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 . This action is responsive to the application filed on 01/31/2024. Claims 1-20 are presented in the case. Claims 1, 19 and 20 are independent claims.
Priority
Applicant's claim for the benefit of a national stage application under 35 U.S.C. § 371 of International Application No. PCT/CN2022/135337, filed November 30, 2022 is acknowledged.
Information Disclosure Statement
The information disclosure statement submitted on 07/16/2024 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Specification
The title of the invention is not descriptive. A new title is required that is clearly indicative of the invention to which the claims are directed. The following title is suggested: Computer-implemented method, apparatus, computer-program product for embedding learning process and predicting association relationships between nodes.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
Claim 17 is rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor, or for pre-AIA the applicant regards as the invention.
Claim 17 recites the limitation "the heterogeneous network". There is insufficient antecedent basis for this limitation in the claim.
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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: Claims 1-18 are directed to a method, claim 19 is directed to an apparatus and claim 20 is directed to a computer-program product. Therefore, the claims are eligible under Step 1 for being directed to a process, a machine and a manufacture respectively.
Independent claims 1, 19 and 20:
Step 2A Prong 1:
Claims recite:
obtaining a bipartite network; a first multi-view homogeneous network; and a second multi-view homogeneous network - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to obtain a bipartite network; a first multi-view homogeneous network; and a second multi-view homogeneous network;
performing an embedding learning process for the bipartite network - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to performing an embedding learning process;
performing an embedding learning process for the first multi-view homogeneous network - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to performing an embedding learning process;
performing an embedding learning process for the second multi-view homogeneous network - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to performing an embedding learning process; and
predicting association relationships between nodes of a first object type and nodes of a second object type - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to predicting association relationships between nodes.
Step 2A Prong 2: This judicial exception is not integrated into a practical application because they recite the additional elements:
An apparatus, comprising: a memory; one or more processors; wherein the memory and the one or more processors are connected with each other; and
the memory stores computer-executable instructions for controlling the one or more processors; A computer-program product, comprising a non-transitory tangible computer-readable medium having computer-readable instructions thereon, the computer-readable instructions being executable by a processor to cause the processor to perform - These limitations amount to components of a general purpose computer that applies a judicial exception, by use of conventional computer functions (see MPEP § 2106.05(b)).
Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are thus directed to the abstract idea.
Step 2B: The claims do not include additional elements that amount to significantly more than the judicial exception.
The additional elements:
An apparatus, comprising: a memory; one or more processors; wherein the memory and the one or more processors are connected with each other; and
the memory stores computer-executable instructions for controlling the one or more processors; A computer-program product, comprising a non-transitory tangible computer-readable medium having computer-readable instructions thereon, the computer-readable instructions being executable by a processor to cause the processor to perform - These limitations amount to components of a general purpose computer that applies a judicial exception, by use of conventional computer functions (see MPEP § 2106.05(b)).
Accordingly, these additional elements do not amount to significantly more than the judicial exception. As such, the claims are ineligible.
Dependent claim 2:
Step 2A Prong 1: The claim recites the abstract ideas of claim 1.
Step 2A Prong 2: This judicial exception is not integrated into a practical application because they recite the additional elements:
wherein the bipartite network comprises nodes of a first object type, nodes of a second object type, and edges connecting the nodes of the first object type and the nodes of the second object type - the step recited at a high level of generality, and amounts to selecting a particular data source or type of data to be manipulated, which is a form of insignificant extra-solution activity (see MPEP § 2106.05(g)).
Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are thus directed to the abstract idea.
Step 2B: The claims do not include additional elements that amount to significantly more than the judicial exception.
The additional elements:
wherein the bipartite network comprises nodes of a first object type, nodes of a second object type, and edges connecting the nodes of the first object type and the nodes of the second object type - viewed individually or in combination, describes selecting a particular data source or type of data to be manipulated similar to selecting information, based on types of information and availability of information in a power-grid environment, for collection, analysis and display described in MPEP § 2106.05(g).
Accordingly, these additional elements do not amount to significantly more than the judicial exception. As such, the claims are ineligible.
Dependent claim 3:
Step 2A Prong 1: The claim recites the abstract ideas of claim 1.
Step 2A Prong 2: This judicial exception is not integrated into a practical application because they recite the additional elements:
wherein the first multi-view homogeneous network comprises a first set of networks composed of nodes of a first object type, in which edges between a same pair of the nodes of the first type are observed in different views - the step recited at a high level of generality, and amounts to selecting a particular data source or type of data to be manipulated, which is a form of insignificant extra-solution activity (see MPEP § 2106.05(g)).
the second multi-view homogeneous network comprises a second set of networks composed of the nodes of the second object type, in which edges between a same pair of nodes of the second type are observed in different views - the step recited at a high level of generality, and amounts to selecting a particular data source or type of data to be manipulated, which is a form of insignificant extra-solution activity (see MPEP § 2106.05(g)).
Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are thus directed to the abstract idea.
Step 2B: The claims do not include additional elements that amount to significantly more than the judicial exception.
The additional elements:
wherein the first multi-view homogeneous network comprises a first set of networks composed of nodes of a first object type, in which edges between a same pair of the nodes of the first type are observed in different views - viewed individually or in combination, describes selecting a particular data source or type of data to be manipulated similar to selecting information, based on types of information and availability of information in a power-grid environment, for collection, analysis and display described in MPEP § 2106.05(g).
the second multi-view homogeneous network comprises a second set of networks composed of the nodes of the second object type, in which edges between a same pair of nodes of the second type are observed in different views - viewed individually or in combination, describes selecting a particular data source or type of data to be manipulated similar to selecting information, based on types of information and availability of information in a power-grid environment, for collection, analysis and display described in MPEP § 2106.05(g).
Accordingly, these additional elements do not amount to significantly more than the judicial exception. As such, the claims are ineligible.
Dependent claim 4:
Step 2A Prong 1:
Claim recites:
obtaining an embedding matrix representing the nodes of the first object type learned in the embedding learning process for the bipartite network - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to obtain an embedding matrix;
obtaining an embedding matrix representing the nodes of the second object type learned in the embedding learning process for the bipartite network - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to obtain an embedding matrix;
obtaining an embedding matrix representing the nodes of the first object type learned in the embedding learning process for the first multi-view homogeneous network - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to obtain an embedding matrix; and
obtaining an embedding matrix representing the nodes of the second object type learned in the embedding learning process for the second multi-view homogeneous network - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to obtain an embedding matrix.
Step 2A Prong 2 & Step 2B: There are no additional elements recited so the claims do not provide a practical application and is not considered to be significantly more. As such, the claims are ineligible.
Dependent claim 5:
Step 2A Prong 1:
Claim recites:
combining an embedding matrix representing the nodes of the first object type learned in the embedding learning process for the bipartite network and an embedding matrix representing the nodes of the first object type learned in the embedding learning process for the first multi-view homogeneous network - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to combine two embedding matrices;
combining an embedding matrix representing the nodes of the second object type learned in the embedding learning process for the bipartite network and an embedding matrix representing the nodes of the second object type learned in the embedding learning process for the second multi-view homogeneous network - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to combine two embedding matrices.
Step 2A Prong 2 & Step 2B: There are no additional elements recited so the claims do not provide a practical application and is not considered to be significantly more. As such, the claims are ineligible.
Dependent claim 6:
Step 2A Prong 1: The claim recites the abstract ideas of claim 1.
Step 2A Prong 2: This judicial exception is not integrated into a practical application because they recite the additional elements:
inputting an embedding matrix representing nodes of a first object type learned in the embedding learning process for the bipartite network into the first multi-view homogeneous network as an initialized node embedding of the first multi-view homogeneous network - the step recited at a high level of generality amounts to mere data gathering which is a form of insignificant extra-solution activity (see MPEP § 2106.05(g)); and
inputting an embedding matrix representing nodes of a second object type learned in the embedding learning process for the bipartite network into the second multi-view homogeneous network as an initialized node embedding of the second multi-view homogeneous network - the step recited at a high level of generality amounts to mere data gathering which is a form of insignificant extra-solution activity (see MPEP § 2106.05(g)).
Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are thus directed to the abstract idea.
Step 2B: The claims do not include additional elements that amount to significantly more than the judicial exception.
The additional elements:
inputting an embedding matrix representing nodes of a first object type learned in the embedding learning process for the bipartite network into the first multi-view homogeneous network as an initialized node embedding of the first multi-view homogeneous network - viewed individually or in combination, describes mere data gathering similar to Presenting offers to potential customers and gathering statistics generated based on the testing about how potential customers responded to the offers; the statistics are then used to calculate an optimized price, OIP Technologies, 788 F.3d at 1363, 115 USPQ2d at 1092-93 described in MPEP § 2106.05(g).; and
inputting an embedding matrix representing nodes of a second object type learned in the embedding learning process for the bipartite network into the second multi-view homogeneous network as an initialized node embedding of the second multi-view homogeneous network - viewed individually or in combination, describes mere data gathering similar to Presenting offers to potential customers and gathering statistics generated based on the testing about how potential customers responded to the offers; the statistics are then used to calculate an optimized price, OIP Technologies, 788 F.3d at 1363, 115 USPQ2d at 1092-93 described in MPEP § 2106.05(g).
Accordingly, these additional elements do not amount to significantly more than the judicial exception. As such, the claims are ineligible.
Dependent claim 7:
Step 2A Prong 1:
Claim recites:
determining weights assigned to M homogeneous views respectively by the attention mechanism - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to determine weights by the attention mechanism.
Step 2A Prong 2: This judicial exception is not integrated into a practical application because they recite the additional elements:
inputting M sets of node embeddings of the first object type learned from M homogeneous view networks in the embedding learning process for the first multi-view homogeneous network into an attention mechanism - the step recited at a high level of generality amounts to mere data gathering which is a form of insignificant extra-solution activity (see MPEP § 2106.05(g)).
Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are thus directed to the abstract idea.
Step 2B: The claims do not include additional elements that amount to significantly more than the judicial exception.
The additional elements:
inputting M sets of node embeddings of the first object type learned from M homogeneous view networks in the embedding learning process for the first multi-view homogeneous network into an attention mechanism - viewed individually or in combination, describes mere data gathering similar to Presenting offers to potential customers and gathering statistics generated based on the testing about how potential customers responded to the offers; the statistics are then used to calculate an optimized price, OIP Technologies, 788 F.3d at 1363, 115 USPQ2d at 1092-93 described in MPEP § 2106.05(g).
Accordingly, these additional elements do not amount to significantly more than the judicial exception. As such, the claims are ineligible.
Dependent claim 8:
Step 2A Prong 1:
Claim recites:
wherein the weights assigned to M homogeneous views respectively are expressed as:
PNG
media_image1.png
400
664
media_image1.png
Greyscale
- Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to assign weights.
Step 2A Prong 2 & Step 2B: There are no additional elements recited so the claims do not provide a practical application and is not considered to be significantly more. As such, the claims are ineligible.
Dependent claim 9:
Step 2A Prong 1:
Claim recites:
wherein determining the weights assigned to the M homogeneous views respectively comprises performing a non-linear transformation on H1M l, to transform H1M l into embeddings of all nodes in a m-th view network of the M homogeneous view networks - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to determine the weights.
Step 2A Prong 2 & Step 2B: There are no additional elements recited so the claims do not provide a practical application and is not considered to be significantly more. As such, the claims are ineligible.
Dependent claim 10:
Step 2A Prong 1:
Claim recites:
fusing, by a l-th layer of the attention mechanism, different low dimensional feature representations of nodes of the first object type under different meta-paths, using the weights assigned to M homogeneous views respectively; and
obtaining a low dimensional embedding representation of the nodes of the first object type from a l-th layer of the attention mechanism:
PNG
media_image2.png
58
188
media_image2.png
Greyscale
- Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to determine the weights.
Step 2A Prong 2 & Step 2B: There are no additional elements recited so the claims do not provide a practical application and is not considered to be significantly more. As such, the claims are ineligible.
Dependent claim 11:
Step 2A Prong 1:
Claim recites:
determining weights assigned to N homogeneous views respectively by the attention mechanism - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to determine weights by the attention mechanism.
Step 2A Prong 2: This judicial exception is not integrated into a practical application because they recite the additional elements:
inputting N sets of node embeddings of the second object type learned from N homogeneous view networks in the embedding learning process for the second multi-view homogeneous network into an attention mechanism - the step recited at a high level of generality amounts to mere data gathering which is a form of insignificant extra-solution activity (see MPEP § 2106.05(g)).
Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are thus directed to the abstract idea.
Step 2B: The claims do not include additional elements that amount to significantly more than the judicial exception.
The additional elements:
inputting N sets of node embeddings of the second object type learned from N homogeneous view networks in the embedding learning process for the second multi-view homogeneous network into an attention mechanism - viewed individually or in combination, describes mere data gathering similar to Presenting offers to potential customers and gathering statistics generated based on the testing about how potential customers responded to the offers; the statistics are then used to calculate an optimized price, OIP Technologies, 788 F.3d at 1363, 115 USPQ2d at 1092-93 described in MPEP § 2106.05(g).
Accordingly, these additional elements do not amount to significantly more than the judicial exception. As such, the claims are ineligible.
Dependent claim 12:
Step 2A Prong 1:
Claim recites:
wherein the weights assigned to N homogeneous views respectively are expressed as:
PNG
media_image3.png
400
664
media_image3.png
Greyscale
- Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to assign weights.
Step 2A Prong 2 & Step 2B: There are no additional elements recited so the claims do not provide a practical application and is not considered to be significantly more. As such, the claims are ineligible.
Dependent claim 13:
Step 2A Prong 1:
Claim recites:
wherein determining the weights assigned to the N homogeneous views respectively comprises performing a non-linear transformation on H2n l, to transform H2n l into embeddings of all nodes in a n-th view network of the N homogeneous view networks - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to determine the weights.
Step 2A Prong 2 & Step 2B: There are no additional elements recited so the claims do not provide a practical application and is not considered to be significantly more. As such, the claims are ineligible.
Dependent claim 14:
Step 2A Prong 1:
Claim recites:
fusing, by a l-th layer of the attention mechanism, different low dimensional feature representations of nodes of the second object type under different meta-paths, using the weights assigned to N homogeneous views respectively; and
obtaining a low dimensional embedding representation of the nodes of the second object type from a l-th layer of the attention mechanism:
PNG
media_image4.png
68
212
media_image4.png
Greyscale
- Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to determine the weights.
Step 2A Prong 2 & Step 2B: There are no additional elements recited so the claims do not provide a practical application and is not considered to be significantly more. As such, the claims are ineligible.
Dependent 15:
Step 2A Prong 1:
Claims recite:
extracting node representations of nodes of the first object type and nodes of the second object type by: learning sequence representations of nodes - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to extract node representations by learning sequence representations of nodes.
Step 2A Prong 2: This judicial exception is not integrated into a practical application because they recite the additional elements:
inputting pairs of nodes - the step recited at a high level of generality, and amounts to mere data inputting, which is a form of insignificant extra-solution activity (see MPEP § 2106.05(g)).
outputting vector representations of nodes - the step recited at a high level of generality, and amounts to mere data outputting, which is a form of insignificant extra-solution activity (see MPEP § 2106.05(g)).
Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are thus directed to the abstract idea.
Step 2B: The claims do not include additional elements that amount to significantly more than the judicial exception.
The additional elements:
inputting pairs of nodes - the step recited at a high level of generality, and amounts to mere data inputting, which is a form of insignificant extra-solution activity (see MPEP § 2106.05(g)).
outputting vector representations of nodes - the step recited at a high level of generality, and amounts to mere data outputting, which is a form of insignificant extra-solution activity (see MPEP § 2106.05(g)).
Accordingly, these additional elements do not amount to significantly more than the judicial exception. As such, the claims are ineligible.
Dependent 16:
Step 2A Prong 1:
Claims recite:
wherein learning sequence representations of nodes is performed using a random walk algorithm; and the vector representations of the nodes are obtained using a skip-gram model - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to learn sequence representations of nodes and obtain the vector representations of the nodes.
Step 2A Prong 2 & Step 2B: There are no additional elements recited so the claims do not provide a practical application and is not considered to be significantly more. As such, the claims are ineligible.
Dependent claim 17:
Step 2A Prong 1:
Claims recite:
wherein the method further comprises a decoding process to reconstruct association relationships between the nodes of a first object type and the nodes of a second object type, thereby predicting drug-disease association - Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to reconstruct association relationships between the nodes.
Step 2A Prong 2: This judicial exception is not integrated into a practical application because they recite the additional elements:
wherein the nodes of a first object type are drug nodes; the nodes of the second object type are disease nodes; wherein the heterogeneous network comprises one or more drug-drug similarity matrixes and one or more disease-disease matrixes - the step recited at a high level of generality, and amounts to selecting a particular data source or type of data to be manipulated, which is a form of insignificant extra-solution activity (see MPEP § 2106.05(g)).
Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are thus directed to the abstract idea.
Step 2B: The claims do not include additional elements that amount to significantly more than the judicial exception.
The additional elements:
wherein the nodes of a first object type are drug nodes; the nodes of the second object type are disease nodes; wherein the heterogeneous network comprises one or more drug-drug similarity matrixes and one or more disease-disease matrixes - viewed individually or in combination, describes selecting a particular data source or type of data to be manipulated similar to selecting information, based on types of information and availability of information in a power-grid environment, for collection, analysis and display described in MPEP § 2106.05(g).
Accordingly, these additional elements do not amount to significantly more than the judicial exception. As such, the claims are ineligible.
Dependent 18:
Step 2A Prong 1:
Claims recite:
further comprising minimizing a weighted binary cross-entropy loss:
PNG
media_image5.png
442
668
media_image5.png
Greyscale
- Under its broadest reasonable interpretation in light of the specification, this limitation encompasses a mathematical concept of a mathematical calculation to minimize a weighted binary cross-entropy loss.
Step 2A Prong 2 & Step 2B: There are no additional elements recited so the claims do not provide a practical application and is not considered to be significantly more. As such, the claims are ineligible.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1-6 and 19-20 are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Fu et al. (hereinafter Fu), “MVGCN: data integration through multi-view graph convolutional network for predicting links in biomedical bipartite networks”, Bioinformatics, Volume 38, Issue 2, January 2022, Pages 426-434, https://doi.org/10.1093/bioinformatics/btab651.
Regarding independent claim 1, Fu teaches a computer-implemented method, comprising:
obtaining a bipartite network; a first multi-view homogeneous network; and a second multi-view homogeneous network (page 427, section “2.1 Construction of multi-view heterogeneous network” discloses constructing a multi view heterogeneous network (MVHN), which consists of a biomedical bipartite network and two similarity networks);
performing an embedding learning process for the bipartite network (pages 427-428, Fig. 1, section “2.1 Construction of multi-view heterogeneous network” discloses an embedding learning process for a biomedical bipartite network comprising two disjoint sets of nodes indicating two domains of bioentities and E denotes the set of edges representing associations or interactions between nodes in A and nodes in B and a binary matrix with regard to the bi partite graph; pages 427-428, Fig. 1, section “2.2 Node attribute pre-training by self-supervised learning” discloses learning node attributes, which will be used as the initial embeddings for the downstream link prediction tasks, constructing the adjacency matrix of the bipartite network);
performing an embedding learning process for the first multi-view homogeneous network (pages 427-428, Fig. 1, section “2.1 Construction of multi-view heterogeneous network” discloses calculating bioentity similarities from features where two tensors denote two domains of multi-view similarities and each slice in one of the tensors is a similarity matrix that is derived from one feature and normalized as stochastic matrix, where each element describes similarity between two bioentities in one domain; page 428, Fig. 1, section “2.3.1 Neighbor information aggregation (NIA)” discloses Graph convolutional network (KipfandWelling,2017) is a multi layer connected neural network architecture, which takes graphs as input and outputs low-dimensional compact representations of nodes. It aggregates neighbor information by repetitively updating the embeddings of nodes through the weights of edges in graphs … Inspired by this basic idea, we define the neighbor information aggregation (NIA) layer to update node embeddings in each view of the MVHN … we first define inter-and intra-domain neighbors for a target node. The inter-domain neighbors are nodes linked to the target node in the bipartite network, while intra-domain neighbors are nodes linked to the target node in similarity networks. Clearly, inter-domain neighbors and the target node are from different domains, and intra-domain neighbors and the target node are from the same domain);
performing an embedding learning process for the second multi-view homogeneous network (pages 427-428, Fig. 1, section “2.1 Construction of multi-view heterogeneous network” discloses calculating bioentity similarities from features where two tensors denote two domains of multi-view similarities and each slice in one of the tensors is a similarity matrix that is derived from one feature and normalized as stochastic matrix, where each element describes similarity between two bioentities in one domain; page 428, Fig. 1, section “2.3.1 Neighbor information aggregation (NIA)” discloses Graph convolutional network (KipfandWelling,2017) is a multi layer connected neural network architecture, which takes graphs as input and outputs low-dimensional compact representations of nodes. It aggregates neighbor information by repetitively updating the embeddings of nodes through the weights of edges in graphs … Inspired by this basic idea, we define the neighbor information aggregation (NIA) layer to update node embeddings in each view of the MVHN … we first define inter-and intra-domain neighbors for a target node. The inter-domain neighbors are nodes linked to the target node in the bipartite network, while intra-domain neighbors are nodes linked to the target node in similarity networks. Clearly, inter-domain neighbors and the target node are from different domains, and intra-domain neighbors and the target node are from the same domain. The NIA layer aggregates information from inter- and intra domain neighbors by integrating two parts: inter-domain message passing (inter-DMP) and intra-domain message passing (intra DMP).Taking one specific view of the MVHN as an example, we denote the embeddings of the node ai and the node bj at lth layer as hðlÞ ai and hðlÞ bj ,respectively); and
predicting association relationships between nodes of a first object type and nodes of a second object type (page 428, Fig. 1, section “2.3 Multi-view graph convolutional network” discloses as shown in Figure1, we will present MVGCN in a modular style: starting with the core design of the GNN propagation layer, followed by mechanisms for integrating information from layers and views, and ending with the exposition of a discriminator and the objective function for link prediction … Fig.1. The overall architecture of MVGCN for predicting links in biomedical bipartite networks. First, MVGCN propagates inter-and intra-domain neighbor information over the heterogeneous network to update the hidden states of nodes. Then, MVGCN combines all NIA layers in each view and integrates multi-view information to obtain the final embeddings of nodes. Finally, the feature of each node pair, emanated from the concatenation of the final embeddings of the corresponding two nodes, is mapped into a probability via a discriminator for the eventual link prediction task; page 429, Fig. 1, section “2.3.3 Model training” discloses For the nodes ai and bi, we concatenate hai and hbj to characterize the node pair, then feed the concatenation into a multi-layer perceptron(MLP) to obtain the predictive score).
Regarding dependent claim 2, Fu teaches all the limitations as set forth in the rejection of claim 1 that is incorporated. Fu teaches wherein the bipartite network comprises nodes of a first object type, nodes of a second object type, and edges connecting the nodes of the first object type and the nodes of the second object type (page 426, section “1 Introduction” discloses In network medicine, there are multifarious biomedical bipartite networks, such as drug–disease association network, disease–gene association network and drug–target interaction network, each of which consists of two distinct types of nodes from two disjoint domains and edges/links between nodes of different domains representing the relationships; pages 427-428, Fig. 1, section “2.1 Construction of multi-view heterogeneous network”
PNG
media_image6.png
146
828
media_image6.png
Greyscale
).
Regarding dependent claim 3, Fu teaches all the limitations as set forth in the rejection of claim 1 that is incorporated. Fu teaches wherein the first multi-view homogeneous network comprises a first set of networks composed of nodes of a first object type, in which edges between a same pair of the nodes of the first type are observed in different views (pages 427-428, Fig. 1, section “2.1 Construction of multi-view heterogeneous network”
PNG
media_image7.png
324
832
media_image7.png
Greyscale
PNG
media_image8.png
198
842
media_image8.png
Greyscale
); and
the second multi-view homogeneous network comprises a second set of networks composed of the nodes of the second object type, in which edges between a same pair of nodes of the second type are observed in different views (pages 427-428, Fig. 1, section “2.1 Construction of multi-view heterogeneous network”
PNG
media_image7.png
324
832
media_image7.png
Greyscale
PNG
media_image8.png
198
842
media_image8.png
Greyscale
).
Regarding dependent claim 4, Fu teaches all the limitations as set forth in the rejection of claim 1 that is incorporated. Fu teaches further comprising:
obtaining an embedding matrix representing the nodes of the first object type learned in the embedding learning process for the bipartite network (pages 427-428, Fig. 1, section “2.1 Construction of multi-view heterogeneous network”
PNG
media_image9.png
136
814
media_image9.png
Greyscale
);
obtaining an embedding matrix representing the nodes of the second object type learned in the embedding learning process for the bipartite network (pages 427-428, Fig. 1, section “2.1 Construction of multi-view heterogeneous network”
PNG
media_image9.png
136
814
media_image9.png
Greyscale
);
obtaining an embedding matrix representing the nodes of the first object type learned in the embedding learning process for the first multi-view homogeneous network (pages 427-428, Fig. 1, section “2.1 Construction of multi-view heterogeneous network”
PNG
media_image10.png
282
864
media_image10.png
Greyscale
); and
obtaining an embedding matrix representing the nodes of the second object type learned in the embedding learning process for the second multi-view homogeneous network (pages 427-428, Fig. 1, section “2.1 Construction of multi-view heterogeneous network”
PNG
media_image10.png
282
864
media_image10.png
Greyscale
).
Regarding dependent claim 5, Fu teaches all the limitations as set forth in the rejection of claim 2 that is incorporated. Fu teaches further comprising:
combining an embedding matrix representing the nodes of the first object type learned in the embedding learning process for the bipartite network and an embedding matrix representing the nodes of the first object type learned in the embedding learning process for the first multi-view homogeneous network (page 428, Fig. 1, section “2.3.1 Neighbor information aggregation (NIA)” we first define inter- and intra-domain neighbors for a target node. The inter-domain neighbors are nodes linked to the target node in the bipartite network, while intra-domain neighbors are nodes linked to the target node in similarity networks. Clearly, inter-domain neighbors and the target node are from different domains, and intra-domain neighbors and the target node are from the same domain. The NIA layer aggregates information from inter- and intra-domain neighbors by integrating two parts: inter-domain message passing (inter-DMP) and intra-domain message passing (intra-DMP); page 429, Fig. 1, section “2.3.3 Model training”
PNG
media_image11.png
182
872
media_image11.png
Greyscale
); and
combining an embedding matrix representing the nodes of the second object type learned in the embedding learning process for the bipartite network and an embedding matrix representing the nodes of the second object type learned in the embedding learning process for the second multi-view homogeneous network (page 428, Fig. 1, section “2.3.1 Neighbor information aggregation (NIA)” we first define inter- and intra-domain neighbors for a target node. The inter-domain neighbors are nodes linked to the target node in the bipartite network, while intra-domain neighbors are nodes linked to the target node in similarity networks. Clearly, inter-domain neighbors and the target node are from different domains, and intra-domain neighbors and the target node are from the same domain. The NIA layer aggregates information from inter- and intra-domain neighbors by integrating two parts: inter-domain message passing (inter-DMP) and intra-domain message passing (intra-DMP); page 429, Fig. 1, section “2.3.3 Model training”
PNG
media_image11.png
182
872
media_image11.png
Greyscale
).
Regarding dependent claim 6, Fu teaches all the limitations as set forth in the rejection of claim 1 that is incorporated. Fu teaches further comprising:
inputting an embedding matrix representing nodes of a first object type learned in the embedding learning process for the bipartite network into the first multi-view homogeneous network as an initialized node embedding of the first multi-view homogeneous network (page 427, right column; we develop a novel link prediction model, multi-view graph convolutional network (abbreviated as MVGCN), which can adaptively incorporate multiple features of bioentities by multi-similarity fusion to infer potential links in multifarious biomedical bipartite networks. We first construct a multi-view heterogeneous network (MVHN) by combining the similarity networks with the bipartite network, and then perform a self-supervised learning strategy to obtain node attributes as initial embeddings; page 428, Fig. 1, section “2.3.1 Neighbor information aggregation (NIA)”
PNG
media_image12.png
230
484
media_image12.png
Greyscale
); and
inputting an embedding matrix representing nodes of a second object type learned in the embedding learning process for the bipartite network into the second multi-view homogeneous network as an initialized node embedding of the second multi-view homogeneous network (page 427, right column; we develop a novel link prediction model, multi-view graph convolutional network (abbreviated as MVGCN), which can adaptively incorporate multiple features of bioentities by multi-similarity fusion to infer potential links in multifarious biomedical bipartite networks. We first construct a multi-view heterogeneous network (MVHN) by combining the similarity networks with the bipartite network, and then perform a self-supervised learning strategy to obtain node attributes as initial embeddings; page 428, Fig. 1, section “2.3.1 Neighbor information aggregation (NIA)”
PNG
media_image12.png
230
484
media_image12.png
Greyscale
).
Regarding independent claim 19, it is an apparatus claim that corresponding to the method of claim 1. Therefore, it is rejected for the same reason as claim 1 above.
Regarding independent claim 20, it is computer-program product claim that corresponding to the method of claim 1. Therefore, it is rejected for the same reason as claim 1 above.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 7-14 are rejected under 35 U.S.C. 103 as being unpatentable over Fu as applied in claim 1, in view of Zhang et al. (hereinafter Zhang), “Multi-view Dynamic Heterogeneous Information Network Embedding”, The Computer Journal, Volume 65, Issue 8, August 2022, Pages 2016-2033, https://doi.org/10.1093/comjnl/bxab041.
Regarding dependent claim 7, Fu teaches all the limitations as set forth in the rejection of claim 1 that is incorporated. Fu does not explicitly teach wherein performing the embedding learning process for the first multi-view homogeneous network comprises:
inputting M sets of node embeddings of the first object type learned from M homogeneous view networks in the embedding learning process for the first multi-view homogeneous network into an attention mechanism; and
determining weights assigned to M homogeneous views respectively by the attention mechanism.
However, in the same field of endeavor, Zhang teaches wherein performing the embedding learning process for the first multi-view homogeneous network comprises:
inputting M sets of node embeddings of the first object type learned from M homogeneous view networks in the embedding learning process for the first multi-view homogeneous network into an attention mechanism (page 2024, section “5.2. Attention based multi-view fusion” After a group of node latent vectors from different views have been obtained, an efficient fusion mechanism is needed to integrate these latent vectors and further vote for the final node vectors. A high-dimension representation can be directly concatenated by all latent vectors. Alternatively, all implicit vectors are averaged (i.e. assigning the same weight to all latent vectors). Considering that different views make different contributions to the final network embedding, an attention mechanism is introduced in the MDHNE framework so that weights of latent vectors that are encoded in node proximities of different views can be automatically learned); and
determining weights assigned to M homogeneous views respectively by the attention mechanism (page 2024, section “5.2. Attention based multi-view fusion”
PNG
media_image13.png
252
684
media_image13.png
Greyscale
PNG
media_image14.png
224
664
media_image14.png
Greyscale
).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of utilizing the attention mechanism to automatically infer the weights of the learned latent vectors during fusion, as suggested in Zhang into Fu’s system because both of these systems are addressing multi-view dynamic heterogeneous information network embedding. This modification would have been motivated by the desire to promote the collaboration of different views, and also automatically infer the weights of views during integration. (Zhang, page 2017, right column).
Regarding dependent claim 8, the combination of Fu and Zhang teaches all the limitations as set forth in the rejection of claim 7 that is incorporated. Zhang teaches wherein the weights assigned to M homogeneous views respectively are expressed as:
(α1 l,α2 l, . . . αM l)=αtt sem(H 11 l ,H 12 l , . . . ,H 1M l),l=0,1,2, . . . ;
wherein (α1 l, α2 l, . . . αM l) stands for a weight matrix comprising the weights assigned to the M homogeneous views respectively;
αttsem stands for a method of performing semantic level attention;
(H11 l, H12 l, . . . , H1M l) stands for feature representations of nodes of the first object type extracted from the M homogeneous views; and
l stands for a l-th layer of the attention mechanism (page 2024, section “5.2. Attention based multi-view fusion”
PNG
media_image13.png
252
684
media_image13.png
Greyscale
PNG
media_image14.png
224
664
media_image14.png
Greyscale
).
Regarding dependent claim 9, the combination of Fu and Zhang teaches all the limitations as set forth in the rejection of claim 8 that is incorporated. Zhang teaches
wherein determining the weights assigned to the M homogeneous views respectively comprises performing a non-linear transformation on H1M l, to transform H1M l into embeddings of all nodes in a m-th view network of the M homogeneous view networks (page 2017, right column; We apply RNN architecture to capture highly complex and temporal features. After a series of non-linear functions in the recurrent layers of RNN, respectively, transformation patterns of structure and dynamically changed proximities can be embedded in latent vector space, and thus the node vectors of multiple views can be updated over time).
Regarding dependent claim 10, the combination of Fu and Zhang teaches all the limitations as set forth in the rejection of claim 7 that is incorporated. Zhang teaches further comprising:
fusing, by a l-th layer of the attention mechanism, different low dimensional feature representations of nodes of the first object type under different meta-paths, using the weights assigned to M homogeneous views respectively (page 2024, section “5.2. Attention based multi-view fusion” After a group of node latent vectors from different views have been obtained, an efficient fusion mechanism is needed to integrate these latent vectors and further vote for the final node vectors. A high-dimension representation can be directly concatenated by all latent vectors); and
obtaining a low dimensional embedding representation of the nodes of the first object type from a l-th layer of the attention mechanism:
H~1l=Σm=1Mαml·H1ml. (page 2024, right column;
PNG
media_image15.png
254
624
media_image15.png
Greyscale
).
Regarding dependent claim 11, Fu teaches all the limitations as set forth in the rejection of claim 1 that is incorporated. Fu does not explicitly teach
wherein performing the embedding learning process for the second multi-view homogeneous network comprises:
inputting N sets of node embeddings of the second object type learned from N homogeneous view networks in the embedding learning process for the second multi-view homogeneous network into an attention mechanism; and
determining weights assigned to N homogeneous views respectively by the attention mechanism.
However, in the same field of endeavor, Zhang teaches wherein performing the embedding learning process for the second multi-view homogeneous network comprises:
inputting N sets of node embeddings of the second object type learned from N homogeneous view networks in the embedding learning process for the second multi-view homogeneous network into an attention mechanism (page 2024, section “5.2. Attention based multi-view fusion” After a group of node latent vectors from different views have been obtained, an efficient fusion mechanism is needed to integrate these latent vectors and further vote for the final node vectors. A high-dimension representation can be directly concatenated by all latent vectors. Alternatively, all implicit vectors are averaged (i.e. assigning the same weight to all latent vectors). Considering that different views make different contributions to the final network embedding, an attention mechanism is introduced in the MDHNE framework so that weights of latent vectors that are encoded in node proximities of different views can be automatically learned); and
determining weights assigned to N homogeneous views respectively by the attention mechanism (page 2024, section “5.2. Attention based multi-view fusion”
PNG
media_image13.png
252
684
media_image13.png
Greyscale
PNG
media_image14.png
224
664
media_image14.png
Greyscale
).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of utilizing the attention mechanism to automatically infer the weights of the learned latent vectors during fusion, as suggested in Zhang into Fu’s system because both of these systems are addressing multi-view dynamic heterogeneous information network embedding. This modification would have been motivated by the desire to promote the collaboration of different views, and also automatically infer the weights of views during integration. (Zhang, page 2017, right column).
Regarding dependent claim 12, the combination of Fu and Zhang teaches all the limitations as set forth in the rejection of claim 11 that is incorporated. Zhang teaches wherein the weights assigned to N homogeneous views respectively are expressed as:
(β1 l,β2 l, . . . ,βN l)=αtt sem(H 21 l ,H 22 l , . . . ,H 2N l),l=0,1,2, . . . ;
wherein (β1 l, β2 l, . . . ,βN l) stands for a weight matrix comprising the weights assigned to the N homogeneous views respectively;
αttsem stands for a method of performing semantic level attention;
(H21 l,H22 l, . . . ,H2N l) stands for feature representations of nodes of the second object type extracted from the N homogeneous views; and
l stands for a l-th layer of the attention mechanism (page 2024, section “5.2. Attention based multi-view fusion”
PNG
media_image13.png
252
684
media_image13.png
Greyscale
PNG
media_image14.png
224
664
media_image14.png
Greyscale
).
Regarding dependent claim 13, the combination of Fu and Zhang teaches all the limitations as set forth in the rejection of claim 12 that is incorporated. Zhang teaches wherein determining the weights assigned to the N homogeneous views respectively comprises performing a non-linear transformation on H2n l, to transform H2n l into embeddings of all nodes in a n-th view network of the N homogeneous view networks (page 2017, right column; We apply RNN architecture to capture highly complex and temporal features. After a series of non-linear functions in the recurrent layers of RNN, respectively, transformation patterns of structure and dynamically changed proximities can be embedded in latent vector space, and thus the node vectors of multiple views can be updated over time).
Regarding dependent claim 14, the combination of Fu and Zhang teaches all the limitations as set forth in the rejection of claim 11 that is incorporated. Zhang teaches further comprising:
fusing, by a l-th layer of the attention mechanism, different low dimensional feature representations of nodes of the second object type under different meta-paths, using the weights assigned to N homogeneous views respectively (page 2024, section “5.2. Attention based multi-view fusion” After a group of node latent vectors from different views have been obtained, an efficient fusion mechanism is needed to integrate these latent vectors and further vote for the final node vectors. A high-dimension representation can be directly concatenated by all latent vectors); and
obtaining a low dimensional embedding representation of the nodes of the second object type from a l-th layer of the attention mechanism:
H~2l=Σn=1Nβnl·H2nl. (page 2024, right column;
PNG
media_image15.png
254
624
media_image15.png
Greyscale
).
Claims 15-16 are rejected under 35 U.S.C. 103 as being unpatentable over Fu as applied in claim 1, in view of Bhuiyan et al. (hereinafter Bhuiyan), US 20220261406 A1.
Regarding dependent claim 15, Fu teaches all the limitations as set forth in the rejection of claim 1 that is incorporated. Fu does not explicitly teach wherein performing the embedding learning process for the bipartite network comprises extracting node representations of nodes of the first object type and nodes of the second object type by:
inputting pairs of nodes;
learning sequence representations of nodes; and
outputting vector representations of nodes.
However, in the same field of endeavor, Bhuiyan teaches wherein performing the embedding learning process for the bipartite network comprises extracting node representations of nodes of the first object type and nodes of the second object type by:
inputting pairs of nodes ([0047] FIG. 4 illustrates an exemplary process 400 for the creation of similar intent groups used in the extended search of FIG. 3. The search catalog 401 contains the products/items/services for example offered by entity (e.g. retailer). The search logs 403 contain the queries previously used to search for products/items/services etc. in the search catalog 401. Using the search catalog 401 and the search logs 403 both of which are accessed with authorization, queries, items and intents (i.e. product types) are mined 405 and pairs are collected 407. Each pair represents a query linked to an intent via an engagement. Engagements may be search returns, views, clicks, add-to-cart, or purchases, etc. For example as a result of a query of Charcoal grills, the user clicks on one of the Costway outdoor BBQ, or adds the product to the cart for purchase, the query and the intent (product type) would be paired via the engagement. The extracted pairs are used to create a bipartite graph as shown in FIG. 5);
learning sequence representations of nodes ([0047] Graph embedding 413, a technique well known in the art, is performed which results in the bipartite graph being represented in vector form, examples of graph embedding as known in the art include Deep walk, Random walk, Word2vec, skip-gram, among others. The similar intents are grouped together 415 based on the vector distance between intents and the resultant groups are stored 416 to be used in the extended search); and
outputting vector representations of nodes ([0047] Graph embedding 413, a technique well known in the art, is performed which results in the bipartite graph being represented in vector form; [0049] Graph embedding methods are undertaken on the bipartite graph in which the graph is converted into a plurality of vectors representing the prior intents).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of a graph embedding method over the bipartite graph which results in the bipartite graph being represented in vector form as suggested in Bhuiyan into Fu’s system because both of these systems are addressing embedding learning process for the bipartite network. This modification would have been motivated by the desire to improve extracting node representation process by capturing the most relevant returns (Bhuiyan, [0016]).
Regarding dependent claim 16, the combination of Fu and Bhuiyan teaches all the limitations as set forth in the rejection of claim 15 that is incorporated. Bhuiyan teaches wherein learning sequence representations of nodes is performed using a random walk algorithm ([0047] Graph embedding 413, a technique well known in the art, is performed which results in the bipartite graph being represented in vector form, examples of graph embedding as known in the art include Deep walk, Random walk, Word2vec, skip-gram, among others); and
the vector representations of the nodes are obtained using a skip-gram model ([0047] Graph embedding 413, a technique well known in the art, is performed which results in the bipartite graph being represented in vector form, examples of graph embedding as known in the art include Deep walk, Random walk, Word2vec, skip-gram, among others).
Claim 17 is rejected under 35 U.S.C. 103 as being unpatentable over Fu as applied in claim 2, in view of Wang et al. (hereinafter Wang), “Drug-Disease Association Prediction Based on Neighborhood Information Aggregation in Neural Networks” in IEEE Access, vol. 7, pp. 50581-50587, 2019.
Regarding dependent claim 17, Fu teaches all the limitations as set forth in the rejection of claim 2 that is incorporated. Fu does not explicitly teach wherein the nodes of a first object type are drug nodes; the nodes of the second object type are disease nodes; wherein the heterogeneous network comprises one or more drug-drug similarity matrixes and one or more disease-disease matrixes; wherein the method further comprises a decoding process to reconstruct association relationships between the nodes of a first object type and the nodes of a second object type, thereby predicting drug-disease association.
However, in the same field of endeavor, Wang teaches wherein the nodes of a first object type are drug nodes (page 50583, section “B. SCHEMATIC OVERVIEW OF HNRD” a) Construct a heterogeneous network based on three standard matrices. The three matrices mainly include the drug similarity link matrix, the disease similarity adjacency matrix, and the correlation matrix of drugs and diseases. The similarity matrix is symmetrical, whereas the drug-disease correlation matrix is asymmetric and binary. Regularize the correlation matrix for each pair. b) Integrate neighborhood information for drugs and diseases, and embed low-dimensional space, each with a low-dimensional representation; page 50583, section “C. HETEROGENEOUS NETWORK” By using the normalized matrices as edges weight, a heterogeneous network is generated which contains two node types {drug, disease} and three edge types {drug-drug, disease-disease, drug-disease});
the nodes of the second object type are disease nodes (page 50583, section “B. SCHEMATIC OVERVIEW OF HNRD” a) Construct a heterogeneous network based on three standard matrices. The three matrices mainly include the drug similarity link matrix, the disease similarity adjacency matrix, and the correlation matrix of drugs and diseases. The similarity matrix is symmetrical, whereas the drug-disease correlation matrix is asymmetric and binary. Regularize the correlation matrix for each pair. b) Integrate neighborhood information for drugs and diseases, and embed low-dimensional space, each with a low-dimensional representation; page 50583, section “C. HETEROGENEOUS NETWORK” By using the normalized matrices as edges weight, a heterogeneous network is generated which contains two node types {drug, disease} and three edge types {drug-drug, disease-disease, drug-disease});
wherein the heterogeneous network comprises one or more drug-drug similarity matrixes and one or more disease-disease matrixes (page 50583, section “B. SCHEMATIC OVERVIEW OF HNRD” a) Construct a heterogeneous network based on three standard matrices. The three matrices mainly include the drug similarity link matrix, the disease similarity adjacency matrix, and the correlation matrix of drugs and diseases. The similarity matrix is symmetrical, whereas the drug-disease correlation matrix is asymmetric and binary. Regularize the correlation matrix for each pair. b) Integrate neighborhood information for drugs and diseases, and embed low-dimensional space, each with a low-dimensional representation; page 50583, section “C. HETEROGENEOUS NETWORK” By using the normalized matrices as edges weight, a heterogeneous network is generated which contains two node types {drug, disease} and three edge types {drug-drug, disease-disease, drug-disease});
wherein the method further comprises a decoding process to reconstruct association relationships between the nodes of a first object type and the nodes of a second object type, thereby predicting drug-disease association (page 50583, section “B. SCHEMATIC OVERVIEW OF HNRD” c) Reconstruct the drug-disease matrices with the captured feature vectors. This step is intended to minimize the different between the reconstructed matrices and the initial matrices. It also can be considered as an embedding process to maximize the extraction of information about the three matrices. e) Finally, predict the drug-disease sequence by reconstructing the matrix).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of a neural-network-based model to predict drug-disease associations as suggested in Wang into Fu’s system because both of these systems are addressing predicting links in biomedical bipartite networks. This modification would have been motivated by the desire to obtain performance improvement in identifying LDA associations. (Wang, page 50584, left column, 1st paragraph).
Claim 18 is rejected under 35 U.S.C. 103 as being unpatentable over Fu, in view of Wang as applied in claim 17, further in view of Yella et al, (hereinafter Yella), “MGATRx: Discovering Drug Repositioning Candidates Using Multi-View Graph Attention”, IEEE/ACM Trans Comput Biol Bioinform. 2022 Sep-Oct;19(5):2596-2604. doi: 10.1109/TCBB.2021.3082466.
Regarding dependent claim 18, the combination of Fu and Wang teaches all the limitations as set forth in the rejection of claim 17 that is incorporated. The combination of Fu and Wang does not explicitly teach further comprising minimizing a weighted binary cross-entropy loss:
PNG
media_image16.png
454
654
media_image16.png
Greyscale
However, in the same field of endeavor, Yella teaches
PNG
media_image16.png
454
654
media_image16.png
Greyscale
(page 2599, section “2.3.3 Loss Function”).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of predicting drug-disease association by calculating binary cross-entropy error of drug-disease association prediction and reconstruction error of the remaining views using mean squared error as suggested in Yella into Fu and Wang’s system because both of these systems are addressing graph based learning to unveil hidden relationships between the nodes in heterogeneous graphs. This modification would have been motivated by the desire to improve prediction performance and have the potential to unveil hidden relationships between the nodes in heterogeneous graphs (Yella, page 2596, section “1 INTRODUCTION”).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Applicant is required under 37 C.F.R. § 1.111(c) to consider these references fully when responding to this action.
MA et al. (US 20230027427 A1) discloses the processing of graph based data using machine learning techniques.
It is noted that any citation to specific pages, columns, lines, or figures in the prior art references and any interpretation of the references should not be considered to be limiting in any way. A reference is relevant for all it contains and may be relied upon for all that it would have reasonably suggested to one having ordinary skill in the art. In re Heck, 699 F.2d 1331, 1332-33, 216 U.S.P.Q. 1038, 1039 (Fed. Cir. 1983) (quoting In re Lemelson, 397 F.2d 1006, 1009, 158 U.S.P.Q. 275, 277 (C.C.P.A. 1968)).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to AMY P HOANG whose telephone number is (469)295-9134. The examiner can normally be reached M-TH 8:30-5:00PM.
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, JENNIFER WELCH can be reached at 571-272-7212. 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.
/AMY P HOANG/Examiner, Art Unit 2143
/JENNIFER N WELCH/Supervisory Patent Examiner, Art Unit 2143
Prosecution Insights