Abstract: A temporal-aware or permutation-dependent Graph Neural Network (GNN) is disclosed. The example GNN is implemented by combining temporal-awareness with multi-layer neighborhood aggregation to further provide the GNN with inductive capabilities with respect to generating embeddings of a dynamic graph, all without creating multiple time snapshots of the graph. By using a temporal-aware message pass scheme involving a temporal-aware and permutation-dependent GNN, a set of temporal-aware local neighborhood aggregator functions may be effective trained and used for generating embeddings for unknow nodes and for providing more accurate embeddings for subsequent prediction tasks.

Abstract: A device may receive source compound simplified molecular-input line-entry (SMILE) data, target compound SMILE data, and a latent space representing compounds, and may project the source compound SMILE data and the target compound SMILE data into the latent space to generate a source compound tensor and a target compound tensor, respectively. The device may process the source compound tensor, with one or more pretrained models, to determine a reward for the source compound tensor, and may determine, based on the reward, a direction and a magnitude to move in the latent space from the source compound tensor. The device may move the direction and the magnitude in the latent space to a new compound tensor, and may determine whether the new compound tensor matches the target compound tensor. The device may return a policy based on the new compound tensor matching the target compound tensor.

Abstract: A device may receive a knowledge graph and SMILE data identifying compounds, and may train embeddings based on the knowledge graph. The device may generate graph embeddings for the SMILE data based on the embeddings, and may encode the SMILE data into a latent space. The device may combine the graph embeddings and the latent space to generate a combined latent-embedding space, and may decode the combined latent-embedding space to generate decoded SMILE data. The device may utilize the decoded SMILE data to train an encoder, and may process source SMILE data, with the trained encoder, to generate a source combined latent-embedding space. The device may search the source combined latent-embedding space to identify new SMILE data, and may decode the new SMILE data to generate decoded new SMILE data. The device may evaluate the decoded new SMILE data to identify particular SMILE data associated with a new compound.

Abstract: Complex computer system architectures are described for utilizing a knowledge data graph comprised of elements, and selecting a discovery element to replace an existing element of a formulation depicted in the knowledge data graph. The substitution process takes advantage of the knowledge data graph structure to improve the computing capabilities of a computing device executing a substitution calculation by translating the knowledge data graph into an embedding space, and determining a discovery element from within the embedding space.

Abstract: Knowledge graph systems are disclosed for enhancing a knowledge graph by generating a new node. The knowledge graph system converts a knowledge graph into an embedding space, and selects a region of interest from within the embedding space. The knowledge graph system further identifies, from the region of interest, one or more gap regions, and calculates a center for each gap region. A node is generated for each gap region, and the information represented by the node is added to the original knowledge graph to generate an updated knowledge graph.

Abstract: Methods, systems and apparatus, including computer programs encoded on computer storage medium, for training a neurosymbolic data imputation system on training data inputs in a domain to impute missing data in a data input from the data domain. In one aspect a method includes, for each training data input, adding random noise to missing fields of the training data input; generating an embedding data input for the training data input using concept embeddings from the domain; processing the noisy data input and the embedding data input through a correlation network to obtain correlation data; applying attention to the noisy training data input and the correlation data to generate a combined data input; processing, by an autoencoder, the combined data input to obtain a decoded data output; computing a difference between the decoded data output and the training data input; and updating parameters of the data imputation system using the difference.

Abstract: The present disclosure relates to systems, methods, and products for identifying candidate molecule. The system includes a non-transitory memory storing instructions; and a processor in communication with the non-transitory memory. The processor executes the instructions to cause the system to receive drug data; convert the drug data into at least one point in a latent space using a grammar variational auto-encoder (VAE) model; receive a query for the at least one candidate molecule; select one or more points in the latent space; and create a k-dimensional tree graph based on the query for the at least one candidate molecule and the selected one or more points; determine a plurality of paths according to an interpolation technique; receive preference data; determine an optimum path; determine at least one candidate point on the optimum path; and determine a drug molecular structure using an inverse of the grammar VAE model.

Abstract: A device may generate a synthetic knowledge graph based on a true knowledge graph, may partition the synthetic knowledge graph into a set of synthetic data partitions, and may determine, using a plurality of teacher models, an aggregated prediction. The aggregated prediction may be based on individual predictions from corresponding individual teacher models included in the plurality of teacher models. The device may determine, using a student model and based on the synthetic knowledge graph and noise, a student prediction. The student model may be trained based on historical synthetic knowledge graphs and historical aggregated predictions associated with the plurality of teacher models. The device may determine an error metric based on the aggregated prediction and the student prediction, and may perform an action associated with the synthetic knowledge graph based on the error metric.

Abstract: A device processes a knowledge graph to select a set of triples and generates a first class based on the set of triples. The device determines a quantity of quasi-identifier attributes in the first class and compares the quantity to a predefined parameter. The device embeds the knowledge graph to generate an embedding space representation, identifies a second class, and determines a first quantity of nodes in the first class and the second class. The device compares the first quantity to the predefined parameter and identifies a third class. The device determines a second quantity of nodes in the first class and the third class and compares the second quantity to the predefined parameter. The device merges the second class or the third class with the first class, based on the comparisons, to generate anonymized nodes for the knowledge graph.

Abstract: A device may generate a synthetic knowledge graph based on a true knowledge graph, may partition the synthetic knowledge graph into a set of synthetic data partitions, and may determine, using a plurality of teacher models, an aggregated prediction. The aggregated prediction may be based on individual predictions from corresponding individual teacher models included in the plurality of teacher models. The device may determine, using a student model and based on the synthetic knowledge graph and noise, a student prediction. The student model may be trained based on historical synthetic knowledge graphs and historical aggregated predictions associated with the plurality of teacher models. The device may determine an error metric based on the aggregated prediction and the student prediction, and may perform an action associated with the synthetic knowledge graph based on the error metric.

Abstract: The present disclosure relates to systems, methods, and products for identifying candidate molecule. The system includes a non-transitory memory storing instructions; and a processor in communication with the non-transitory memory. The processor executes the instructions to cause the system to receive drug data; convert the drug data into at least one point in a latent space using a grammar variational auto-encoder (VAE) model; receive a query for the at least one candidate molecule; select one or more points in the latent space; and create a k-dimensional tree graph based on the query for the at least one candidate molecule and the selected one or more points; determine a plurality of paths according to an interpolation technique; receive preference data; determine an optimum path; determine at least one candidate point on the optimum path; and determine a drug molecular structure using an inverse of the grammar VAE model.

Abstract: Complex computer system architectures are described for analyzing data elements of a knowledge graph, and predicting new surprising or unforeseen facts from relational learning applied to the knowledge graph. This discovery process takes advantage of the knowledge graph structure to improve the computing capabilities of a device executing a discovery calculation by applying both training and inference analysis techniques on the knowledge graph within an embedding space, and generating a scoring strategy for predicting surprising facts that may be discoverable from the knowledge graph.

Abstract: The disclosure enables latent space exploration of a dataset based on drug molecular-structure data and drug biological-treatment data for a set of drug compounds in order to determine optimal drug compounds for treating diseases. Regional interpolation, including a linear interpolation (LERP) operation and a non-linear interpolation operation such as a spherical linear interpolation (SLERP), along with quantitative structure-activity relationship (QSAR) models may be utilized to navigate through a latent space generated from a variational auto-encoder (VAE), in accordance with certain embodiments.

Abstract: Complex computer system architectures are described for utilizing a knowledge data graph comprised of elements, and selecting a discovery element to replace an existing element of a formulation depicted in the knowledge data graph. The substitution process takes advantage of the knowledge data graph structure to improve the computing capabilities of a computing device executing a substitution calculation by translating the knowledge data graph into an embedding space, and determining a discovery element from within the embedding space.

Abstract: Knowledge graph systems are disclosed for enhancing a knowledge graph by generating a new node. The knowledge graph system converts a knowledge graph into an embedding space, and selects a region of interest from within the embedding space. The knowledge graph system further identifies, from the region of interest, one or more gap regions, and calculates a center for each gap region. A node is generated for each gap region, and the information represented by the node is added to the original knowledge graph to generate an updated knowledge graph.

Abstract: A device may receive individual information associated with individual activities of an individual, and may aggregate the individual information, based on a time period, to generate aggregated individual information. The device may identify patterns in the aggregated individual information, and may determine states for the patterns based on state information associated with activities capable of being performed by individuals. The device may generate a sequential knowledge graph based on modifying a knowledge graph with the states and adding a sequence of activities to the knowledge graph, and may determine embeddings for the individual activities based on the sequential knowledge graph.