title, 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.
Joint Inventors, Common Ownership Presumed
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 at the time any inventions covered therein were effectively filed 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 at the time a later invention was effectively filed 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.
Test for Obviousness
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 or pre-AIA 35 U.S.C. 103(a) 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.
Claims 1-5, 8, 11-20 are rejected under 35 U.S.C. 103 as being unpatentable over Qin et al (publication number 2013/0211905), hereinafter Qin, and further in view of Martinez et al (publication number 2009/0177644), hereinafter Martinez, and further in view of Behtash et al (publication number 2021/0109958), hereinafter Behtash.
Regarding claim 1, Qin teaches a system (please refer to system represented in Qin par. 17 in reference to Fig. 1) comprising:
one or more processors (Qin Fig. 2 and [0024] processors 202); and
a memory that stores instructions that, when executed by the one or more processors, cause the one or more processors to perform operations (Qin Fig. 2 and [0024] processors 202, memory 204, computer readable instructions, [0065] operations) comprising:
providing a user communication to a trained machine learning model as input (Claimed "user communication" equates to Qin's online advertisement 106 in Fig. 1 and [0018] The user click inference engine 102 performs an analysis of the online advertisement 106 to generate a user click probability 112 for the online advertisement 106. The arrow from 106 to 102 means that 106 is an input for 102. Claimed "trained machine learning model" equates to Qin's inference engine 102 with its models 122-128 in Fig. 1 and [0055], [0056]: A training module 212 is used to learn the parameters and facilitate the training of the click behavior model 122. [0023] The user click inference engine 102 is implemented by the computer 104);
receiving, from the trained machine learning model, a predicted measure of user interaction (Qin [0003], [0012], [0081]: predicting a user click probability by taking into account both the relevance of the online advertisement to a user search query. [0022] The user click inference engine 102 generates the user click probability 112 for the online advertisement 106 using a click behavior model 122) for a set of words in the user communication (Qin [0013] The attractiveness of an online advertisement is contingent upon the ability the words in the online advertisement to attract the attention of a user. [0014] The advertisement attractiveness model is developed from a word-level attractiveness model that measures the attractiveness of individual words in the online advertisement); and
causing presentation of a user interface comprising a ranking (Qin [0062] [0063] The user may select online advertisements to be analyzed by the user click inference engine 102 via the user interface module 218. [0021] The user click inference engine 102 extracts a set of relevance features 120 that are visible to users, such as topical page rank).
Qin does not explicitly teach "attention levels for at least a subset of the set of words".
Martinez uses the following terms and acronyms:
[0025] A real-world entity (RWE) refers to a person, device, location, or other physical thing known to the W4 COMN. Each RWE known to the W4 COMN is assigned or otherwise provided with a unique W4 identification number that absolutely identifies the RWE within the W4 COMN. [0027] Examples of RWEs include physical entities, such as people, locations (e.g., states, cities, houses, buildings, airports, roads, etc.) and things (e.g., animals, pets, livestock, gardens, physical objects, cars, airplanes, works of art, etc.)
[0188], [0215]: A physical entity is described by a text, for example, "Bonzo's Pizzeria". [0222] Physical entities are pre-associated with text that describes the physical entity.
[0089] Everything in the world can be assigned an attention rank by combining and weighting the online data about that RWE. Attention is recorded online by browsers and devices and carriers and Network operators, transactions and instrumented pages or networks. [0100] World rank module 708 models real-world users' attention through devices.
Martinez teaches (please refer to Martinez Figs. 10, 11): causing presentation of a user interface comprising a ranking of attention levels (Martinez [0170], [0190] - [0192] FIG. 10 illustrates a W4 engine for displaying one or more physical entities and indicating the attention rank for each physical entity. [0216] FIG. 11 illustrates a method for displaying one or more physical entities and indicating the attention rank for each physical entity) for at least a subset of the set of words (Martinez [0188], [0215]: A physical entity is described by a text, for example, "Bonzo's Pizzeria". [0222] Physical entities are preassociated with text that describes the physical entity); and a machine learning model (Martinez [0161] machine learning techniques; Hidden Markov Models (HMMs), Support Vector Machines (SVMs); learning algorithms populated with data models).
Therefore, it would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to modify the disclosure of Qin, by incorporating the teachings of Martinez into the disclosure of Qin, because a great deal of information is generated when people use electronic devices, such as when people use mobile phones and cable set-top boxes. Such information, such as location, applications used, social network, physical and online locations visited, could be used to deliver useful services and information to end users, and provide commercial opportunities to advertisers and retailers. However, most of this information is effectively abandoned due to deficiencies. There exists a need for methods to collect and communicate data associated with users and their electronic devices (Martinez [0001]).
Qin as modified does not explicitly teach "trained".
Behtash teaches a trained machine learning model (Behtash [0252] So far, the machine learning models have been trained with supervised learning. [0311] [0336] Machine learning model trained).
Therefore, it would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to modify the disclosure of Qin as modified, by incorporating the teachings of Behtash into the disclosure of Qin as modified, because what is needed is a more efficient research tool that only needs a small amount of resources, consumes less electricity per query, and has a smaller carbon footprint compared to existing research tools such as those discussed above (Behtash [0018]).
Regarding claim 2, Qin teaches wherein the operations further comprise: accessing a training set comprising a plurality of user communications (Claimed "user communication" equates to Qin's online advertisement 106 in Fig. 1 and [0018] The user click inference engine 102 performs an analysis of the online advertisement 106 to generate a user click probability 112 for the online advertisement 106. Qin meets claimed "set" because Qin's disclosure comprises a plurality of online ads 106), each user communication of the plurality of user communications annotated with a measure of user interaction of the user communication (Qin Fig. 2 and [0020], [0021], [0028] As shown in the factor graph 222, score s depends on the relevance score r of an advertisement to the query and the attractiveness score a of the online advertisement. Therefore, each ad corresponds to a set of scores. [0064] The data store 220 stores advertisement attractiveness scores, relevance scores, and probability of clicks for online advertisements); and training (Qin Fig. 1 and [0055], [0056]: A training module 212 is used to learn the parameters and facilitate the training of the click behavior model 122. [0023] The user click inference engine 102 is implemented by the computer 104), based on the accessed training set, the machine learning model to predict the measure of user interaction for input user communications (Qin [0003], [0012], [0081]: predicting a user click probability by taking into account both the relevance of the online advertisement to a user search query. [0022] The user click inference engine 102 generates the user click probability 112 for the online advertisement 106 using a click behavior model 122).
Regarding claims 3, 12, 19, Qin teaches wherein: the plurality of user communications comprises a plurality of messages (Claimed "user communication" equates to Qin's online advertisement 106 in Fig. 1 and [0013] The attractiveness of an online advertisement is contingent upon the ability the words in the online advertisement to attract the attention of a user).
Regarding claims 4, 13, 16, 20, Qin as modified teaches wherein the ranking of the attention levels (Martinez [0170], [0190] - [0192] FIG. 10 illustrates a W4 engine for displaying one or more physical entities and indicating the attention rank for each physical entity. [0216] FIG. 11 illustrates a method for displaying one or more physical entities and indicating the attention rank for each physical entity).
Qin as modified does not explicitly teach "based on internal attention states of the trained machine learning model".
Behtash teaches wherein the ranking of the attention levels is based on internal attention states of the trained machine learning model (Behtash [0243] The attention mechanism 2703 produces a coefficient (weight) for each internal state—also called attention coefficient—which measures the amount of attention to be given to them. [0245] Process 2704 multiplies the attention coefficients αi by hi that means every internal state is weighted by the amount of attention received).
Therefore, it would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to modify the disclosure of Qin as modified, by incorporating the teachings of Behtash into the disclosure of Qin as modified, because what is needed is a more efficient research tool that only needs a small amount of resources, consumes less electricity per query, and has a smaller carbon footprint compared to existing research tools such as those discussed above (Behtash [0018]).
Regarding claims 5, 14, 18, Qin does not explicitly teach wherein the plurality of user communications comprises a plurality of short messaging system (SMS) messages.
Martinez teaches wherein the plurality of user communications comprises a plurality of short messaging system (SMS) messages (Martinez [0057], [0077] With respect to the interaction data, communications between any RWEs may be, for example, any data associated with an incoming or outgoing short message service (SMS) message. See also Table 2 after [0082]).
Therefore, it would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to modify the disclosure of Qin, by incorporating the teachings of Martinez into the disclosure of Qin, because a great deal of information is generated when people use electronic devices, such as when people use mobile phones and cable set-top boxes. Such information, such as location, applications used, social network, physical and online locations visited, could be used to deliver useful services and information to end users, and provide commercial opportunities to advertisers and retailers. However, most of this information is effectively abandoned due to deficiencies. There exists a need for methods to collect and communicate data associated with users and their electronic devices (Martinez [0001]).
Regarding claim 8, Qin teaches a method (please refer to system represented in Qin par. 17 in reference to Fig. 1; [0065] FIGS. 3-5 describe various example processes) comprising:
providing, by one or more processors (Qin Fig. 2 and [0024] processors 202), a user communication to a trained machine learning model as input (Claimed "user communication" equates to Qin's online advertisement 106 in Fig. 1 and [0018] The user click inference engine 102 performs an analysis of the online advertisement 106 to generate a user click probability 112 for the online advertisement 106. The arrow from 106 to 102 means that 106 is an input for 102. Claimed "trained machine learning model" equates to Qin's inference engine 102 with its models 122-128 in Fig. 1 and [0055], [0056]: A training module 212 is used to learn the parameters and facilitate the training of the click behavior model 122. [0023] The user click inference engine 102 is implemented by the computer 104);
receiving, from the trained machine learning model, a predicted measure of user interaction (Qin [0003], [0012], [0081]: predicting a user click probability by taking into account both the relevance of the online advertisement to a user search query. [0022] The user click inference engine 102 generates the user click probability 112 for the online advertisement 106 using a click behavior model 122) for a set of words in the user communication (Qin [0013] The attractiveness of an online advertisement is contingent upon the ability the words in the online advertisement to attract the attention of a user. [0014] The advertisement attractiveness model is developed from a word-level attractiveness model that measures the attractiveness of individual words in the online advertisement); and
causing presentation of a user interface comprising a ranking (Qin [0062] [0063] The user may select online advertisements to be analyzed by the user click inference engine 102 via the user interface module 218. [0021] The user click inference engine 102 extracts a set of relevance features 120 that are visible to users, such as topical page rank).
Qin does not explicitly teach "attention levels for at least a subset of the set of words".
Martinez teaches (please refer to Martinez Figs. 10, 11): causing presentation of a user interface comprising a ranking of attention levels (Martinez [0170], [0190] - [0192] FIG. 10 illustrates a W4 engine for displaying one or more physical entities and indicating the attention rank for each physical entity. [0216] FIG. 11 illustrates a method for displaying one or more physical entities and indicating the attention rank for each physical entity) for at least a subset of the set of words (Martinez [0188], [0215]: A physical entity is described by a text, for example, "Bonzo's Pizzeria". [0222] Physical entities are preassociated with text that describes the physical entity); and a machine learning model (Martinez [0161] machine learning techniques; Hidden Markov Models (HMMs), Support Vector Machines (SVMs); learning algorithms populated with data models).
Therefore, it would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to modify the disclosure of Qin, by incorporating the teachings of Martinez into the disclosure of Qin, because a great deal of information is generated when people use electronic devices, such as when people use mobile phones and cable set-top boxes. Such information, such as location, applications used, social network, physical and online locations visited, could be used to deliver useful services and information to end users, and provide commercial opportunities to advertisers and retailers. However, most of this information is effectively abandoned due to deficiencies. There exists a need for methods to collect and communicate data associated with users and their electronic devices (Martinez [0001]).
Qin as modified does not explicitly teach "trained".
Behtash teaches a trained machine learning model (Behtash [0252] So far, the machine learning models have been trained with supervised learning. [0311] [0336] machine learning model trained).
Therefore, it would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to modify the disclosure of Qin as modified, by incorporating the teachings of Behtash into the disclosure of Qin as modified, because what is needed is a more efficient research tool that only needs a small amount of resources, consumes less electricity per query, and has a smaller carbon footprint compared to existing research tools such as those discussed above (Behtash [0018]).
Regarding claim 11, Qin teaches accessing, from a database, a training set comprising a plurality of user communications (Claimed "user communication" equates to Qin's online advertisement 106 in Fig. 1 and [0018] The user click inference engine 102 performs an analysis of the online advertisement 106 to generate a user click probability 112 for the online advertisement 106. Qin meets claimed "set" because Qin's disclosure comprises a plurality of online ads 106), each user communication of the plurality of user communications annotated with a measure of user interaction of the user communication (Qin Fig. 2 and [0020], [0021], [0028] As shown in the factor graph 222, score s depends on the relevance score r of an advertisement to the query and the attractiveness score a of the online advertisement. Therefore, each ad corresponds to a set of scores. [0064] The data store 220 stores advertisement attractiveness scores, relevance scores, and probability of clicks for online advertisements); and training (Qin Fig. 1 and [0055], [0056]: A training module 212 is used to learn the parameters and facilitate the training of the click behavior model 122. [0023] The user click inference engine 102 is implemented by the computer 104), by the one or more processors and based on the accessed training set, the machine learning model to predict the measure of user interaction for input user communications (Qin [0003], [0012], [0081]: predicting a user click probability by taking into account both the relevance of the online advertisement to a user search query. [0022] The user click inference engine 102 generates the user click probability 112 for the online advertisement 106 using a click behavior model 122).
Regarding claim 15, Qin teaches a non-transitory machine-readable medium that stores instructions which, when executed by one or more processors, cause the one or more processors to perform operations (please refer to system represented in Qin par. 17 in reference to Fig. 1. Qin Fig. 2 and [0024] processors 202, memory 204, computer readable instructions, [0065] operations) comprising:
providing a user communication to a trained machine learning model as input (Claimed "user communication" equates to Qin's online advertisement 106 in Fig. 1 and [0018] The user click inference engine 102 performs an analysis of the online advertisement 106 to generate a user click probability 112 for the online advertisement 106. The arrow from 106 to 102 means that 106 is an input for 102. Claimed "trained machine learning model" equates to Qin's inference engine 102 with its models 122-128 in Fig. 1 and [0055], [0056]: A training module 212 is used to learn the parameters and facilitate the training of the click behavior model 122. [0023] The user click inference engine 102 is implemented by the computer 104);
receiving, from the trained machine learning model, a predicted measure of user interaction (Qin [0003], [0012], [0081]: predicting a user click probability by taking into account both the relevance of the online advertisement to a user search query. [0022] The user click inference engine 102 generates the user click probability 112 for the online advertisement 106 using a click behavior model 122) for a set of words in the user communication (Qin [0013] The attractiveness of an online advertisement is contingent upon the ability the words in the online advertisement to attract the attention of a user. [0014] The advertisement attractiveness model is developed from a word-level attractiveness model that measures the attractiveness of individual words in the online advertisement); and
causing presentation of a user interface comprising a ranking (Qin [0062] [0063] The user may select online advertisements to be analyzed by the user click inference engine 102 via the user interface module 218. [0021] The user click inference engine 102 extracts a set of relevance features 120 that are visible to users, such as topical page rank).
Qin does not explicitly teach "attention levels for at least a subset of the set of words".
Martinez teaches (please refer to Martinez Figs. 10, 11): causing presentation of a user interface comprising a ranking of attention levels (Martinez [0170], [0190] - [0192] FIG. 10 illustrates a W4 engine for displaying one or more physical entities and indicating the attention rank for each physical entity. [0216] FIG. 11 illustrates a method for displaying one or more physical entities and indicating the attention rank for each physical entity) for at least a subset of the set of words (Martinez [0188], [0215]: A physical entity is described by a text, for example, "Bonzo's Pizzeria". [0222] Physical entities are preassociated with text that describes the physical entity); and a machine learning model (Martinez [0161] machine learning techniques; Hidden Markov Models (HMMs), Support Vector Machines (SVMs); learning algorithms populated with data models).
Therefore, it would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to modify the disclosure of Qin, by incorporating the teachings of Martinez into the disclosure of Qin, because a great deal of information is generated when people use electronic devices, such as when people use mobile phones and cable set-top boxes. Such information, such as location, applications used, social network, physical and online locations visited, could be used to deliver useful services and information to end users, and provide commercial opportunities to advertisers and retailers. However, most of this information is effectively abandoned due to deficiencies. There exists a need for methods to collect and communicate data associated with users and their electronic devices (Martinez [0001]).
Qin as modified does not explicitly teach "trained".
Behtash teaches a trained machine learning model (Behtash [0252] So far, the machine learning models have been trained with supervised learning. [0311] [0336] machine learning model trained).
Therefore, it would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to modify the disclosure of Qin as modified, by incorporating the teachings of Behtash into the disclosure of Qin as modified, because what is needed is a more efficient research tool that only needs a small amount of resources, consumes less electricity per query, and has a smaller carbon footprint compared to existing research tools such as those discussed above (Behtash [0018]).
Regarding claim 17, Qin teaches accessing a training set comprising a plurality of user communications (Claimed "user communication" equates to Qin's online advertisement 106 in Fig. 1 and [0018] The user click inference engine 102 performs an analysis of the online advertisement 106 to generate a user click probability 112 for the online advertisement 106. Qin meets claimed "set" because Qin's disclosure comprises a plurality of online ads 106), each user communication of the plurality of user communications annotated with a measure of user interaction of the user communication (Qin Fig. 2 and [0020], [0021], [0028] As shown in the factor graph 222, score s depends on the relevance score r of an advertisement to the query and the attractiveness score a of the online advertisement. Therefore, each ad corresponds to a set of scores. [0064] The data store 220 stores advertisement attractiveness scores, relevance scores, and probability of clicks for online advertisements); and training (Qin Fig. 1 and [0055], [0056]: A training module 212 is used to learn the parameters and facilitate the training of the click behavior model 122. [0023] The user click inference engine 102 is implemented by the computer 104), based on the accessed training set, the machine learning model to predict the measure of user interaction for input user communications (Qin [0003], [0012], [0081]: predicting a user click probability by taking into account both the relevance of the online advertisement to a user search query. [0022] The user click inference engine 102 generates the user click probability 112 for the online advertisement 106 using a click behavior model 122).
Claims 6-7, 9-10 are rejected under 35 U.S.C. 103 as being unpatentable over Qin, in view of Martinez, in view of Behtash, and further in view of Cheng et al (publication number 2020/0265466), hereinafter Cheng.
Regarding claims 6, 9, Qin as modified does not explicitly teach "wherein the machine learning model comprises a multi-headed attention layer".
Cheng teaches wherein the machine learning model comprises a multi-headed attention layer (Cheng [0034] A hierarchical attention layer includes two parts as follows: (i) a transformer 320 with multi-head self-attention; and (ii) a series of hierarchical attention layers 330, 340, 345, 350. The second-order of attention is multi-head self-attention).
Therefore, it would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to modify the disclosure of Qin as modified, by incorporating the teachings of Cheng into the disclosure of Qin as modified, in order to provide reliable explanations together with accurate, trustworthy recommendations by quantifying the effects of feature combinations of arbitrary orders by a hierarchical attention mechanism and explains the recommending decision according to learned feature salience (Cheng [0003], [0013]).
Regarding claims 7, 10, Qin as modified does not explicitly teach "wherein the machine learning model comprises a self-attention layer".
Cheng teaches wherein the machine learning model comprises a self-attention layer (Cheng [0034] A hierarchical attention layer includes two parts as follows: (i) a transformer 320 with multi-head self-attention; and (ii) a series of hierarchical attention layers 330, 340, 345, 350. The second-order of attention is multi-head self-attention).
Therefore, it would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to modify the disclosure of Qin as modified, by incorporating the teachings of Cheng into the disclosure of Qin as modified, in order to provide reliable explanations together with accurate, trustworthy recommendations by quantifying the effects of feature combinations of arbitrary orders by a hierarchical attention mechanism and explains the recommending decision according to learned feature salience (Cheng [0003], [0013]).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to RONALD EISNER whose telephone number is (571)270-3334. The examiner can normally be reached on Monday and Tuesday from 9:00 AM to 5:30 PM.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Kathy Wang-Hurst, can be reached at telephone number (571) 270-5371. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from Patent Center. Status information for published applications may be obtained from Patent Center. Status information for unpublished applications is available through Patent Center for authorized users only. Should you have questions about access to Patent Center, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
Examiner interviews are available via a variety of formats see MPEP § 713.01. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) Form at https://www.uspto.gov/InterviewPractice.
/RONALD EISNER/
Primary Examiner, Art Unit 2644
Prosecution Insights