Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for realizing a multimodal image classifier. In an aspect, a method includes, for each image of a plurality of images: processing the image by a textual generator model to obtain a set of phrases that are descriptive of the content of the image, wherein each phrase is one or more terms, processing the set of phrases by a textual embedding model to obtain an embedding of predicted text for the image, and processing the image using an image embedding model to obtain an embedding of image pixels of the image. Then a multimodal image classifier is trained on the embeddings of predicted text for the images and the embeddings of image pixels for the images to produce, as output, labels of an output taxonomy to classify an image based on the image as input.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for realizing a multimodal image classifier. In an aspect, a method includes, for each image of a plurality of images: processing the image by a textual generator model to obtain a set of phrases that are descriptive of the content of the image, wherein each phrase is one or more terms, processing the set of phrases by a textual embedding model to obtain an embedding of predicted text for the image, and processing the image using an image embedding model to obtain an embedding of image pixels of the image. Then a multimodal image classifier is trained on the embeddings of predicted text for the images and the embeddings of image pixels for the images to produce, as output, labels of an output taxonomy to classify an image based on the image as input.

Abstract: A transformation on raw data is applied to produce transformed data, where the transformation includes at least one selected from among a summary of the raw data or a transform of the raw data between different domains. In response to a query to access data, the query is processed using the transformed data.

Abstract: Example implementations relate to updating an inference graph absent node locking. For example, a processor executing a first thread may receive a first task for updating a node of an inference graph stored by a storage device accessible to a second thread, the first task being assigned during a first iteration of a graph update loop. Absent locking the node from access by the second thread, the processor may generate a value for the node and update the node with the value. Based on detecting that each node of the inference graph has been updated, the processor may continue with a second iteration of the graph update loop.

Abstract: Providing a targeted security alert can include collecting participant data from a plurality of participants within a threat exchange community, calculating, using a threat exchange server, a threat relevancy score of a participant among the plurality of participants within the threat exchange community using the collected participant data, and providing, from the threat exchange server to the participant, the targeted security alert based on the calculated threat relevancy score via a communication link within the threat exchange community.

Abstract: Storing time series data for a search query includes identifying a time series whose representation is to be pre-computed based on available memory storage, pre-computing at least one representation of the identified time series, and storing the at least one representation in the memory storage.

Abstract: An example technique includes assigning partitions of a dataset of multidimensional points to a plurality of local memory nodes of a multicore machine and using the local memory nodes for a search query to determine similarity matches in the dataset for a given multidimensional point. The using includes parallel searching with the local memory nodes in the assigned partitions to identify candidate similarity matches to the given multidimensional point using indexes derived from the multidimensional points, the parallel searching for each node progressing through a sequence of search distances and providing an ongoing search result for each search distance from the given multidimensional point and regulating an extent of the parallel searching based on the ongoing search results.

Abstract: Examples herein involve graph processing using a shared memory. An example method includes distributing vertices of a graph to a plurality of graph partition processors of a system, the graph partition processors to process non-overlapping sets of vertices of the graph; storing a master copy of a vertex state of the graph in a shared memory of the system; instructing the graph partition processors to iteratively process respective vertices of the graph partitions based on local copies of the state of the graph stored in respective local memories of the graph partition processors; and updating the master copy of the state of the graph in the shared memory based on the iterative processing of the partitions of the vertices by the graph partition processors until convergence of the graph.

Abstract: Example implementations relate to updating an inference graph absent node locking. For example, a processor executing a first thread may receive a first task for updating a node of an inference graph stored by a storage device accessible to a second thread, the first task being assigned during a first iteration of a graph update loop. Absent locking the node from access by the second thread, the processor may generate a value for the node and update the node with the value. Based on detecting that each node of the inference graph has been updated, the processor may continue with a second iteration of the graph update loop.

Abstract: A technique includes performing graph inference in a graph inference engine that includes multiple processing nodes to determine assignments for vertices of a graph. Performing the graph inference includes controlling remote memory accesses within the engine, including storing first data in a local memory of the first processing node, where the first data represents at least assignments for a plurality of vertices of the graph; in the first processing node, determining updates for the assignments for a subset of the plurality of vertices of a partition of the graph assigned to the first processing node and modifying the first data based on the updates; and communicating the updates to at least one other processing node of the multiple processing nodes, where at least one other partition of the graph is assigned to the other processing node(s).

Abstract: Method, systems, and computer-readable storage devices for updating a prediction model are described. In one aspect, a statistical analysis group assignment may be received. The statistical analysis group assignment may group partition-level worker node and a first set of partition-level worker nodes as a statistical analysis group. A statistical analysis phase may then be executed where a group-level decision tree is generated from statistical data and other statistical data received from the first set of partition-level worker nodes. A decision tree analysis phase may then be executed, where a step decision tree may be generated based on a selection from the group-level tree and other group-level trees received from other statistical analysis groups. The prediction model may be caused to be updated using the step decision tree.

Abstract: A transformation on raw data is applied to produce transformed data, where the transformation includes at least one selected from among a summary of the raw data or a transform of the raw data between different domains. In response to a query to access data, the query is processed using the transformed data.

Abstract: Examples herein involve graph processing using a shared memory. An example method includes distributing vertices of a graph to a plurality of graph partition processors of a system, the graph partition processors to process non-overlapping sets of vertices of the graph; storing a master copy of a vertex state of the graph in a shared memory of the system; instructing the graph partition processors to iteratively process respective vertices of the graph partitions based on local copies of the state of the graph stored in respective local memories of the graph partition processors; and updating the master copy of the state of the graph in the shared memory based on the iterative processing of the partitions of the vertices by the graph partition processors until convergence of the graph.

Abstract: An example technique includes assigning partitions of a dataset of multidimensional points to a plurality of local memory nodes of a multicore machine and using the local memory nodes for a search query to determine similarity matches in the dataset for a given multidimensional point. The using includes parallel searching with the local memory nodes in the assigned partitions to identify candidate similarity matches to the given multidimensional point using indexes derived from the multidimensional points, the parallel searching for each node progressing through a sequence of search distances and providing an ongoing search result for each search distance from the given multidimensional point and regulating an extent of the parallel searching based on the ongoing search results.

Abstract: A method for event correlation includes capturing events and arranging the events sequentially in at least one dimension. An event correlator implemented by a computational device convolves a kernel density function with each of the events to produce a convolved function for each event. Co-occurrences between events are found by calculating overlap between convolved functions.

Abstract: Storing time series data for a search query includes identifying a time series whose representation is to be pre-computed based on available memory storage, pre-computing at least one representation of the identified time series, and storing the at least one representation in the memory storage.

Abstract: Disclosed herein are a system, non-transitory computer readable medium, and method to handle a routing problem. It is determined whether a fastest route can be resolved using static routing or time dependent routing.

Abstract: Sharing information can include identifying, utilizing a threat exchange server, a security occurrence associated with a participant within a threat exchange community. Sharing information can also include determining what participant-related information to share with the threat exchange server in response to the identified security occurrence, and receiving, at the threat exchange server, information associated with the determined participant-related information via communication links within the threat exchange community.

Abstract: Providing a targeted security alert can include collecting participant data from a plurality of participants within a threat exchange community, calculating, using a threat exchange server, a threat relevancy score of a participant among the plurality of participants within the threat exchange community using the collected participant data, and providing, from the threat exchange server to the participant, the targeted security alert based on the calculated threat relevancy score via a communication link within the threat exchange community.

Abstract: Probable anomalies associated with at least one data metric may be detected across a series of windows of time series data by comparison of data to a threshold. An estimated probability of anomalies for each of the windows of time series data may be determined based on the detected probable anomalies and the threshold. The windows of time series data may be ranked based on the estimated probabilities. Probable anomalies associated with highest ranked windows of time series data may be output to a user.