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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea (mental process) without significantly more.
Regarding claim 1, in Step 1 of the 101 analysis set forth in MPEP 2106, the claim recites a method that makes a prediction based on data. A method is one of the four statutory categories of invention.
In Step 2a Pong 1 of the 101 analysis set forth in the MPEP 2106, the examiner has determined that the following limitations recite a process that, under the broadest reasonable interpretation, covers a mental process but for recitation of generic computer components:
generating, by the analysis computer, using a first recurrent neural network and an initial vector representation for a first user node from the plurality of user nodes, an updated vector representation for the first user node in response to a new interaction involving the first user node; (a person can mentally generate a representation based on processed data as a process of simply evaluating data and making a judgement based on the observed data. (MPEP 2106))
performing, by the analysis computer, a first prediction of a future interaction based on the updated vector representation for the first user node; and performing, by the analysis computer, an action based on the future interaction. (a person can mentally make a prediction and an action based on the prediction as a process of simply evaluating data and making a judgement based on the evaluation of the data. (MPEP 2106))
If claim limitations, under their broadest reasonable interpretation, covers performance of the limitations as a mental process but for the recitation of generic computer components, then it
falls within the mental process grouping of abstract ideas. According, the claim “recites” an abstract idea.
In Step 2a Prong 2 of the 101 analysis set forth in MPEP 2106, the examiner has determined that the following additional elements do not integrate this judicial exception into a practical application:
A method comprising: (Generally linking the use of the judicial exception to a particular technological environment or field of use (MPEP 2106.05(h))).
receiving, by an analysis computer, a graph comprising a plurality of user nodes for a plurality of users, a plurality of resource provider nodes for a plurality of resource providers, and a plurality of interaction edges between the plurality of user nodes and the plurality of resource provider nodes, the interaction edges representing a plurality of interactions between the plurality of users and the plurality of resource providers; (Adding insignificant extra-solution activity (mere data gathering) to the judicial exception (MPEP 2106.05(g)).
extracting, by the analysis computer, a dataset including initial vector representations for each of the plurality of user nodes and for each of the plurality of resource provider nodes; (Adding insignificant extra-solution activity (mere data gathering) to the judicial exception (MPEP 2106.05(g)).
Since the claim does not contain any other additional elements that are indicative of integration into a practical application, the claim is “directed” to an abstract idea.
In Step 2b of the 101 analysis set forth in the 2019 PEG, the examiner has determined that the claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
As discussed above, the additional element (iii) recites generally linking the use of the judicial exception to a particular technological environment or field of use, which is not indicative of significantly more. Additional elements (iv) and (v) recite insignificant extra solution activities. Further, elements (iv) and (v) recite steps of receiving data via a network, which has been determined by the courts to recite a well understood, routine and conventional activity which is not indicative of significantly more (See MPEP 2106.05(d)(II). Considering the additional elements individually and in combination, and the claim as a whole, the additional elements do not provide significantly more than the abstract idea. Therefore, the claim is not patent eligible.
Regarding claim 2, it is dependent upon claim 1, and thereby incorporates the limitations of, and corresponding analysis applied to claim 1. Further, claim 2 recites The method of claim 1, wherein generating the updated vector representation for the first user node is based on inputs to the first recurrent neural network, where the inputs to the first recurrent neural network include the initial vector representation for the first user node, a vector representation for a resource provider node associated with the new interaction, and features of the new interaction. (Generally linking the use of the judicial exception to a particular technological environment or field of use (MPEP 2106.05(h))).. Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 3, it is dependent upon claim 2, and thereby incorporates the limitations of, and corresponding analysis applied to claim 2. Further, claim 3 recites The method of claim 2, wherein the inputs to the first recurrent neural network further include one or more vector representations corresponding to one or more neighbor nodes of the first user node. (Generally linking the use of the judicial exception to a particular technological environment or field of use (MPEP 2106.05(h))).. Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 4, it is dependent upon claim 3, and thereby incorporates the limitations of, and corresponding analysis applied to claim 3. Further, claim 4 recites The method of claim 3, wherein the one or more neighbor nodes of the first user node include one-hop neighbors or two-hop neighbors. (Generally linking the use of the judicial exception to a particular technological environment or field of use (MPEP 2106.05(h))). Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 5, it is dependent upon claim 4, and thereby incorporates the limitations of, and corresponding analysis applied to claim 4. Further, claim 5 recites The method of claim 1, wherein performing the first prediction of the future interaction includes: determining, using a third recurrent neural network, a time-updated vector representation for the first user node based on an amount of time elapsed since a most recent interaction involving the first user node and a current time, wherein the future interaction is predicted based on the time-updated vector representation for the first user node. (In step 2A, prong 1, this recites a mental process but for recitation of generic computer components which is not indicative of integration into a practical application). Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 6, it is dependent upon claim 5, and thereby incorporates the limitations of, and corresponding analysis applied to claim 5. Further, claim 6 recites The method of claim 5, wherein predicting the future interaction includes: predicting, using a fourth recurrent neural network and the time- updated vector representation for the first user node, a vector; and determining, a resource provider node from the plurality of resource provider nodes with a vector representation that is closest to the predicted vector, wherein the future interaction is predicted to include the first user node and the determined resource provider node. (In step 2A, prong 1, this recites a mental process but for recitation of generic computer components which is not indicative of integration into a practical application). Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 7, it is dependent upon claim 6, and thereby incorporates the limitations of, and corresponding analysis applied to claim 6. Further, claim 7 recites The method of claim 6, further comprising: performing, by the analysis computer, a second prediction of features for the future interaction, where the second prediction is performed using a fifth recurrent neural network and inputs to the fifth recurrent neural network including the updated vector representation for the first user node and the vector representation of the determined resource provider node. (In step 2A, prong 1, this recites a mental process but for recitation of generic computer components which is not indicative of integration into a practical application). Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 8, it is dependent upon claim 7, and thereby incorporates the limitations of, and corresponding analysis applied to claim 7. Further, claim 8 recites The method of claim 7, further comprising: performing, by the analysis computer, a third prediction of an outcome for the future interaction, the third prediction is performed using a sixth recurrent neural network and inputs to the sixth recurrent neural network including the predicted features for the future interaction. (In step 2A, prong 1, this recites a mental process but for recitation of generic computer components which is not indicative of integration into a practical application). Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 9, it is dependent upon claim 8, and thereby incorporates the limitations of, and corresponding analysis applied to claim 8. Further, claim 9 recites The method of claim 8, wherein the predicted outcome for the future interaction is a probability of approval. (Generally linking the use of the judicial exception to a particular technological environment or field of use (MPEP 2106.05(h))). Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 10, it is dependent upon claim 9, and thereby incorporates the limitations of, and corresponding analysis applied to claim 9. Further, claim 10 recites The method of claim 9, wherein performing the action includes notifying a resource provider associated with the determined resource provider node about at least one of the future interaction, the predicted features for the future interaction, and the probability of approval. (Adding insignificant extra-solution activity (mere data gathering) to the judicial exception (MPEP 2106.05(g)).Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 11, it is dependent upon claim 10, and thereby incorporates the limitations of, and corresponding analysis applied to claim 10. Further, claim 11 recites The method of claim 10, wherein notifying includes providing a recommendation to take one or more subsequent actions, the subsequent actions including at least one of submitting an interaction for approval when the future interaction is initiated, not submitting the interaction for approval when the future interaction is initiated, and waiting until a recommended later time to submit the interaction for approval. (Adding insignificant extra-solution activity (mere data gathering) to the judicial exception (MPEP 2106.05(g)).Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 12, it is dependent upon claim 11, and thereby incorporates the limitations of, and corresponding analysis applied to claim 11. Further, claim 12 recites The method of claim 11, wherein the first recurrent neural network, the third recurrent neural network, the fourth recurrent neural network, the fifth recurrent neural network, and the sixth recurrent neural network each include corresponding learned coefficients trained using a machine learning model and known historical interaction data. (Generally linking the use of the judicial exception to a particular technological environment or field of use (MPEP 2106.05(h))). Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 13, it is dependent upon claim 12, and thereby incorporates the limitations of, and corresponding analysis applied to claim 12. Further, claim 13 recites The method of claim 12, further comprising: training, by the analysis computer, each of the first recurrent neural network, the third recurrent neural network, the fourth recurrent neural network, the fifth recurrent neural network, and the sixth recurrent neural network using the known historical interaction data. (In step 2A pong 2, training a system is a mere application of a computer tool (machine learning model), which is not indicative of integration into a practical application. In step 2B, merely applying a computer tool is not indicative of significantly more.). Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 14, it is dependent upon claim 13, and thereby incorporates the limitations of, and corresponding analysis applied to claim 13. Further, claim 14 recites The method of claim 13, wherein the training is based on a loss function that includes a term designed to minimize differences between predicted vectors and corresponding known resource provider vectors. (Generally linking the use of the judicial exception to a particular technological environment or field of use (MPEP 2106.05(h))). which is not indicative of integration into a practical application. In step 2B, merely applying a computer tool is not indicative of significantly more.). Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 16, it is dependent upon claim 15 which is a mirror of claim 1, and thereby incorporates the limitations of, and corresponding analysis applied to claim 15. Further, claim 16 recites The analysis computer of claim 15, wherein the graph is a bipartite graph. (Generally linking the use of the judicial exception to a particular technological environment or field of use (MPEP 2106.05(h))). which is not indicative of integration into a practical application. In step 2B, merely applying a computer tool is not indicative of significantly more.). Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 17, it is dependent upon claim 15 which is a mirror of claim 1, and thereby incorporates the limitations of, and corresponding analysis applied to claim 15. Further, claim 17 recites The analysis computer of claim 15, wherein each interaction edge from the plurality of interaction edges includes an associated feature vector containing values for one or more features including one or more of an amount, a time, a location, a type, and an outcome. (Generally linking the use of the judicial exception to a particular technological environment or field of use (MPEP 2106.05(h))). which is not indicative of integration into a practical application. In step 2B, merely applying a computer tool is not indicative of significantly more.). Since the claim does not recite additional elements that either integrate the judicial exception into a practical application, nor provide significantly more than the judicial exception, the claim is not patent eligible.
Regarding claim 15, it comprises of limitations similar to those of claim 1 and is therefor rejected for similar rationale. Regarding claims 18-20, they comprise of limitations similar to those of claims 2-4 and are therefor rejected for similar rationale.
Claim Rejections - 35 USC § 103
Claim(s) 1-4, 15-16, and 18-20 are rejected under 35 U.S.C. 103 as being unpatentable over HE(“Bipartite Graph Neural Networks for Efficient Node Representation Learning”) in view of LIU (“A hybrid method of recurrent neural network and graph neural network for next-period prescription prediction”).
Regarding claim 1, HE teaches the invention substantially as claimed, including:
A method comprising: receiving, by an analysis computer, a graph comprising a plurality of user nodes for a plurality of users, a plurality of resource provider nodes for a plurality of resource providers, and a plurality of interaction edges between the plurality of user nodes and the plurality of resource provider nodes, the interaction edges representing a plurality of interactions between the plurality of users and the plurality of resource providers; ((Abstract paragraph) In this paper, we propose the Bipartite Graph Neural Network (BGNN) that is domain-consistent, unsupervised, and efficient. BGNN consists of two central operations: interdomain message passing and intra-domain alignment, as illustrated in Figure 2. Suppose the two domains to be U and V , respectively. In each layer (depth) of BGNN, we formulate two kinds of information flows, one from U to V and the other from V to U, each of which is equipped with different weight filter. In this way, we attain inter-domain representation for each node.
(figure 1)
PNG
media_image1.png
161
241
media_image1.png
Greyscale
(as seen by the written and illustrated description, the information flow can be seen as edges between two node neighborhoods and it can be seen in the context of e-commerce having users and resource providers)) extracting, by the analysis computer, a dataset including initial vector representations for each of the plurality of user nodes and for each of the plurality of resource provider nodes; ((section 3, overall framework, paragraph 1) Here we introduce the overall framework of our model for better readability. In general, bipartite graph representation learning is to learn the embedding representations Hu 2 RP and Hv 2 RQ for nodes in group U and V , respectively.)
While HE does teach extracting the dataset and a graph of user and provider nodes, it does not explicitly teach:
generating, by the analysis computer, using a first recurrent neural network and an initial vector representation for a first user node from the plurality of user nodes, an updated vector representation for the first user node in response to a new interaction involving the first user node; performing, by the analysis computer, a first prediction of a future interaction based on the updated vector representation for the first user node; and performing, by the analysis computer, an action based on the future interaction.
However, in analogous art that similarly handles user and provider datasets, LIU teaches:
generating, by the analysis computer, using a first recurrent neural network and an initial vector representation for a first user node from the plurality of user nodes, an updated vector representation for the first user node in response to a new interaction involving the first user node; ((section 3.2 paragraph 1)The structure of our proposed hybrid method of decomposed LSTM and GNN (RGNN) is shown in Fig. 2. The medical history xpk of patient, pk is represented by decomposed LSTM and GNN separately, and then the representations from the two views are combined in different ways. Subsequently, the final representation of xpk (denoted by hpk) ) performing, by the analysis computer, a first prediction of a future interaction based on the updated vector representation for the first user node; and performing, by the analysis computer, an action based on the future interaction. ((section 3.2 paragraph 1) and the demographics dpk of patient pk are concatenated, transformed by a Rectified Linear Unit (ReLU) activation function, and fed into a Sigmod function to predict medications in the next time.)
It would have been obvious to a person skilled in the art before the effective filing date of the invention to have combined with LIU’s making a prediction with a recurrent network and, with HE’s teaching of a gathering user and provider datasets from a graph, to realize, with a reasonable expectation of success, a method that gathers datasets relating to users and providers to make a representation, as in HE, and updating that representation to make a prediction, as in LIU. A person of ordinary skill would have been motivated to make this combination to be prepared to better utilize available data (LIU section 1, paragraph 3).
Regarding claim 2, LIU further teaches:
The method of claim 1, wherein generating the updated vector representation for the first user node is based on inputs to the first recurrent neural network, where the inputs to the first recurrent neural network include the initial vector representation for the first user node, a vector representation for a resource provider node associated with the new interaction, and features of the new interaction. ((section 3.1, paragraph 1) Next-period prescription prediction can be recognized as a multi-label classification problem as follows: given a set of patients P = { p1,p2,…,pN } with their medical histo ries X = { xp1 , xp2 , …, xpN } and demographics D = { dp1 , dp2 , …, dpN }, we need to predict their prescriptions M={mp1 ,mp2 ,…,mpN} in the next time. For patient pk, xpk is a visit sequence that comprises sequences of different types of medical events such as diagnosis sequence, labo ratory test sequence, prescription sequence, etc. In this study, following Jin et al.’s work, we only consider labora tory test sequence and prescription sequence, denoted by xpk m = xpkm1 ,xpkm2 , …,xpkmt and xpk l tively, where xpk mi and xpk li = xpk l1 ,xpk l2 , … , xpk lt respec are medications and laboratory tests at the time i, represented by a one-hot vector of medi cation vocabulary (denoted as ∑) and a one-hot vector of laboratory test item vocabulary (denoted as L). We adopt xpk i = xpkmi , xpk li to denote the status of pk at the time t. In the case of M, mpk = xpkmt+1 . Figure 1 gives an example of medi cation vocabulary, where the numbers in parentheses are indices of items in the vocabularies.)
Regarding claim 3, HE further teaches:
The method of claim 2, wherein the inputs to the first recurrent neural network further include one or more vector representations corresponding to one or more neighbor nodes of the first user node. ((section 3.4) We summarize our overall implementation framework form BGNN model in Algorithm 1 which is consistent to Figure 3(a). The processes for set U and V are symmetric. Each step in the outmost loop proceeds as follows, where k represents the current layer and H(k) u ,H(k) v are hidden representations in layer k. For every epoch, sampling is conducted on these hidden representations to get mini-batch as input. After several epochs of training, the embedding representations of depth k can be learned and saved for k+1 layer training.)
Regarding claim 4, HE further teaches:
The method of claim 3, wherein the one or more neighbor nodes of the first user node include one-hop neighbors or two-hop neighbors. ((figure 2, description) Given the inputs of two domainsXu andXv, we obtain their unsupervised-embedded representations as Hu and Hv via inter-domain message passing and inter-domain distribution alignment. To enable multi-hop neighbor information aggregation, we stack multiple layers to formulate a deep BGNN whose layers are trained in a cascaded manner.)
Regarding claim 16, HE further teaches:
The analysis computer of claim 15, wherein the graph is a bipartite graph. ((Abstract) Existing Graph Neural Networks (GNNs) mainly focus on general structures, while the specific architecture on bipartite graphs—a crucial practical data form that consists of two distinct domains of nodes—is seldom studied. In this paper, we propose Bipartite Graph Neural Network (BGNN),)
Regarding claim 15, it comprises of limitations similar to those of claim 1 and is therefore rejected for similar rationale. Regarding claims 18-20, they comprise of limitations similar to those of claims 2-4 and are therefore rejected for similar rationale.
Claim(s) 5-8 are rejected under 35 U.S.C. 103 as being unpatentable over HE(“Bipartite Graph Neural Networks for Efficient Node Representation Learning”) in view of LIU (“A hybrid method of recurrent neural network and graph neural network for next-period prescription prediction”). In further view of GHAEINI (U.S. Pub. No. US 20200320387 A1)
Regarding claim 5, while HE as modified by LIU does teach claim 1, which claim 5 is dependent upon, it does not explicitly teach:
The method of claim 1, wherein performing the first prediction of the future interaction includes: determining, using a third recurrent neural network
However, in analogous art that similarly handles predictions, GHAEINI teaches:
The method of claim 1, wherein performing the first prediction of the future interaction includes: determining, using a third recurrent neural network (processing the data indicative of the premise independently using a third recurrent network to generate third independent premise data;),
It would have been obvious to a person skilled in the art before the effective filing date of the invention to have combined with GHAEINI’s usage of multiple recurrent networks and, with HE’s, as modified by LIU, teaching of utilizing the recurrent networks to make a prediction and update a vector representation, to realize, with a reasonable expectation of success, a method that generates a representation based on gathered data sets, as in HE as modified by LIU, which are processed using multiple recurrent networks, as in GHAEINI. A person of ordinary skill would have been motivated to make this combination to be prepared to better preserve information (GHAEINI [0004]).
LIU further teaches:
a time-updated vector representation for the first user node based on an amount of time elapsed since a most recent interaction involving the first user node and a current time, wherein the future interaction is predicted based on the time-updated vector representation for the first user node. ((section 3.2, paragraph 3) ( xpk i ) are linked to medical events at time i + 1 ( xpk i+1 ). The graph can be denoted by A∈∑×L×T(T = {1, 2, …, t}) and Ai∈∑×L is the adjacency matrix regarding xpk i and xpk i+1 . After obtaining co-occurrence graph A, we deploy GCN, a kind of GNN, to represent A as follows: where D . is the degree matrix of A, I is the identity matrix, Ws are weight matrices, and MEAN{.} is the mean function. Considering different time intervals between two neigh bor time points, we build a time-aware graph A by replacing ajk=1∈Ai by ajk= 1 Δti ∈Ai , where Δti is the time interval between the time i and time i + 1.)
Regarding claim 6, GHAEINI further teaches:
The method of claim 5, wherein predicting the future interaction includes: predicting, using a fourth recurrent neural network and the time- updated vector representation for the first user node, a vector([0011] In various embodiments, the method may further include wherein the first recurrent network is a first bidirectional long short term memory (Bi-LSTM) network, the second recurrent network is second Bi-LSTM network, the third recurrent network is a third Bi-LSTM network, the fourth recurrent network is a fourth Bi-LSTM network, the fifth recurrent network is a fifth Bi-LSTM network, the sixth recurrent network is a sixth Bi-LSTM network [0041] Image 300 can contain four Bi-LSTM blocks which in a variety of embodiments, can work together to independently and dependently read the premise and hypothesis. Bi-LSTM block 314 can independently read hypothesis 312 to generate an independent hypothesis vector space 322. Similarly, Bi-LSTM block 318 can independently read premise 310 to generate independent premise vector space 326. [0006] processing the data indicative of the hypothesis dependently with the third independent premise data using a fourth recurrent network to generate fourth dependent hypothesis data; pooling the second dependent premise data and the third independent premise independent data to combine independent and dependent premise data and generate pooled premise data;);
LIU further teaches:
and determining, a resource provider node from the plurality of resource provider nodes with a vector representation that is closest to the predicted vector, wherein the future interaction is predicted to include the first user node and the determined resource provider node. ((section 3.2 paragraph 3)( xpk i ) are linked to medical events at time i + 1 ( xpk i+1 ). The graph can be denoted by A∈∑×L×T(T = {1, 2, …, t}) and Ai∈∑×L is the adjacency matrix regarding xpk i and xpk i+1 . After obtaining co-occurrence graph A, we deploy GCN, a kind of GNN, to represent A as follows: where D . is the degree matrix of A, I is the identity matrix, Ws are weight matrices, and MEAN{.} is the mean function. Considering different time intervals between two neigh bor time points, we build a time-aware graph A by replacing ajk=1∈Ai by ajk= 1 Δti ∈Ai , where Δti is the time interval between the time i and time i + 1.)
Regarding claim 7, GHAEINI further teaches:
The method of claim 6, further comprising: performing, by the analysis computer, a second prediction of features for the future interaction, where the second prediction is performed using a fifth recurrent neural network and inputs to the fifth recurrent neural network including the updated vector representation for the first user node and the vector representation of the determined resource provider node. ([0008] In various embodiments, the method may further include processing the concatenated hypothesis vectors data independently using a fifth recurrent network to generate fifth hypothesis independent recurrent network data;)
Regarding claim 8 GHAEINI further teaches:
The method of claim 7, further comprising: performing, by the analysis computer, a third prediction of an outcome for the future interaction, the third prediction is performed using a sixth recurrent neural network and inputs to the sixth recurrent neural network including the predicted features for the future interaction. ([0008] processing the concatenated premise vectors data dependently with the fifth hypothesis independent recurrent network data using a sixth recurrent network to generate sixth premise dependent recurrent network data;)
Claim(s) 9-11 are rejected under 35 U.S.C. 103 as being unpatentable over HE(“Bipartite Graph Neural Networks for Efficient Node Representation Learning”) in view of LIU (“A hybrid method of recurrent neural network and graph neural network for next-period prescription prediction”), GHAEINI (U.S. Pub. No. US 20200320387 A1) in further view of LEVINE (U.S. Pub. No. US 20200357000 A1)
Regarding claim 9, while HE as modified by LIU and GHAEINI does teach claim 8, which claim 9 is dependent upon, it does not explicitly teach:
The method of claim 8, wherein the predicted outcome for the future interaction is a probability of approval.
However, in analogous art that similarly handles model predictions, LEVINE teaches:
The method of claim 8, wherein the predicted outcome for the future interaction is a probability of approval. ([0068] The prediction 138 output by the model 134 may be in the form of a confidence (probability) in the decision that the request 12 will be approved and/or denied and/or held.)
It would have been obvious to a person skilled in the art before the effective filing date of the invention to have combined with LEVINE’s prediction output and, with HE’s, as modified by LIU and GHAVEINI, teaching of utilizing the recurrent networks to make a prediction, to realize, with a reasonable expectation of success, a method that generates a representation based on gathered data sets and makes a prediction using those representations, as in HE as modified by LIU and GHAEINI, where the prediction includes a probability of approval, as in LEVINE. A person of ordinary skill would have been motivated to make this combination to be prepared to assist in the decision making process (LEVINE [0005]).
LEVINE further teaches:
The method of claim 9, wherein performing the action includes notifying a resource provider associated with the determined resource provider node about at least one of the future interaction, the predicted features for the future interaction, and the probability of approval. ([0112] At S114, the request 12 and related proposal 18 are output from the system 10, by the submission component 120, e.g., to a regulatory agency for making an agency decision on the request. The submission component 120 may incorporate the proposed decision 20 on the new request, the confidence measure 22, and the rationale 24 for the decision into a proposal template 138, which may be auto-filled with other information relating to the request, such as an ID for the request, requestor information, or the like, in the event that the request 12 and the proposal 18 become separated. Alternatively, the proposal 18 may be incorporated into the submitted request 12, e.g., as an attachment.)
LEVINE further teaches:
The method of claim 10, wherein notifying includes providing a recommendation to take one or more subsequent actions, the subsequent actions including at least one of submitting an interaction for approval when the future interaction is initiated, not submitting the interaction for approval when the future interaction is initiated, and waiting until a recommended later time to submit the interaction for approval. ([0113] At S116, a template 152 for the regulator's response may be prefilled with information from the request and proposal and provided to the regulatory agency. A reviewer, after receiving and reviewing the request 12, related proposal 18, and the template response 152, may modify the template response 152 to generate the agency response 26, accept the template response 152 as the agency response 26, or otherwise generate the agency response 26. The agency response 26 may be sent to the requestor directly or via the system 10.)
Claim(s) 12-14 are rejected under 35 U.S.C. 103 as being unpatentable over HE(“Bipartite Graph Neural Networks for Efficient Node Representation Learning”) in view of LIU (“A hybrid method of recurrent neural network and graph neural network for next-period prescription prediction”), GHAEINI (U.S. Pub. No. US 20200320387 A1), LEVINE (U.S. Pub. No. US 20200357000 A1),in further view of SADEGH (U.S. Pub. No. US 20120150764 A1)
Regarding claim 12, while HE as modified by LIU, GHAEINI, and LEVINE, does teach claim 11, which claim 12 is dependent upon, it does not explicitly teach:
The method of claim 11, wherein the first recurrent neural network, the third recurrent neural network, the fourth recurrent neural network, the fifth recurrent neural network, and the sixth recurrent neural network each include corresponding learned coefficients trained using a machine learning model and known historical interaction data.
However, in analogous art that similarly trains a model, SADEGH teaches:
The method of claim 11, wherein the first recurrent neural network, the third recurrent neural network, the fourth recurrent neural network, the fifth recurrent neural network, and the sixth recurrent neural network each include corresponding learned coefficients trained using a machine learning model and known historical interaction data. ([0046] Maximum likelihood, least squares, or similar estimation algorithms may be used to train the parameters/coefficients of the model using historical data.)
It would have been obvious to a person skilled in the art before the effective filing date of the invention to have combined with SADEGH’s training and, with HE’s, as modified by LIU, GHAVEINI, and LEVINE, recurrent models, to realize, with a reasonable expectation of success, a method utilizes many recurrent models to generate a prediction, as in HE as modified by LIU and GHAEINI, and LEVINE, which have coefficients that are trained, as in SADEGH. A person of ordinary skill would have been motivated to make this combination to be prepared to reduce costs (SADEGH [0005]).
SADEGH further teaches:
The method of claim 12, further comprising: training, by the analysis computer, each of the first recurrent neural network, the third recurrent neural network, the fourth recurrent neural network, the fifth recurrent neural network, and the sixth recurrent neural network using the known historical interaction data. ([0042] The mathematical input-output structures in the business relations, as well as historical time series data in the database, may be used to train and build mathematical models.)
HE further teaches:
The method of claim 13, wherein the training is based on a loss function that includes a term designed to minimize differences between predicted vectors and corresponding known resource provider vectors. ((section 3.2, paragraph 1) We introduce IDA from the perspective of domain U. As mentioned previously in Eq. (4), we design two types of alignment losses to align Hv!u to Hu.)
Claim(s) 17 is rejected under 35 U.S.C. 103 as being unpatentable over HE(“Bipartite Graph Neural Networks for Efficient Node Representation Learning”) in view of LIU (“A hybrid method of recurrent neural network and graph neural network for next-period prescription prediction”). In further view of MIAO (C.N. Pub. No. CN 109636608 A)
Regarding claim 17, while HE as modified by LIU does teach claim 15, which claim 17 is dependent upon, it does not explicitly teach:
The analysis computer of claim 15, wherein each interaction edge from the plurality of interaction edges includes an associated feature vector containing values for one or more features including one or more of an amount, a time, a location, a type, and an outcome.
However, in analogous art that similarly teaches a graph network, MIAO teaches:
The analysis computer of claim 15, wherein each interaction edge from the plurality of interaction edges includes an associated feature vector containing values for one or more features including one or more of an amount, a time, a location, a type, and an outcome. ((Page 4, paragraph 9) In data processing and account behavioural characteristic vector building module, before converting and being expressed as account behavioural characteristic vector Initial data feature source it is varied, these features may come from coding of accounts and essential information, membership number and base This information, the commission order information of account, order deal message, the statistical information held position, transaction terminal information etc..Wherein, with The relevant feature of commission order may include commission time, commission direction, commission quantity, commission seat number, commission price, commission Batch number, commission order type are opened and put down mark, remove simple correlation information, membership information, triggered time, product and contract number etc. ； With The relevant feature of trade order may include conclusion of the business serial number, closing time, concluded price, conclusion of the business quantity, conclusion of the business analogue letter Breath, conclusion of the business seat number, conclusion of the business membership number, dealing are opened f lat mark, product and contract number etc. ； Feature relevant to information of holding position can be with Including the same day dealing amount of holding position, held position with yesterday increment, guarantee fund's amount of money, product and contract number etc.)
It would have been obvious to a person skilled in the art before the effective filing date of the invention to have combined with MIAO’s information and, with HE’s, as modified by LIU, teaching of recurrent networks used to make predictions, to realize, with a reasonable expectation of success, a method that generates a prediction, as in HE as modified by LIU, using information that includes various information types, as in MIAO. A person of ordinary skill would have been motivated to make this combination to be make a more expansive and manageable information base (MIAO Background paragraph 3).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SKIELER A KOWALIK whose telephone number is (571)272-1850. The examiner can normally be reached 8-5.
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, Mariela D Reyes can be reached at (571)270-1006. 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.
/SKIELER ALEXANDER KOWALIK/Examiner, Art Unit 2142 /Mariela Reyes/Supervisory Patent Examiner, Art Unit 2142
Prosecution Insights