Abstract: Embodiments of the present invention provide systems, methods, and computer storage media for determining an increased matching for large graphs in which an increased matching is generated for the graph by leveraging an initial matching for a small fraction of edges of the large graph. An initial matching for a random subset of edges of an input graph is leveraged to generate alternating paths based on the initially matched edges and the remaining edges, not included in the random subset. An increased matching for the entire graph includes the alternating paths without the initial matched edges, thus increasing the number of matched edges in the increased matching by at least one for every initially matched edge. Graph-based tasks may then be triggered based on the increased matching.

Abstract: System and methods for relational time-series learning are provided. Unlike traditional time series forecasting techniques, which assume either complete time series independence or complete dependence, the disclosed system and method allow time series forecasting that can be performed on multivariate time series represented as vertices in graphs with arbitrary structures and predicting a future classification for data items represented by one of nodes in the graph. The system and methods also utilize non-relational, relational, temporal data for classification, and allow using fast and parallel classification techniques with linear speedups. The system and methods are well-suited for processing data in a streaming or online setting and naturally handle training data with skewed or unbalanced class labels.

Abstract: A system and method for automatically adjusting computing resources provisioned for a computer service or application by applying historical resource usage data to a predictive model to generate predictive resource usage. The predictive resource usage is then simulated for various service configurations, determining scaling requirements and resource wastage for each configuration. A cost value is generated based on the scaling requirement and resource wastage, with the cost value for each service configuration used to automatically select a configuration to apply to the service. Alternatively, the method for automatically adjusting computer resources provisioned for a service may include receiving resource usage data of the service, applying it to a linear quadratic regulator (LQR) to find an optimal stationary policy (treating the resource usage data as states and resource-provisioning variables as actions), and providing instructions for configuring the service based on the optimal stationary policy.

Abstract: In some embodiments, a network analysis system receives network data in the form of a temporal graph that includes nodes and edges. Each node represents an entity involved in a network. An edge connects two nodes to indicate an association between the two nodes. Each edge also has a temporal value indicating a time point when the association between the two nodes was created. The network analysis system generates a sequence of nodes by traversing the nodes in the temporal graph along edges with non-decreasing temporal values or with non-increasing temporal values. The network analysis system further replaces the identifiers of the nodes in the sequence to generate a sequence of feature values. Based on the sequence of feature values, the network analysis system determines network embeddings for the nodes in the temporal graph. Using the network embeddings, the network analysis system identifies two or more of the nodes in the temporal graph that belong to the same entity.

Abstract: System and methods for relational time-series learning are provided. Unlike traditional time series forecasting techniques, which assume either complete time series independence or complete dependence, the disclosed system and method allow time series forecasting that can be performed on multivariate time series represented as vertices in graphs with arbitrary structures and predicting a future classification for data items represented by one of nodes in the graph. The system and methods also utilize non-relational, relational, temporal data for classification, and allow using fast and parallel classification techniques with linear speedups. The system and methods are well-suited for processing data in a streaming or online setting and naturally handle training data with skewed or unbalanced class labels.

Abstract: Embodiments of the present invention are generally directed to generating figure captions for electronic figures, generating a training dataset to train a set of neural networks for generating figure captions, and training a set of neural networks employable to generate figure captions. A set of neural networks is trained with a training dataset having electronic figures and corresponding captions. Sequence-level training with reinforced learning techniques are employed to train the set of neural networks configured in an encoder-decoder with attention configuration. Provided with an electronic figure, the set of neural networks can encode the electronic figure based on various aspects detected from the electronic figure, resulting in the generation of associated label map(s), feature map(s), and relation map(s).

Abstract: Techniques are provided for training graph neural networks with heterophily datasets and generating predictions for such datasets with heterophily. A computing device receives a dataset including a graph data structure and processes the dataset using a graph neural network. The graph neural network defines prior belief vectors respectively corresponding to nodes of the graph data structure, executes a compatibility-guided propagation from the set of prior belief vectors and using a compatibility matrix. The graph neural network predicts predicting a class label for a node of the graph data structure based on the compatibility-guided propagations and a characteristic of at least one node within a neighborhood of the node. The computing device outputs the graph data structure where it is usable by a software tool for modifying an operation of a computing environment.

Abstract: A visualization recommendation system generates recommendation scores for multiple visualizations that combine data attributes of a dataset with visualization configurations. The visualization recommendation system maps meta-features of the dataset to a meta-feature space and configuration attributes of the visualization configurations to a configuration space. The visualization recommendation system generates meta-feature vectors that describe the mapped meta-features, and generates configuration attribute sets that describe the attributes of the visualization configurations. The visualization recommendation system applies multiple scoring models to the meta-feature vectors and configuration attribute sets, including a wide scoring model and a deep scoring model. In some cases, the visualization recommendation system trains the multiple scoring models using the meta-feature vectors and configuration attribute sets.

Abstract: The present disclosure describes systems, non-transitory computer-readable media, and methods for utilizing hash partitions to determine local densities and distances among users (or among other represented data points) for clustering sparse data into segments. For instance, the disclosed systems can generate hash signatures for users in a sparse dataset and can map users to hash partitions based on the hash signatures. The disclosed systems can further determine local densities and separation distances for particular users (or other represented data points) within the hash partitions. Upon determining local densities and separation distances for datapoints from the dataset, the disclosed systems can select a segment (or cluster of data points) grouped according to a hierarchy of a clustering algorithm, such as a density-peaks-clustering algorithm.

Abstract: The present disclosure describes systems and methods for information retrieval. Embodiments of the disclosure provide a retrieval network that leverages external knowledge to provide reformulated search query suggestions, enabling more efficient network searching and information retrieval. For example, a search query from a user (e.g., a query mention of a knowledge graph entity that is included in a search query from a user) may be added to a knowledge graph as a surrogate entity via entity linking. Embedding techniques are then invoked on the updated knowledge graph (e.g., the knowledge graph that includes additional edges between surrogate entities and other entities of the original knowledge graph), and entities neighboring the surrogate entity are retrieved based on the embedding (e.g., based on a computed distance between the surrogate entity and candidate entities in the embedding space). Search results can then be ranked and displayed based on relevance to the neighboring entity.

Abstract: In some embodiments, a data visualization system detects insights from a dataset and computes insight scores for respective insights. The data visualization system further computes insight type scores, from the insight scores, for insight types in the detected insights. The data visualization system determines a selected insight type for the dataset having a higher insight type score than unselected insight types and determines, for the selected insight type, a set of selected insights that have higher insight scores than unselected insights. The data visualization system determines insight visualizations for the set of selected insights and generates, for inclusion in a user interface of the data visualization system, selectable interface elements configured for invoking an editing tool for updating the determined insight visualizations from the dataset. The selectable interface elements are arranged in the user interface according to the insight scores of the set of selected insights.

Abstract: This disclosure describes one or more embodiments of systems, non-transitory computer-readable media, and methods that intelligently and automatically analyze input data and generate visual data stories depicting graphical visualizations from data insights determined from the input data. For example, the disclosed systems automatically extract data insights utilizing an in-depth statistical analysis of dataset groups from data-attribute categories within the input data. Based on the data insights, the disclosed systems can automatically generate exportable visual data stories to visualize the data insights, provide textual or audio-based natural language summaries of the data insights, and animate such data insights in videos. In some embodiments, the disclosed systems generate a visual-data-story graph comprising nodes representing visual data stories and edges representing similarities between the visual data stories.

Abstract: In implementations of systems for suggesting content components, a computing device implements a design system to receive input data describing a feature of a content component to be included in a hypertext markup language (HTML) document. The design system represents that feature of the content component as a document object model (DOM) element and determines a hash value for the DOM element using locality-sensitive hashing. Manhattan distances are computed between the has value and has values described by a segment of content component data. The hash values were determined using the locality-sensitive hashing for DOM elements extracted from a corpus of HTML documents. The design system generates indications, for display in a user interface, of candidate content components for inclusion in the HTML document based on the Manhattan distances.

Abstract: Embodiments described herein involve selecting outlier-detection programs that are specific to meta-features of datasets. For instance, a computing system constructs a performance vector from a U vector and a reference V matrix. Vector elements of the performance vector identify estimated performance values of various outlier-detection programs with respect to an input dataset. The U vector is generated using meta-features of the input dataset. The reference V matrix is generated from a training process in which performance values of the various outlier-detection programs with respect to training input datasets are used to obtain the reference V matrix via a UV decomposition. The computing system selects an outlier-detection program having a greater estimated performance value in the performance vector as compared to other outlier-detection programs' respective estimated performance values.

Abstract: A system and method for fast, accurate, and scalable typed graphlet estimation. The system and method utilizes typed edge sampling and typed path sampling to estimate typed graphlet counts in large graphs in a small fraction of the computing time of existing systems. The obtained unbiased estimates of typed graphlets are highly accurate, and have applications in the analysis, mining, and predictive modeling of massive real-world networks. During operation, the system obtains a dataset indicating nodes and edges of a graph. The system samples a portion of the graph and counts a number of graph features in the sampled portion of the graph. The system then computes an occurrence frequency of a typed graphlet pattern and a total number of typed graphlets associated with the typed graphlet pattern in the graph.

Abstract: Systems and methods for personalized visualization recommendation are described.

Abstract: Deriving network embeddings that represent attributes of, and relationships between, different nodes in a network while preserving network data temporal and structural properties is described. A network representation system generates a plurality of graph time-series representations of network data that each includes a subset of nodes and edges included in a time segment of the network data, constrained either by time or a number of edges included in the representation. A temporal graph of the network data is generated by implementing a temporal model that incorporates temporal dependencies into the graph time-series representations. From the temporal graph, network embeddings for the network data are derived, where the network embeddings capture temporal dependencies between nodes, as indicated by connecting edges, as well as temporal structural properties of the network data.

Abstract: The present disclosure describes systems, non-transitory computer-readable media, and methods for utilizing hash partitions to determine local densities and distances among users (or among other represented data points) for clustering sparse data into segments. For instance, the disclosed systems can generate hash signatures for users in a sparse dataset and can map users to hash partitions based on the hash signatures. The disclosed systems can further determine local densities and separation distances for particular users (or other represented data points) within the hash partitions. Upon determining local densities and separation distances for datapoints from the dataset, the disclosed systems can select a segment (or cluster of data points) grouped according to a hierarchy of a clustering algorithm, such as a density-peaks-clustering algorithm.

Abstract: In implementations of deep hybrid graph-based forecasting systems, a computing device implements a forecast system to receive time-series data describing historic computing metric values for a plurality of processing devices. The forecast system determines dependency relationships between processing devices of the plurality of processing devices based on time-series data of the processing devices. Time-series data of each processing device is represented as a node of a graph and the nodes are connected based on the dependency relationships. The forecast system generates an indication of a future computing metric value for a particular processing device by processing a first set of the time-series data using a relational global model and processing a second set of the time-series data using a relational local model. The first and second sets of the time-series data are determined based on a structure of the graph.

Abstract: In implementations of systems for generating occurrence contexts for objects in digital content collections, a computing device implements a context system to receive context request data describing an object that is depicted with additional objects in digital images of a digital content collection. The context system generates relationship embeddings for the object and each of the additional objects using a representation learning model trained to predict relationships for objects. A relationship graph is formed for the object that includes a vertex for each relationship between the object and the additional objects indicated by the relationship embeddings. The context system clusters the vertices of the relationship graph into contextual clusters that each represent an occurrence context of the object in the digital images of the digital content collection.