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 .
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 1-20 are rejected under 35 U.S.C. 101 for containing an abstract idea without significantly more.
Regarding Claim 1:
Step 1 - Is the claim directed to a process, a machine, manufacture or composition of matter? - Yes, the claim is directed to a process.
Step 2A - Prong 1 - Does the claim recite an abstract idea, law of nature, or natural phenomenon? - Yes, the claim recites the abstract ideas:
validating, by the one or more processors and using a second Al model, the classification code for the data object that is assigned to the data object via the first Al model, wherein the second Al model is trained using positive data objects that include assigned classification codes that are correct and negative data objects that include assigned classification codes that are incorrect: - This limitation is directed to the abstract idea of a mental process, as the process of validating the classification code for the data object that is assigned to the data object is a thought process that can be performed in a human mind by observing, evaluating and judging (concepts performed in the human mind (including an observation, evaluation, judgment, opinion) (see MPEP § 2106.04(a)(2), subsection III)). The use of a second AI model is discussed next at step 2 – prong 2.
Step 2A - Prong 2 - Does the claim recite additional elements that integrate the judicial exception into a practical application? - No, there are no additional elements that integrate the judicial exception into a practical application. The additional elements:
receiving, by one or more processors, a data object associated with an entity including a classification code that is assigned to the data object via a first artificial intelligence (Al) model: - This limitation is directed to mere data gathering and outputting. The courts (as per Ultramercial, 772 F.3d at 715, 112 USPQ2d at 1754) have recognized mere data gathering and outputting as insignificant extra-solution activity (see MPEP 2106.05(g)(3)) and therefore fails to integrate the exception into a practical application;
… using a second Al model…: - This limitation does not integrate a judicial exception into a practical application as the AI model is recited at a high level of generality, therefore this amounts to mere instructions to implement an abstract idea 2106.05(f);
and transmitting, by the one or more processors and to a user device, a validation result based on validating the classification code for the data object using the second Al model: - This limitation is directed to mere data gathering and outputting. The courts (as per Ultramercial, 772 F.3d at 715, 112 USPQ2d at 1754) have recognized mere data gathering and outputting as insignificant extra-solution activity (see MPEP 2106.05(g)(3)) and therefore fails to integrate the exception into a practical application.
Step 2B - Does the claim recite additional elements that amount to significantly more than the judicial exception? - No, there are no additional elements that amount to significantly more than the judicial exception. The additional elements:
receiving, by one or more processors, a data object associated with an entity including a classification code that is assigned to the data object via a first artificial intelligence (Al) model: - This limitation is directed to receiving or transmitting data over a network. The courts (as per Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362) have recognized receiving or transmitting data over a network as well-understood, routine, and conventional functions when they are claimed in a merely generic manner (e.g., at a high level of generality) (see MPEP 2106.05(d) II);
… using a second Al model…: - This limitation does not amount to significantly more than the judicial exception as the AI model is recited at a high level of generality, therefore this amounts to mere instructions to implement an abstract idea 2106.05(f);
and transmitting, by the one or more processors and to a user device, a validation result based on validating the classification code for the data object using the second Al model: - This limitation is directed to receiving or transmitting data over a network. The courts (as per Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362) have recognized receiving or transmitting data over a network as well-understood, routine, and conventional functions when they are claimed in a merely generic manner (e.g., at a high level of generality) (see MPEP 2106.05(d) II).
Regarding Claim 2:
Step 1 - Is the claim directed to a process, a machine, manufacture or composition of matter? - Yes, the claim is directed to a process.
Step 2A - Prong 1 - Does the claim recite an abstract idea, law of nature, or natural phenomenon? - Yes, the claim is dependent on claim 1 which included an abstract idea (see rejection for claim 1).
Step 2A - Prong 2 - Does the claim recite additional elements that integrate the judicial exception into a practical application? - No, there are no additional elements that integrate the judicial exception into a practical application. The additional elements:
wherein the second Al model includes a bidirectional encoder representations from transformers (BERT) model that is configured to generate a semantic representation of the data object: - This limitation does not integrate a judicial exception into a practical application as the AI model is recited at a high level of generality, therefore this amounts to mere instructions to implement an abstract idea 2106.05(f).
Step 2B - Does the claim recite additional elements that amount to significantly more than the judicial exception? - No, there are no additional elements that amount to significantly more than the judicial exception. The additional elements:
wherein the second Al model includes a bidirectional encoder representations from transformers (BERT) model that is configured to generate a semantic representation of the data object: - This limitation does not amount to significantly more than the judicial exception as the AI model is recited at a high level of generality, therefore this amounts to mere instructions to implement an abstract idea 2106.05(f).
Regarding Claim 3:
Step 1 - Is the claim directed to a process, a machine, manufacture or composition of matter? - Yes, the claim is directed to a process.
Step 2A - Prong 1 - Does the claim recite an abstract idea, law of nature, or natural phenomenon? - Yes, the claim is dependent on claim 2 which included an abstract idea (see rejection for claim 2). Additionally, claim 3 recites the abstract ideas:
wherein the second Al model further includes a feed-forward network (FFN) that is configured to map the semantic representation and the classification code to a space where contrastive loss is applied: - This claim is directed to the abstract idea of mathematical concepts, as the process of applying the contrastive loss function to the feed-forward network (FFN) that is configured to map the semantic representation and the classification code is a process of organizing information and manipulating information through mathematical correlation (see MPEP 2106.04(a)(2)(I) subsection A), therefore the claim amounts to mathematical concepts.
Step 2A - Prong 2 - Does the claim recite additional elements that integrate the judicial exception into a practical application? - No, there are no additional elements that integrate the judicial exception into a practical application.
Step 2B - Does the claim recite additional elements that amount to significantly more than the judicial exception? - No, there are no additional elements that amount to significantly more than the judicial exception.
Regarding Claim 4:
Step 1 - Is the claim directed to a process, a machine, manufacture or composition of matter? - Yes, the claim is directed to a process.
Step 2A - Prong 1 - Does the claim recite an abstract idea, law of nature, or natural phenomenon? - Yes, the claim is dependent on claim 3 which included an abstract idea (see rejection for claim 3). Additionally, claim 4 recites the abstract ideas:
wherein the second Al model further includes a contrastive loss function configured to identify whether the classification code is correct: - This claim is directed to the abstract idea of mathematical concepts, as the process of using the contrastive loss function configured to identify whether the classification code is correct is a process of organizing information and manipulating information through mathematical correlation (see MPEP 2106.04(a)(2)(I) subsection A), therefore the claim amounts to mathematical concepts.
Step 2A - Prong 2 - Does the claim recite additional elements that integrate the judicial exception into a practical application? - No, there are no additional elements that integrate the judicial exception into a practical application.
Step 2B - Does the claim recite additional elements that amount to significantly more than the judicial exception? - No, there are no additional elements that amount to significantly more than the judicial exception.
Regarding Claim 5:
Step 1 - Is the claim directed to a process, a machine, manufacture or composition of matter? - Yes, the claim is directed to a process.
Step 2A - Prong 1 - Does the claim recite an abstract idea, law of nature, or natural phenomenon? - Yes, the claim is dependent on claim 1 which included an abstract idea (see rejection for claim 1).
Step 2A - Prong 2 - Does the claim recite additional elements that integrate the judicial exception into a practical application? - No, there are no additional elements that integrate the judicial exception into a practical application. The additional element:
wherein the classification code is a hierarchical condition category (HCC) code: - This limitation is directed to selection of a particular data source or type of data to be manipulated as it is merely adding data, being insignificant extra-solution activity (see MEPEP n2106.05(g)).
Step 2B - Does the claim recite additional elements that amount to significantly more than the judicial exception? - No, there are no additional elements that amount to significantly more than the judicial exception.
wherein the classification code is a hierarchical condition category (HCC) code: - This limitation is analogous to electronic recordkeeping because it’s adding data to the training set and keeping record of it (see MPEP 2106.05(d) II (iii)).
Regarding Claim 6:
Step 1 - Is the claim directed to a process, a machine, manufacture or composition of matter? - Yes, the claim is directed to a process.
Step 2A - Prong 1 - Does the claim recite an abstract idea, law of nature, or natural phenomenon? - Yes, the claim is dependent on claim 5 which included an abstract idea (see rejection for claim 5).
Step 2A - Prong 2 - Does the claim recite additional elements that integrate the judicial exception into a practical application? - No, there are no additional elements that integrate the judicial exception into a practical application. The additional element:
wherein the data object is an electronic health record (EHR) and the entity is a patient: - This limitation is directed to selection of a particular data source or type of data to be manipulated as it is merely adding data, being insignificant extra-solution activity (see MEPEP n2106.05(g)).
Step 2B - Does the claim recite additional elements that amount to significantly more than the judicial exception? - No, there are no additional elements that amount to significantly more than the judicial exception. The additional element:
wherein the data object is an electronic health record (EHR) and the entity is a patient: - This limitation is analogous to electronic recordkeeping because it’s adding data to the training set and keeping record of it (see MPEP 2106.05(d) II (iii)).
Regarding Claim 7:
Step 1 - Is the claim directed to a process, a machine, manufacture or composition of matter? - Yes, the claim is directed to a process.
Step 2A - Prong 1 - Does the claim recite an abstract idea, law of nature, or natural phenomenon? - Yes, the claim is dependent on claim 1 which included an abstract idea (see rejection for claim 1).
Step 2A - Prong 2 - Does the claim recite additional elements that integrate the judicial exception into a practical application? - No, there are no additional elements that integrate the judicial exception into a practical application. The additional element:
wherein the second Al model is trained using a larger number of the negative data objects than the positive data objects per anchor: - This limitation does not integrate a judicial exception into a practical application as the AI model is recited at a high level of generality, therefore this amounts to mere instructions to implement an abstract idea 2106.05(f).
Step 2B - Does the claim recite additional elements that amount to significantly more than the judicial exception? - No, there are no additional elements that amount to significantly more than the judicial exception.
wherein the second Al model is trained using a larger number of the negative data objects than the positive data objects per anchor: - This limitation does not amount to significantly more than the judicial exception as the AI model is recited at a high level of generality, therefore this amounts to mere instructions to implement an abstract idea 2106.05(f).
Regarding claim 8, this claim is directed to a device and is rejected on the same basis as claim 1 since they are analogous claims.
Regarding claim 9, this claim is directed to a device and is rejected on the same basis as claim 2 since they are analogous claims.
Regarding claim 10, this claim is directed to a device and is rejected on the same basis as claim 3 since they are analogous claims.
Regarding claim 11, this claim is directed to a device and is rejected on the same basis as claim 4 since they are analogous claims.
Regarding claim 12, this claim is directed to a device and is rejected on the same basis as claim 5 since they are analogous claims.
Regarding claim 13, this claim is directed to a device and is rejected on the same basis as claim 6 since they are analogous claims.
Regarding claim 14, this claim is directed to a device and is rejected on the same basis as claim 7 since they are analogous claims.
Regarding claim 15, this claim is directed to a non-transitory computer-readable medium and is rejected on the same basis as claim 1 since they are analogous claims.
Regarding claim 16, this claim is directed to a non-transitory computer-readable medium and is rejected on the same basis as claim 2 since they are analogous claims.
Regarding claim 17, this claim is directed to a non-transitory computer-readable medium and is rejected on the same basis as claim 3 since they are analogous claims.
Regarding claim 18, this claim is directed to a non-transitory computer-readable medium and is rejected on the same basis as claim 4 since they are analogous claims.
Regarding claim 19, this claim is directed to a non-transitory computer-readable medium and is rejected on the same basis as claim 5 since they are analogous claims.
Regarding claim 20, this claim is directed to a non-transitory computer-readable medium and is rejected on the same basis as claim 6 since they are analogous claims.
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.
Claims 1, 4, 8, 11, 15, 18 are rejected under 35 U.S.C. 103 as being unpatentable over Zhang et al (US-12182524-B2- hereinafter Zhang) in view of Lee et al (KR-102313479-B1 - hereinafter Lee).
Referring to Claim 1, Zhang teaches:
validating, by the one or more processors and using a second Al model, the classification code for the data object that is assigned to the data object via the first Al model, wherein the second Al model is trained using positive data objects that include assigned classification codes that are correct and negative data objects that include assigned classification codes that are incorrect (see Zhang at Cl. 10 Ln. 20-32: “According to some aspects, fine-tuning component 325 trains the encoder 330 and/or the classification network 335 in a second training phase using a second contrastive learning loss based on a labeled positive sample pair including a first labeled hidden representation having a ground truth label and a second labeled hidden representation having the same ground truth label. For example, fine-tuning component 325 may perform supervised fine-tuning when there are limited training examples available to training unit 315 (for example, five and ten examples for an intent). Fine-tuning component 325 uses a supervised contrastive learning method to understand similar user intents. According to some aspects, fine-tuning component 325 trains at least one neural network such as encoder 330 and classification network 335 via a supervised contrastive learning method with an intent classification loss. In some cases, two utterances from a same class are treated by fine-tuning component 325 as a positive pair and two utterances across different classes are treated by fine-tuning component 325 as a negative pair for the purpose of contrastive learning. For example, same utterances could be a positive pair, and the positive pair can be input by training unit 315 to encoder 330 and/or classification network 335”. Examiner interprets the classification network (interpreted as second AI) being trained by fine-tuning component 325 via the supervised contrastive learning method with an intent classification loss and the pretraining of the encoder (interpreted as first AI) to be equivalent as the claimed “validating, by the one or more processors and using a second Al model, the classification code for the data object that is assigned to the data object via the first Al model”. Two utterances from the same class being treated by fine-tuning component 325 as a positive pair and two utterances across different classes being treated by fine-tuning component 325 as a negative pair for the purpose of contrastive learning is interpreted to be equivalent as the claimed “the second Al model is trained using positive data objects that include assigned classification codes that are correct and negative data objects that include assigned classification codes that are incorrect”);
and transmitting, by the one or more processors and to a user device, a validation result based on validating the classification code for the data object using the second Al model (see Zhang at Cl. 6 Ln. 31-67 and Cl. 7 Ln. 1-9: “At operation 205, the system provides a text phrase. In some cases, the operations of this step refer to, or may be performed by, a user as described with reference to FIG. 1. For example, a user can input a user utterance to a user device via hardware such a keyboard, mouse, touchscreen, microphone, etc., and/or software such as a graphical user interface, virtual keyboard, etc. For example, a user may be operating software on the user device, and may input an utterance into a prompt, text box, pop-up, etc. For example, the user may input an utterance such as “help me crop this photo” into photo-editing software.
(29) At operation 210, the system encodes the text phrase. In some cases, the operations of this step refer to, or may be performed by, a server as described with reference to FIG. 1. For example, the server can encode the text phrase using an encoder to obtain a hidden representation of the text phrase. In some embodiments, the encoder is trained during a first training phrase using self-supervised learning based on a first contrastive loss and during a second training phrase using supervised learning based on a second contrastive learning loss.
(30) At operation 215, the system identifies intent of the encoded text phrase. In some cases, the operations of this step refer to, or may be performed by, a server as described with reference to FIG. 1. For example, the server can identify an intent of the text phrase from a predetermined set of intent labels using a classification network. In some examples, the classification network is jointly trained with the encoder in the second training phase.
(31) At operation 220, the system generates response to text phrase. In some cases, the operations of this step refer to, or may be performed by, a server as described with reference to FIG. 1. For example, once the server has identified an intent of the text phrase, the server can generate an intent-accurate response. For example, after identifying the intent of the example utterance “help me to crop this photo”, the server can generate a response that instructs the photo-editing software to provide an appropriate prompt to a user that intends to crop a photograph.
(32) At operation 225, the system provides the response to user. In some cases, the operations of this step refer to, or may be performed by, a server as described with reference to FIG. 1. For example, the server can instruct the user device to display, via a display and the example photo-editing software, the example appropriate prompt relating to cropping a photo in the software”. Examiner interprets the classification network trained jointly with the encoder in the second training phase to be equivalent as the claimed “using the second Al model”. The server has identified an intent of the text phrase and generated the intent-accurate response is interpreted to be equivalent as the claimed “a validation result based on validating the classification code for the data object”. The response being provided to the user device by the server is interpreted to be equivalent as the claimed “transmitting, by the one or more processors and to a user device”).
However, Zhang fails to teach:
receiving, by one or more processors, a data object associated with an entity including a classification code that is assigned to the data object via a first artificial intelligence (Al) model.
Lee teaches, in analogous system,
receiving, by one or more processors, a data object associated with an entity including a classification code that is assigned to the data object via a first artificial intelligence (Al) model (see Lee at Pg. 130-132: “For example, before labeling information about an object in a work site video, the server (100) can provide a classification code management UI (20) (e.g., FIG. 9) for managing data of targets for action by type of object to a worker terminal (200), and through the classification code management UI (20), the name of the object, the code for the name, the target for action for the object, and the code for the target for action can be received as input, and according to the received data, "name of object - name code - target for action - target for action code" can be mapped and stored. Here, the server (100) can receive a setting on whether each object is a target for artificial intelligence model recognition through the classification code management UI (20), and can perform object extraction and labeling only on objects set as "targets". In various embodiments, the server (100) can perform individual labeling for each of the plurality of objects included in the work site image as illustrated in FIG. 10. For example, if the server (100) has a first object that is a worker and the attribute information for the first object is that it is not wearing protective gear, and the safety level for the first object is dangerous, it can perform labeling by coding the first object in the form of "Protective Gear_Safety Helmet_Safety Level Dangerous", that is, according to the object type action target data set through the classification code management UI (20), in the form of "PSN_SFH_X". Examiner interprets the server (server is interpreted as comprising processor) having a first object that is a worker (worker interpreted as entity) and the server performing labeling by coding the first object through the classification code management UI in the form of “PSN_SFH_X” as well as the server receiving a setting on whether each object is a target for artificial intelligence model recognition through the classification code management UI to be equivalent as the claimed “receiving, by one or more processors, a data object associated with an entity including a classification code that is assigned to the data object via a first artificial intelligence (Al) model”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Zhang with the above teachings of Lee by transmitting the validation result based on validating the classification code for the data object using the second Al model, as taught by Zhang, and receiving the data object associated with an entity including the classification code that is assigned to the data object via the first artificial intelligence model, as taught by Lee. The modification would have been obvious because one of ordinary skill in the art would be motivated to map the data of targets for action by type of object and store them to the worker terminal (as suggested by Lee at Pg. 130: “For example, before labeling information about an object in a work site video, the server (100) can provide a classification code management UI (20) (e.g., FIG. 9) for managing data of targets for action by type of object to a worker terminal (200), and through the classification code management UI (20), the name of the object, the code for the name, the target for action for the object, and the code for the target for action can be received as input, and according to the received data, "name of object - name code - target for action - target for action code" can be mapped and stored”).
Referring to Claim 4, Zhang teaches the method of claim 3:
wherein the second Al model further includes a contrastive loss function configured to identify whether the classification code is correct (see Zhang at Cl. 1 Ln. 50-64: “A method, apparatus, and non-transitory computer readable medium for natural language processing are described. One or more aspects of the method, apparatus, and non-transitory computer readable medium include receiving a text phrase; encoding the text phrase using an encoder to obtain a hidden representation of the text phrase, wherein the encoder is trained during a first training phrase using self-supervised learning based on a first contrastive loss and during a second training phrase using supervised learning based on a second contrastive teaming loss; identifying an intent of the text phrase from a predetermined set of intent labels using a classification network based on the hidden representation, wherein the classification network is jointly trained with the encoder in the second training phase; and generating a response to the text phrase based on the intent”. Examiner interprets the classification network to be equivalent as the claimed “second Al model”. The classification network is jointly trained with the encoder in the second training phase, it is said previously that the encoder is trained based on the contrastive teaming loss during the second training phase to identify the intent of the text phrase, therefore, Examiner interprets these statements to be equivalent as the claimed “the second Al model further includes a contrastive loss function configured to identify whether the classification code is correct”).
Referring to independent Claim 8, this claim is rejected on the same basis as independent claim 1 since they are analogous claims.
Referring to independent Claim 15, this claim is rejected on the same basis as independent claim 1 since they are analogous claims.
Referring to dependent Claim 11, this claim is rejected on the same basis as dependent claim 4 since they are analogous claims.
Referring to dependent Claim 18, this claim is rejected on the same basis as dependent claim 4 since they are analogous claims.
Claims 2, 9, 16 are rejected under 35 U.S.C. 103 as being unpatentable over Zhang et al (US-12182524-B2- hereinafter Zhang) in view of Lee et al (KR-102313479-B1 - hereinafter Lee) in further view of Liang et al (US-11947920-B2 – hereinafter Liang).
Referring to Claim 2, Zhang-Lee teaches the method of claim 1.
However, Zhang-Lee fails to teach:
wherein the second Al model includes a bidirectional encoder representations from transformers (BERT) model that is configured to generate a semantic representation of the data object.
Liang teaches, in analogous system,
wherein the second Al model includes a bidirectional encoder representations from transformers (BERT) model that is configured to generate a semantic representation of the data object (see Liang at Cl. 5 Ln. 39-58: “Further, in some examples, using the user dialogue sentence ContextX0, the goal type GT0 and the goal entity GE0 of the user dialogue sentence ContextX0 and the knowledge base data{SRO1, SRO2, SRO3 . . . } as the input of the second neural network module of the neural network system, and generating the reply sentence Y by performing the feature extraction and classification through the second neural network module, includes: the user dialogue sentence ContextX0 and the back-propagated reply sentence Y being input to a natural language processing model of the second neural network module, and encoded by the natural language processing model to extract semantics for generating a second vector Xy0; in some examples, the natural language processing model adopts a BERT (Bidirectional Encoder Representations from Transformers) model, and the encoder of the BERT model encodes the input user dialogue sentence ContextX0 and the back-propagated reply sentence Y to extract semantics for generating the second vector Xy0”. Examiner interprets the user dialogue sentence ContextX0 and the back-propagated reply sentence Y being input to the natural language processing model of the second neural network module, and encoded by the natural language processing model to extract semantics for generating a second vector Xy0 and the natural language processing model being the BERT (Bidirectional Encoder Representations from Transformers) model to be equivalent as the claimed “the second Al model includes a bidirectional encoder representations from transformers (BERT) model that is configured to generate a semantic representation of the data object”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Zhang and Lee with the above teachings of Liang by receiving and transmitting the validation result based on validating the classification code for the data object using the second Al model, as taught by Zhang and Lee, and configuring the bidirectional encoder representations from transformers (BERT) model to generate the semantic representation of the data object, as taught by Liang. The modification would have been obvious because one of ordinary skill in the art would be motivated to generate the reply sentence (as suggested by Liang at Cl. 5 Ln. 39-58: “Further, in some examples, using the user dialogue sentence ContextX0, the goal type GT0 and the goal entity GE0 of the user dialogue sentence ContextX0 and the knowledge base data{SRO1, SRO2, SRO3 . . . } as the input of the second neural network module of the neural network system, and generating the reply sentence Y by performing the feature extraction and classification through the second neural network module, includes: the user dialogue sentence ContextX0 and the back-propagated reply sentence Y being input to a natural language processing model of the second neural network module, and encoded by the natural language processing model to extract semantics for generating a second vector Xy0; in some examples, the natural language processing model adopts a BERT (Bidirectional Encoder Representations from Transformers) model, and the encoder of the BERT model encodes the input user dialogue sentence ContextX0 and the back-propagated reply sentence Y to extract semantics for generating the second vector Xy0”).
Referring to dependent Claim 9, this claim is rejected on the same basis as dependent claim 2 since they are analogous claims.
Referring to dependent Claim 16, this claim is rejected on the same basis as dependent claim 2 since they are analogous claims.
Claims 3, 10, 17 are rejected under 35 U.S.C. 103 as being unpatentable over Zhang et al (US-12182524-B2- hereinafter Zhang) in view of Lee et al (KR-102313479-B1 - hereinafter Lee) in further view of Liang et al (US-11947920-B2 – hereinafter Liang) and in further view of Wang et al (US-12315278-B2 – hereinafter Wang).
Referring to Claim 3, Zhang-Lee-Liang teaches the method of claim 2.
However, Zhang-Lee-Liang fails to teach:
wherein the second Al model further includes a feed-forward network (FFN) that is configured to map the semantic representation and the classification code to a space where contrastive loss is applied.
Wang teaches, in analogous system,
wherein the second Al model further includes a feed-forward network (FFN) that is configured to map the semantic representation and the classification code to a space where contrastive loss is applied (see Wang at CL. 18 Ln. 13-34: “After the image encoding component in the candidate text recognition model outputs the labeled semantic information, the labeled semantic information can be inputted into the feedforward network (may be understood as the classification network layer). The feedforward network can output the prediction text recognition result corresponding to the labeled images, where, input of the feedforward network is the labeled semantic information outputted by the image encoding component, output of the feedforward network is a vector, a dimensionality of the vector is equal to the number of text character categories, and if the candidate text recognition model is applicable to recognizing 300 text character categories, the output of the feedforward network may be a vector with a dimensionality being 300. The output vector of the feedforward network may serve as the prediction text recognition result of the labeled images in the candidate text recognition model, then, the error between the tag information of the labeled images and the prediction text recognition result can be calculated to calculate losses, optimize the network parameters of the candidate text recognition model, and the target text recognition model finally trained is obtained”. Examiner interprets the labeled semantic information being inputted into the feedforward network (classification network layer) after the recognition model outputs the labeled semantic information, and the error between the tag information of the labeled images and the prediction text recognition result being calculated to calculate losses to be equivalent as the claimed “wherein the second Al model further includes a feed-forward network (FFN) that is configured to map the semantic representation and the classification code to a space where contrastive loss is applied”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Zhang, Lee and Liang with the above teachings of Wang by validating the classification code for the data object using the second Al model configured with bidirectional encoder representations from transformers (BERT) model, as taught by Zhang, Lee and Liang, and configuring the second Al model based on the feed-forward network (FFN) to map the semantic representation and the classification code to a space where contrastive loss is applied, as taught by Wang. The modification would have been obvious because one of ordinary skill in the art would be motivated to calculate the losses caused by the error between the tag information of the labeled images and the prediction text recognition result (as suggested by Wang at CL. 18 Ln. 13-34: “After the image encoding component in the candidate text recognition model outputs the labeled semantic information, the labeled semantic information can be inputted into the feedforward network (may be understood as the classification network layer). The feedforward network can output the prediction text recognition result corresponding to the labeled images, where, input of the feedforward network is the labeled semantic information outputted by the image encoding component, output of the feedforward network is a vector, a dimensionality of the vector is equal to the number of text character categories, and if the candidate text recognition model is applicable to recognizing 300 text character categories, the output of the feedforward network may be a vector with a dimensionality being 300. The output vector of the feedforward network may serve as the prediction text recognition result of the labeled images in the candidate text recognition model, then, the error between the tag information of the labeled images and the prediction text recognition result can be calculated to calculate losses, optimize the network parameters of the candidate text recognition model, and the target text recognition model finally trained is obtained”).
Referring to dependent Claim 10, this claim is rejected on the same basis as dependent claim 3 since they are analogous claims.
Referring to dependent Claim 17, this claim is rejected on the same basis as dependent claim 3 since they are analogous claims.
Claims 5-6, 12-13, 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over Zhang et al (US-12182524-B2- hereinafter Zhang) in view of Lee et al (KR-102313479-B1 - hereinafter Lee) in further view of Kattan et al (US-20240079119-A1 – hereinafter Kattan).
Referring to Claim 5, Zhang-Lee teaches the method of claim 1.
However, Zhang-Lee fails to teach:
wherein the classification code is a hierarchical condition category (HCC) code.
Kattan teaches, in analogous system,
wherein the classification code is a hierarchical condition category (HCC) code (see Kattan at Pg. 32: “The medical data (e.g., held by any of the data sources 170, and/or the database 118) may include any medical data. Examples of the medical data include: disease classification codes (e.g., International Classification of Diseases (ICD) codes); Health and Human Services (HHS)-Hierarchical Condition Categories (HHS-HCC) data; Centers for Medicare & Medicaid Services (CMS)-HCC data; information of a hospice stay; dates of death; information of stays at a skilled nursing facility (SNF); information of stays at a residential boarding house (BH); Chronic Conditions Wearhouse (CCW) data; ages of the patients; numbers of emergency department (ED) visits; indications of lines of business (LOBs) being any of commercial, Medicare, or Medicare; information of a diagnosis related group (DRG); and/or information of a behavioral health inpatient stay. In addition, the medical data may be respective medical data; that is, the medical data corresponds to individual patients, and thus may be used to classify the individual patients”. Examiner interprets the disease classification codes including Health and Human Services (HHS)-Hierarchical Condition Categories (HHS-HCC) data to be equivalent as the claimed “the classification code is a hierarchical condition category (HCC) code”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Zhang and Lee with the above teachings of Kattan by receiving and transmitting the validation result based on validating the classification code for the data object using the second Al model, as taught by Zhang and Lee, and the classification code being the hierarchical condition category (HCC) code, as taught by Kattan. The modification would have been obvious because one of ordinary skill in the art would be motivated to use the hierarchical condition category (HCC) code to classify the individual patients (as suggested by Kattan at Pg. 32: “The medical data (e.g., held by any of the data sources 170, and/or the database 118) may include any medical data. Examples of the medical data include: disease classification codes (e.g., International Classification of Diseases (ICD) codes); Health and Human Services (HHS)-Hierarchical Condition Categories (HHS-HCC) data; Centers for Medicare & Medicaid Services (CMS)-HCC data; information of a hospice stay; dates of death; information of stays at a skilled nursing facility (SNF); information of stays at a residential boarding house (BH); Chronic Conditions Wearhouse (CCW) data; ages of the patients; numbers of emergency department (ED) visits; indications of lines of business (LOBs) being any of commercial, Medicare, or Medicare; information of a diagnosis related group (DRG); and/or information of a behavioral health inpatient stay. In addition, the medical data may be respective medical data; that is, the medical data corresponds to individual patients, and thus may be used to classify the individual patients”).
Referring to Claim 6, Zhang-Lee teaches the method of claim 5.
However, Zhang-Lee fails to teach:
wherein the data object is an electronic health record (EHR) and the entity is a patient.
Kattan teaches, in analogous system,
wherein the data object is an electronic health record (EHR) and the entity is a patient (see Kattan at Pg. 82: “The method 600 may optionally include the patient classifier 124 adding the classifications to electronic medical records (EMRs) of the patients (block 630). Advantageously, in some implementations, this allows for technical improvements. For example, a user may wish to predict an optimal staffing level of a particular type of caregiver. Here, if the categories have been added to the EMRs of the patients, the organizational predictor 126 is able to simply draw the data from the EMRs patients of the healthcare provider 150 for which the prediction is being made rather than having the patient classifier 124 classify every patient based on medical data”. Examiner interprets the electronic medical records (EMRs) of the patients to be equivalent as the claimed “the data object is an electronic health record (EHR) and the entity is a patient”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Zhang and Lee with the above teachings of Kattan by receiving and transmitting the validation result based on validating the classification code for the data object using the second Al model, as taught by Zhang and Lee, and the data object being the electronic health record (EHR) and the entity is a patient, as taught by Kattan. The modification would have been obvious because one of ordinary skill in the art would be motivated to predict the optimal staffing level of a particular type of caregiver (as suggested by Kattan at Pg. 32: “The method 600 may optionally include the patient classifier 124 adding the classifications to electronic medical records (EMRs) of the patients (block 630). Advantageously, in some implementations, this allows for technical improvements. For example, a user may wish to predict an optimal staffing level of a particular type of caregiver. Here, if the categories have been added to the EMRs of the patients, the organizational predictor 126 is able to simply draw the data from the EMRs patients of the healthcare provider 150 for which the prediction is being made rather than having the patient classifier 124 classify every patient based on medical data”).
Referring to dependent Claim 12, this claim is rejected on the same basis as dependent claim 5 since they are analogous claims.
Referring to dependent Claim 19, this claim is rejected on the same basis as dependent claim 5 since they are analogous claims.
Referring to dependent Claim 13, this claim is rejected on the same basis as dependent claim 6 since they are analogous claims.
Referring to dependent Claim 20, this claim is rejected on the same basis as dependent claim 6 since they are analogous claims.
Claims 7, 14 are rejected under 35 U.S.C. 103 as being unpatentable over Zhang et al (US-12182524-B2- hereinafter Zhang) in view of Lee et al (KR-102313479-B1 - hereinafter Lee) in further view of Voloshynovskiy et al (US-12293007-B2– hereinafter Voloshynovskiy).
Referring to Claim 7, Zhang-Lee teaches the method of claim 1.
However, Zhang-Lee fails to teach:
wherein the second Al model is trained using a larger number of the negative data objects than the positive data objects per anchor.
Voloshynovskiy teaches, in analogous system,
wherein the second Al model is trained using a larger number of the negative data objects than the positive data objects per anchor (see Voloshynovskiy at Cl. 18 Ln. 39-55: “The trained embedding is used for the training of the classifier. Alternatively, the minimization/maximization of distances can be extended to the maximization/minimization of corresponding mutual information such as will be explained in more detail in the following description. This system resembles a triplet loss architecture [86] that is well known for example in biometrics applications and content retrieval systems. In addition, the training can also incorporate not only pairwise distances between the pairs of fingerprints representing original and fakes but also more general cases, when the distance between a particular fingerprint and several examples of fakes are used or between a particular fingerprint and several fakes corresponding to other fingerprints close to xi. The examples of training strategies of such kind of systems can be based on multi-class N-pair loss [87] or NT-Xent loss [88] also extended to the supervised case [89]”. Examiner interprets the classifier to be equivalent as the claimed “second Al model”. The distance between a particular fingerprint (interpreted as positive data) and several examples of fakes (interpreted as negative data) are used or the distance between a particular fingerprint (interpreted as positive data) and several fakes corresponding to other fingerprints (interpreted as negative data) close to xi (interpreted as anchor). The previous statement shows that the number of fakes fingerprints is greater than the particular fingerprint close to xi, therefore, the statement is interpreted to be equivalent as the claimed “a larger number of the negative data objects than the positive data objects per anchor”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Zhang and Lee with the above teachings of Voloshynovskiy by receiving and transmitting the validation result based on validating the classification code for the data object using the second Al model, as taught by Zhang and Lee, and training the second Al model using a larger number of the negative data objects than the positive data objects per anchor, as taught by Voloshynovskiy. The modification would have been obvious because one of ordinary skill in the art would be motivated to maximize the distance between a particular fingerprint and several examples of fakes (as suggested by Voloshynovskiy at Cl. 18 Ln. 39-55: “The trained embedding is used for the training of the classifier. Alternatively, the minimization/maximization of distances can be extended to the maximization/minimization of corresponding mutual information such as will be explained in more detail in the following description. This system resembles a triplet loss architecture [86] that is well known for example in biometrics applications and content retrieval systems. In addition, the training can also incorporate not only pairwise distances between the pairs of fingerprints representing original and fakes but also more general cases, when the distance between a particular fingerprint and several examples of fakes are used or between a particular fingerprint and several fakes corresponding to other fingerprints close to xi. The examples of training strategies of such kind of systems can be based on multi-class N-pair loss [87] or NT-Xent loss [88] also extended to the supervised case [89]”).
Referring to dependent Claim 14, this claim is rejected on the same basis as dependent claim 7 since they are analogous claims.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to AWADAGBE G HOUNTON whose telephone number is (571)270-0670. The examiner can normally be reached Monday-Friday 8am-5pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, David Yi can be reached at (571) 270-7519. 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.
/AWADAGBE G HOUNTON/Examiner, Art Unit 2126
/DAVID YI/Supervisory Patent Examiner, Art Unit 2126
Prosecution Insights