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 .
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.
Claims 1-10 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception without significantly more.
Regarding Claim 1:
Step 1 – Is the claim to a process, machine, manufacture, or composition of matter?
Yes
Step 2A – Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
Yes, the claim recites the abstract ideas of:
calculate a composite feature vector indicating features of a composite object obtained by combining the plurality of component objects; This limitation is directed to the abstract idea of a mathematical concepts, as calculating a feature vector is analogous to a mathematical calculation (see MPEP 2106.04(a)(2) I. C.)
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.
An information processing system comprising: at least one processor, wherein the at least one processor is configured to: This limitation recites generic computer components such as processor, which invokes a system merely as a tool for performing an existing process [see MPEP 2106.05(f)(2)] and therefore fails to integrate the exception into a practical application.
acquire a numerical representation and a combination ratio for each of a plurality of component objects; This limitation is directed to mere data gathering, which is an insignificant extra-solution activity [see MPEP 2106.05(g)(3)] and therefore fails to integrate the judicial exception into a practical application.
execute, based on a plurality of the numerical representations and a plurality of the combination ratios corresponding to the plurality of component objects, machine learning and application of the plurality of combination ratios. The claim further recites using machine learning model and training data. However, using machine learning merely provides instructions to apply the the mathematical concept and therefore does not integrate the judicial exception into a practical application. The claim does not improve the functioning of a computer or another technology (MPEP 2106.05(f)).
output the composite feature vector. This limitation recites outputting the vector, which merely as performing to output data [see MPEP 2106.05(f)(2)] and therefore fails to integrate the exception into a practical application.
Step 2B – Does the claim recite any 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.
An information processing system comprising: at least one processor, wherein the at least one processor is configured to: This limitation invokes a system merely as a tool for performing an existing process [see MPEP 2106.05(f)(2)] and therefore fails to amount to significantly more than the judicial exception.
acquire a numerical representation and a combination ratio for each of a plurality of component objects; This limitation is directed to receiving or transmitting data over a network, which the courts have recognized as well-understood, routine, conventional activity when they are claimed at a high level of generality or as insignificant extra-solution activity [see MPEP 2106.05(d) II. i] and therefore fails to amount to significantly more than the judicial exception.
execute, based on a plurality of the numerical representations and a plurality of the combination ratios corresponding to the plurality of component objects, machine learning and application of the plurality of combination ratios. The additional elements, machine learning model and training data, do not amount to significantly more than the abstract idea. A machine learning model, merely provides instructions to apply the mathematical concept. The claim does not improve the functioning of a computer or another technology (MPEP 2016.04(d)(1)).
output the composite feature vector. This limitation recites outputting the vector merely performing to output data [see MPEP 2106.05(f)(2)] and therefore fails to amount to significantly more than the judicial exception.
Step 2A Prong Two and Step 2B:
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B. The claim is ineligible.
Regarding Claim 2:
Step 1 – Is the claim to a process, machine, manufacture, or composition of matter?
Yes
Step 2A – Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
Yes, the claim recites the abstract ideas of:
calculate a feature vector of each of the plurality of component objects; This limitation is directed to the abstract idea of a mathematical concepts, as calculating a feature vector is analogous to a mathematical calculation (see MPEP 2106.04(a)(2) I. C.)
calculate the composite feature vector. This limitation is directed to the abstract idea of a mathematical concepts, as calculating a feature vector is analogous to a mathematical calculation (see MPEP 2106.04(a)(2) I. C.)
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.
input the plurality of numerical representations into a machine learning model. This limitation recites as an insignificant extra solution activity, as inputting the data into a device, under BRI, is mere data gathering per MPEP 2106.05(g)(3).
execute the application of the plurality of combination ratios in association with the machine learning model; The claim further recites using machine learning model and training data. However, the limitation merely provides applying mathematical calculations, and therefore does not integrate the judicial exception into a practical application. The claim does not improve the functioning of a computer or another technology (MPEP 2106.05(f)).
input a plurality of the feature vectors reflecting the plurality of combination ratios into an aggregation function. This limitation recites as an insignificant extra solution activity, as inputting the data into a device, under BRI, is mere data gathering per MPEP 2106.05(g)(3).
Step 2B – Does the claim recite any 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.
input the plurality of numerical representations into a machine learning model. This limitation is directed to receiving data, which the courts have recognized as well-understood, routine, conventional activity when they are claimed at a high level of generality or as insignificant extra-solution activity [see MPEP 2106.05(d) II. i] and therefore fails to amount to significantly more than the judicial exception.
execute the application of the plurality of combination ratios in association with the machine learning model. The additional elements, machine learning model and training data, do not amount to significantly more than the abstract idea. A machine learning model, merely provides instructions to apply the mathematical concept. The claim does not improve the functioning of a computer or another technology (MPEP 2016.04(d)(1)).
input a plurality of the feature vectors reflecting the plurality of combination ratios into an aggregation function. This limitation is directed to receiving data, which the courts have recognized as well-understood, routine, conventional activity when they are claimed at a high level of generality or as insignificant extra-solution activity [see MPEP 2106.05(d) II. i] and therefore fails to amount to significantly more than the judicial exception.
Step 2A Prong Two and Step 2B:
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B. The claim is ineligible.
Regarding Claim 3:
Step 1 – Is the claim to a process, machine, manufacture, or composition of matter?
Yes
Step 2A – Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
Yes, the claim recites the abstract ideas of:
calculate a first feature vector of each of the plurality of component objects; This limitation is directed to the abstract idea of a mathematical concepts, as calculating a feature vector is analogous to a mathematical calculation (see MPEP 2106.04(a)(2) I. C.)
calculate a second feature vector of each of the plurality of component objects; This limitation is directed to the abstract idea of a mathematical concepts, as calculating a feature vector is analogous to a mathematical calculation (see MPEP 2106.04(a)(2) I. C.)
calculate the composite feature vector. This limitation is directed to the abstract idea of a mathematical concepts, as calculating a feature vector is analogous to a mathematical calculation (see MPEP 2106.04(a)(2) I. C.)
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.
input the plurality of numerical representations into a first machine learning model. This limitation recites as an insignificant extra solution activity, as inputting the data into a device, under BRI, is mere data gathering per MPEP 2106.05(g)(3).
input a plurality of the first feature vectors into a second machine learning model. This limitation recites as an insignificant extra solution activity, as inputting the data into a device, under BRI, is mere data gathering per MPEP 2106.05(g)(3).
execute the application of the plurality of combination ratios in association with at least one machine learning model selected from the first machine learning model and the second machine learning model. The claim further recites using machine learning model and training data. However, the limitation merely provides instructions to apply mathematical calculations and therefore does not integrate the judicial exception into a practical application. The claim does not improve the functioning of a computer or another technology (MPEP 2106.05(f)).
input a plurality of the second feature vectors reflecting the plurality of combination ratios into an aggregation function. This limitation recites as an insignificant extra solution activity, as inputting the data into a device, under BRI, is mere data gathering per MPEP 2106.05(g)(3).
Step 2B – Does the claim recite any 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.
input the plurality of numerical representations into a first machine learning model. This limitation is directed to receiving data, which the courts have recognized as well-understood, routine, conventional activity when they are claimed at a high level of generality or as insignificant extra-solution activity [see MPEP 2106.05(d) II. i] and therefore fails to amount to significantly more than the judicial exception.
input a plurality of the first feature vectors into a second machine learning model. This limitation is directed to receiving data, which the courts have recognized as well-understood, routine, conventional activity when they are claimed at a high level of generality or as insignificant extra-solution activity [see MPEP 2106.05(d) II. i] and therefore fails to amount to significantly more than the judicial exception.
execute the application of the plurality of combination ratios in association with at least one machine learning model selected from the first machine learning model and the second machine learning model. The additional elements, machine learning model and training data, do not amount to significantly more than the abstract idea. A machine learning model, merely provides instructions to apply the mathematical concept. The claim does not improve the functioning of a computer or another technology (MPEP 2016.04(d)(1)).
input a plurality of the second feature vectors reflecting the plurality of combination ratios into an aggregation function. This limitation is directed to receiving data, which the courts have recognized as well-understood, routine, conventional activity when they are claimed at a high level of generality or as insignificant extra-solution activity [see MPEP 2106.05(d) II. i] and therefore fails to amount to significantly more than the judicial exception.
Step 2A Prong Two and Step 2B:
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B. The claim is ineligible.
Regarding Claim 4:
Step 1 – Is the claim to a process, machine, manufacture, or composition of matter?
Yes
Step 2A – Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
a machine learning model which generates the first feature vector that is a fixed-length vector from the numerical representation. This limitation is directed to the abstract idea of a mathematical concepts, as generating a fixed-length vector is analogous to a mathematical calculation. (see MPEP 2106.04(a)(2) I. C.)
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.
wherein the first machine learning model is a machine learning model which generates the first feature vector that is a fixed-length vector from the numerical representation that is unstructured data. This limitation recites the machine learning model generating a fixed length vector from unstructured data. Therefore, this limitation amounts to merely indicating a field of use or technological environment [see MPEP 2106.05(h)] and fails to integrate the judicial exception into a practical application.
Step 2B – Does the claim recite any 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 first machine learning model is a machine learning model which generates the first feature vector that is a fixed-length vector from the numerical representation that is unstructured data. This limitation recites description of the model generating a fixed length vector from unstructured data. Therefore, this limitation amounts to merely indicating a field of use or technological environment [see MPEP 2106.05(h)] and therefore fails to amount to significantly more than the judicial exception.
Step 2A Prong Two and Step 2B:
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B. The claim is ineligible.
Regarding Claim 5:
Step 1 – Is the claim to a process, machine, manufacture, or composition of matter?
Yes
Step 2A – Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
Claim 5 does not recite abstract ideas other than the ones recited at 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.
wherein the application of the plurality of combination ratios in association with the machine learning model comprises applying the plurality of combination ratios to output data of an intermediate layer of the machine learning model. This limitation recites an insignificant extra solution activity, as outputting information, under BRI, is mere data outputting per MPEP 2106.05(g)(3).
Step 2B – Does the claim recite any 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 application of the plurality of combination ratios in association with the machine learning model comprises applying the plurality of combination ratios to output data of an intermediate layer of the machine learning model. This limitation is directed to outputting/transmitting data, which the courts have recognized as well-understood, routine, conventional activity when they are claimed at a high level of generality or as insignificant extra-solution activity [see MPEP 2106.05(d) II.i] and therefore fails to amount to significantly more than the judicial exception.
Step 2A Prong Two and Step 2B:
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B. The claim is ineligible.
Regarding Claim 6:
Step 1 – Is the claim to a process, machine, manufacture, or composition of matter?
Yes
Step 2A – Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
calculate a predicted value of characteristics of the composite object; This limitation is directed to the abstract idea of a mathematical concepts, as calculating a feature vector is analogous to a mathematical calculation (see MPEP 2106.04(a)(2) I. C.)
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.
input the composite feature vector into another machine learning model. This limitation is directed to receiving data, which the courts have recognized as well-understood, routine, conventional activity when they are claimed at a high level of generality or as insignificant extra-solution activity [see MPEP 2106.05(d) II. i] and therefore fails to amount to significantly more than the judicial exception.
output the predicted value. This limitation recites an insignificant extra solution activity, as outputting information, under BRI, is mere data outputting per MPEP 2106.05(g)(3).
Step 2B – Does the claim recite any 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.
input the composite feature vector into another machine learning model. This limitation is directed to receiving data, which the courts have recognized as well-understood, routine, conventional activity when they are claimed at a high level of generality or as insignificant extra-solution activity [see MPEP 2106.05(d) II. i] and therefore fails to amount to significantly more than the judicial exception.
output the predicted value. This limitation is directed to outputting/transmitting data, which the courts have recognized as well-understood, routine, conventional activity when they are claimed at a high level of generality or as insignificant extra-solution activity [see MPEP 2106.05(d) II.i] and therefore fails to amount to significantly more than the judicial exception.
Step 2A Prong Two and Step 2B:
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B. The claim is ineligible.
Regarding Claim 7:
Step 1 – Is the claim to a process, machine, manufacture, or composition of matter?
Yes
Step 2A – Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
Claim 7 does not recite abstract ideas other than the ones recited at 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.
wherein the component object is a material, and the composite object is a multi-component substance. This limitation recites the type of component object and the composite object. Therefore, this limitation amounts to merely indicating a field of use or technological environment [see MPEP 2106.05(h)] and fails to integrate the judicial exception into a practical application.
Step 2B – Does the claim recite any 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 component object is a material, and the composite object is a multi-component substance. This limitation recites the type of component object and the composite object. Therefore, this limitation amounts to merely indicating a field of use or technological environment [see MPEP 2106.05(h)] and therefore fails to amount to significantly more than the judicial exception.
Step 2A Prong Two and Step 2B:
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B. The claim is ineligible.
Regarding Claim 8:
Step 1 – Is the claim to a process, machine, manufacture, or composition of matter?
Yes
Step 2A – Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
Claim 8 does not recite abstract ideas other than the ones recited at 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.
wherein the material is a polymer, and the multi-component substance is a polymer alloy. This limitation recites the type of the polymer and the multi-component substance. Therefore, this limitation amounts to merely indicating a field of use or technological environment [see MPEP 2106.05(h)] and fails to integrate the judicial exception into a practical application.
Step 2B – Does the claim recite any 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 material is a polymer, and the multi-component substance is a polymer alloy. wherein the component object is a material, and the composite object is a multi-component substance. This limitation recites the type of the polymer and the multi-component substance. Therefore, this limitation amounts to merely indicating a field of use or technological environment [see MPEP 2106.05(h)] and therefore fails to amount to significantly more than the judicial exception.
Step 2A Prong Two and Step 2B:
Thus, the judicial exception is not integrated into a practical application (see MPEP 2106.04(d) I.), failing step 2A prong 2. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception under step 2B. The claim is ineligible.
Regarding claims 9- 10
Claims 9 - 10 recites analogous limitations to claims 1 (respectively) and therefore they are rejected on the same grounds as claims 1.
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 non-obviousness.
Claims 1 – 6 and claims 9 - 10, are rejected under 35 U.S.C. 103 as being unpatentable over Goodall (NPL, “Predicting materials properties without crystal structure: Deep representation learning from stoichiometry (v2)” dated on 10/29/2019, by Goodall et al – hereinafter Goodall) in view of Sandler (WIPO, WO2018084974A1, by Sandler et al - hereinafter Sandler).
Referring to Claim 1, Goodall teaches:
acquire a numerical representation and a combination ratio for each of a plurality of component objects; See Goodall at [Page 2, mid-left] “For each material, we then construct a dense graph where the vertices are labelled by both the elemental feature vector from the surrogate embedding and the fraction of the element in the material.” Examiner interprets the elemental feature as equivalent to the numerical representation, and the fraction of the element in the material as equivalent to the combination ratio since percentages can be represented in fractional form.
execute, based on a plurality of the numerical representations and a plurality of the combination ratios corresponding to the plurality of component objects, machine learning and application of the plurality of combination ratios to calculate a composite feature vector indicating features of a composite object obtained by combining the plurality of component objects; See Goodall at [II.A. Message Passing Model, from Page 2 bottom-left to Page 3 top-left]:” For each element in the model’s input domain, we start with an embedding vector that differentiates between different elements. In this work, we use embedding vectors derived from applying the Word2Vec algorithm to a corpus of scientific abstracts…This process allows for material-specific representations for the elements in a given material to be learned automatically. Mathematically;
PNG
media_image1.png
38
322
media_image1.png
Greyscale
where hi is the feature vector for the ith element… Ut(h) is the element update function… The first stage of the attention mechanism is to compute unnormalised coefficients, eij , across pairs of elements in the material.
PNG
media_image2.png
46
304
media_image2.png
Greyscale
where f(…) is a single hidden layer neural network, the j index runs over all the elements in vi, and ||is the concatenation operation. We then normalise these coefficients using a weighted softmax function where the weights, wj , are the fraction of the neighbouring element in the total material.
PNG
media_image3.png
62
348
media_image3.png
Greyscale
The elemental representations are then updated with new pair-dependent perturbations that are weighted by the soft-attention coefficients.
PNG
media_image4.png
66
380
media_image4.png
Greyscale
where gn(…) are again single hidden layer neural networks... A fixed-length representation for each material is then determined via a similar weighted soft-attention based pooling operation that considers each element in a material in turn and decides, given its representation, how much attention to pay to its presence when constructing the material’s representation.” Examiner interprets the embedding vectors for elements as equivalent to the numerical representations; fractions of the element as equivalent to the combination ratios; message passing neural network as the machine learning; and the fixed-length representation as equivalent to the composite feature vector. Specifically, Goodall starts with embedding vectors for elements, maps the elemental feature vector, computes weighted attention coefficients aij using fractions wj of neighboring elements, updates the feature vectors using weighted aggregation, and then determines a fixed-length representation for each material using a weighted soft-attention pooling operation.
output the composite feature vector. Examiner interprets the fixed-length representation in the above limitation, which is the result of the machine learning model, as equivalent to an output as claimed.
however, it fails to teach
An information processing system comprising: at least one processor, wherein the at least one processor is configured to.
Sandler teaches,
An information processing system comprising: at least one processor, wherein the at least one processor is configured to. See Sandler at [0007]:” Another example aspect of the present disclosure is directed to one or more tangible, non-transitory computer-readable media storing computer-readable instructions that when executed by one or more processors cause the one or more processors to perform operations. The operations include accessing data indicative of a connection pattern associated with a convolutional neural network.” Examiner interprets the computer readable media, processors and instructions as equivalent as the system and the computer components as claimed.
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 Goodall with the above teachings of Sandler by utilizing machine learning to combine the numerical features and ratios of multiple component objects to calcite and output a composite feature vector, as taught by Goodall, an information processing system comprising computer components, as taught by Sandler. The modification would have been obvious because one of ordinary skill in the art would be motivated to improve the generalization capability of the models being trained. See Sandler at [0033]: “The training computing system 150 can include a model trainer 160 that trains the CNNs 140 stored at the server computing system 130 using various training or learning techniques, such as, for example, backwards propagation of errors. In some implementations, performing backwards propagation of errors can include performing truncated backpropagation through time. The model trainer 160 can perform a number of generalization techniques (e.g., weight decays, dropouts, etc.) to improve the generalization capability of the models being trained. In particular, the model trainer 160 can train a CNN 140 based on a set of training data 162.”
Referring to Claim 2, Goodall-Sandler teaches the information processing system of claim 1, Goodall also teaches:
input the plurality of numerical representations into a machine learning model to calculate a feature vector of each of the plurality of component objects; See Goodall at [Page 2, bottom-left]:” In this work, we use embedding vectors derived from applying the Word2Vec algorithm to a corpus of scientific abstracts [21, 22]. We map these initial feature vectors onto a surrogate embedding by using a trainable weight matrix.” Also see Goodall [Page 2, bottom-left]:” The next stages use message-passing operations to update these elemental features, thereby propagating contextual information about the different elements present in the material throughout its graph. This process allows for material-specific representations for the elements in a given material tobe learned automatically. Mathematically; where hi is the feature vector for the ith element.”
PNG
media_image5.png
65
404
media_image5.png
Greyscale
Examiner interprets mapping and updating the feature vector (h, formular (1)) as equivalent calculating the feature vector as claimed.
execute the application of the plurality of combination ratios in association with the machine learning model; See Goodall at [Page 2, mid-right]:” See Goodall at [Page 2, mid-right]:” We then normalise these coefficients using a weighted softmax function where the weights, wj , are the fraction of the neighbouring element in the total material.”
PNG
media_image3.png
62
348
media_image3.png
Greyscale
Examiner interprets Goodall executing the application of a plurality of combination ratios in association with the machine learning model because Goodall applies weighted fractions wj of neighboring elements within a weighted soft attention mechanism of the message-passing neural network. Specifically, the attention coefficients aij are computed using the fractions wj in equation (3).
input a plurality of the feature vectors reflecting the plurality of combination ratios into an aggregation function to calculate the composite feature vector. See Goodall at [Page 2, bottom-right]:” The elemental representations are then updated with new pair-dependent perturbations that are weighted by the soft-attention coefficients. where gn(…) are again single hidden layer neural networks.”
PNG
media_image6.png
70
372
media_image6.png
Greyscale
Examiner interprets Goodall updates hi , the feature vectors in equation (4), by using sum operation (∑) to aggregating the aij coefficients which is determined by the fraction in equation (3). Examiner interprets Goodall teaches inputting the feature vectors reflecting the combination ratios. Also, see Goodall at [Page 3, top-left]:” A fixed-length representation for each material is then determined via a similar weighted soft-attention based pooling operation that considers each element in a material in turn and decides, given its representation, how much attention to pay to its presence when constructing the material’s representation.” Examiner interprets the sum operation and the pooling operation as equivalent as the aggregation function, and determining the fixed-length representation as equivalent as calculating the composite feature vector since it is well known in the art that pooling operation is a type of local aggregation operation used to reduce the spatial dimensionality of feature maps while retaining essential information.
Referring to Claim 3, Goodall- Sandler teaches the information processing system of claim 1, Goodall also teaches:
input the plurality of numerical representations into a first machine learning model to calculate a first feature vector of each of the plurality of component objects; See Goodall at [Page 2, mid-left]:” A. Message Passing Model …For each element in the model’s input domain, we start with an embedding vector that differentiates between different elements. In this work, we use embedding vectors derived from applying the Word2Vec algorithm to a corpus of scientific abstracts [21, 22]. We map these initial feature vectors onto a surrogate embedding by using a trainable weight matrix. The dimensionality of the surrogate embedding, d, determines the size of the hidden states for the message passing.” Also, See Goodall writes at [Page 2, bottom-left]:” This process allows for material-specific representations for the elements in a given material to be learned automatically. Mathematically;
PNG
media_image5.png
65
404
media_image5.png
Greyscale
Where hi is the feature vector for the ith element. Vi ={hα, hβ, hϒ …} is the set of elements that are the neighbours of hi, and Ut(h)() is the element update function.” Examiner interprets hi as equivalent as the first feature vector; Ut(h)(), as equivalent as an update function, performing the function of first machine learning model, which is as equivalent as calculating a first feature vector as claimed. Thus, Goodall teaches the limitation.
input a plurality of the first feature vectors into a second machine learning model to calculate a second feature vector of each of the plurality of component objects. See Goodall at [Page 2, mid-right]:” The first stage of the attention mechanism is to compute unnormalised coefficients, eij , across pairs of elements in the material…
PNG
media_image2.png
46
304
media_image2.png
Greyscale
where f(…) is a single hidden layer neural network, the j index runs over all the elements in vi, and || is the concatenation operation. We then normalise these coefficients using a weighted softmax function where the weights, wj , are the fraction of the neighbouring element in the total material.
PNG
media_image3.png
62
348
media_image3.png
Greyscale
The elemental representations are then updated with new pair-dependent perturbations that are weighted by the soft-attention coefficients.
PNG
media_image6.png
70
372
media_image6.png
Greyscale
where gn(…) are again single hidden layer neural networks. Examiner interprets the functions and calculations of the network from equation (2) to equation (4) as equivalent as the second machine learning model. The feature vector hi is updated and calculated in equation (2) and (4) as a result of second feature vector as claimed.
execute the application of the plurality of combination ratios in association with at least one machine learning model selected from the first machine learning model and the second machine learning model; See Goodall at [Page 2, mid-right]:” We then normalise these coefficients using a weighted softmax function where the weights, wj , are the fraction of the neighbouring element in the total material.
PNG
media_image3.png
62
348
media_image3.png
Greyscale
The elemental representations are then updated with new pair-dependent perturbations that are weighted by the soft-attention coefficients. Examiner interprets using the fraction as the weights wj to normalize the coefficients aij in the equation (3) as equivalent as executing the application of combination ratios in the second machine learning model.
input a plurality of the second feature vectors reflecting the plurality of combination ratios into an aggregation function to calculate the composite feature vector. See Goodall at [Page 3, top-left]:” A fixed-length representation for each material is then determined via a similar weighted soft-attention based pooling operation that considers each element in a material in turn and decides, given its representation, how much attention to pay to its presence when constructing the material’s representation.” Examiner interprets the representations of each element in material as equivalent as second feature vectors, the pooling operation as equivalent as the aggregation function, and determining the fixed-length representation as equivalent as calculating the composite feature vector. Thus, Goodall teaches the limitation.
Referring to Claim 4, Goodall- Sandler teaches the information processing system of claim 1, Goodall also teaches:
wherein the first machine learning model is a machine learning model which generates the first feature vector that is a fixed-length vector from the numerical representation that is unstructured data. See Goodall at [Page 2, mid-left]:” We map these initial feature vectors onto a surrogate embedding by using a trainable weight matrix. The dimensionality of the surrogate embedding, d, determines the size of the hidden states for the message passing.” And, see Goodall at [Page 3, top-left]:” In this work, we adopt the same architecture for all the problems investigated. We set the surrogate embedding dimension to 64, and have 3 message passing layers each with 3 attention heads.” Goodall teaches that the first feature vector is a fixed-length vector because the elemental representation hi is a hidden-state vector whose size is determined by the surrogate embedding dimension d, and Goodall expressly sets the surrogate embedding dimension to 64. Then, the message-passing neural network generates updated elemental feature vectors hit+1having a fixed length of 64 dimensions from the input elemental/numerical representations. Also, see Goodall at [Page 1, bottom-right]:” Our approach is inspired by breakthrough machine learning methods in chemistry that directly take a molecular graph as input and infer the optimal molecule-to-descriptor map from data [18, 19].” Examiner interprets the input such as a molecular graph, as equivalent to the unstructured data. Thus, Goodall teaches the limitation.
Referring to Claim 5, Goodall - Sandler teaches the information processing system of claim 1, Goodall also teaches:
wherein the application of the plurality of combination ratios in association with the machine learning model comprises applying the plurality of combination ratios to output data of an intermediate layer of the machine learning model. See Goodall at [Page 2, mid-right]:” The first stage of the attention mechanism is to compute unnormalised coefficients, eij , across pairs of elements in the material.
PNG
media_image2.png
46
304
media_image2.png
Greyscale
where f (…) is a single hidden layer neural network, the j index runs over all the elements in vi, and || is the concatenation operation. We then normalise these coefficients using a weighted softmax function where the weights, wj , are the fraction of the neighbouring element in the total material.
PNG
media_image3.png
62
348
media_image3.png
Greyscale
The elemental representations are then updated with new pair-dependent perturbations that are weighted by the soft-attention coefficients.
PNG
media_image6.png
70
372
media_image6.png
Greyscale
where gn(…) are again single hidden layer neural networks.” Goodall computes intermediate attention coefficients eij in equation (2), and then applies weighted fraction wj corresponding to fractions of neighboring elements in Equation (3) to generate normalized attention coefficients aij. Then, the combination ratios are applied to intermediate outputs of the machine learning model prior to updating the feature vectors in Equation (4). Examiner interprets equation (3) applying the plurality of combination ratios to output data of an intermediate layer of the model.
Referring to Claim 6, Goodall-Sandler teaches the information processing system of claim 1, Goodall also teaches:
input the composite feature vector into another machine learning model to calculate a predicted value of characteristics of the composite object; and output the predicted value. See Goodall at [Page 3, top-left]:” These material representations are then taken as the input to an output neural network that computes the model predictions such that the whole model is end-to-end differentiable.” Examiner interprets inputting the material representations into an output neural network as equivalent as inputting composite feature vector into another model to calculate a predicated value. Also, the network computes the predicated value as an output.
Referring to claim 9 and claim 10, the claim is rejected on the same basis as claim 1, mutatis mutandis, since they are analogous claims.
Claim 7 and claim 8 are rejected under 35 U.S.C. 103 as being unpatentable over Goodall -Sandler in view of Helms (WIPO: WO2019099944A1, 05/23/2019, by Helms et al - hereinafter Helms).
Referring to Claim 7, Goodall - Sandler teaches the system of claim 1, it fails to teach,
the component object is a material, and the composite object is a multi- component substance.
Helms teaches, in an analogous system:
the component object is a material, and the composite object is a multi- component substance. See Helms at [0060]:” The present invention provides for a polymer alloy comprising a mixture or two or more polymers; wherein one or more of said polymers is the composition of the present invention; wherein at least of the two or more polymers optionally comprises one or more of the following: polyurethane, polyurea, epoxy, phenolic resin, polyolefin, silicone, rubber, polyacrylate, polymethacrylate, polycyanoacrylate, polyester, polycarbonate, polyimide, polyamide, vitrimer, poly(vinylogous amide), poly(vinylogous urethane), and/or thermoplastic elastomers.” Examiner interprets the polymer as equivalent as the material, and one or more polymers is the composition of the present invention as equivalent as multi-component substance as claimed.
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 Goodall - Sandler with the above teachings of Helms by information processing system to calculate the composite feature vector, as taught by Goodall and Sandler, the component object is a material, and the composite object is a multi- component substance, as taught by Helms. The modification would have been obvious because one of ordinary skill in the art would be motivated to provide an extreme precision characterization of material formulations, which functions as a high-quality, fine-grained training dataset, See Helms at [0053]:” Where a range of values is provided, it is understood that each intervening value, to the tenth of the unit of the lower limit unless the context clearly dictates otherwise, between the upper and lower limits of that range is also specifically disclosed. Each smaller range between any stated value or intervening value in a stated range and any other stated or intervening value in that stated range is encompassed within the invention. The upper and lower limits of these smaller ranges may independently be included or excluded in the range, and each range where either, neither or both limits are included in the smaller ranges is also encompassed within the invention, subject to any specifically excluded limit in the stated range. Where the stated range includes one or both of the limits, ranges excluding either or both of those included limits are also included in the invention.”
Referring to Claim 8, Goodall - Sandler teaches the system of claim 1, it fails to teach: the material is a polymer, and the multi-component substance is a polymer alloy.
Helms teaches, in an analogous system:
the material is a polymer, and the multi-component substance is a polymer alloy. See Helms at [0060]:” The present invention provides for a polymer alloy comprising a mixture or two or more polymers; wherein one or more of said polymers is the composition of the present invention; wherein at least of the two or more polymers optionally comprises one or more of the following: polyurethane, polyurea, epoxy, phenolic resin, polyolefin, silicone, rubber, polyacrylate, polymethacrylate, polycyanoacrylate, polyester, polycarbonate, polyimide, polyamide, vitrimer, poly(vinylogous amide), poly(vinylogous urethane), and/or thermoplastic elastomers.” Examiner interprets one or more polymers is the composition of the invention as equivalent as material is a polymer, and polymers optionally comprises polyurethane, polyurea, epoxy…as equivalent as multi-component substance is a polymer alloy.
The same motivation that was utilized for combining Goodall-Sandler with Helms as set forth in claim 7 is equally applicable to claim 8.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JIAYUE MA whose telephone number is (571)272-9658. The examiner can normally be reached between 9 am to 5 pm.
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.
/Jiayue Ma/
Examiner, Art Unit 2126
/DAVID YI/Supervisory Patent Examiner, Art Unit 2126
Prosecution Insights