Abstract: Example storage systems and methods provide data storage management using parallel processing of filtered transaction logs. Transaction logs are comprised of log entries corresponding to storage operations for at least one storage node. Sets of log entries are sequentially retrieved from the transaction log and filtered through multiple transaction log filters to generate multiple subsets of the log entries. Different metadata operations are executed in parallel using the different filtered subsets of log entries.

Abstract: A program stored on non-transitory computer-readable storage medium executes a method of evaluating a graph over a query. Decomposition instructions decompose the query into a plurality of subqueries. Evaluation instructions evaluate a subquery of the plurality of subqueries and generate a substitution multiset representing a result of the evaluation of the subquery. Filtration instructions or expansion instructions may operate upon the generated substitution set before passing the substitution set to a next subquery to be evaluated. The filtration instructions identify one or more mappings in the substitution multiset that cannot be safely passed to the second subquery and delete the identified one or more mappings from the substitution multiset. The expansion instructions determine, in a case where the subquery is operated upon by a non-distributive query operator, an expansion of the substitution multiset based at least on adding one or more new substitutions to the substitution multiset.

Abstract: Embodiments of systems and methods for information retrieval are disclosed. Embodiments of such systems and methods may perform information retrieval based on a language model that is used to generate a single vector for the search terms of a query. Similarly, a single vector representation of each of the data records to be searched is obtained and the single vector representing the search terms of the query compared to the single vector of each data record to determine a similarity metric. The resulting similarity metrics associated with each of the data records can be used to rank, present or return one or more data records.

Abstract: A computer-implemented method includes receiving a query specifying an operation to perform on a first table of a plurality of data blocks stored. Each data block in the first table includes a respective reference count indicating a number of tables referencing the data block. The method also includes determining that the operation specified by the query includes copying the plurality of data blocks in the first table into a second table and, in response, for each data block of the plurality of data blocks in the first table copied into the second table, incrementing, the respective reference count associated with the data block in the first table, appending, by the data processing hardware, into metadata of the second table, a reference of the corresponding data block copied into the second table.

Abstract: Systems and methods are provided to ingest and integrate data objects for use in one or more system operations including providing a renderable data object to a user and updating a data item database.

Abstract: In a computer-implemented method for maintaining a time series database including a plurality of time series database schemas, time series data including data points are received at an ingestion node of a time series database, the data points comprising a plurality of dimensions. A plurality of time series database schemas of the time series database is determined for storing the time series data. The time series data is ingested according to the plurality of time series database schemas, wherein each data point is stored according to each time series database schema of the plurality of time series database schemas, such that the time series database comprises multiple instances of each data point.

Abstract: The described technology is generally directed towards managing data retention policy for stream data stored in a streaming storage system. When a request to truncate a data stream from a certain position (e.g., from a request-specified stream cut) is received, an evaluation is made to determine whether the requested position is within a data retention period as specified by data retention policy. If any data prior to the stream cut position (corresponding to a stream cut time) is within the data retention period, the truncation request is blocked. Otherwise truncation from the stream cut point is allowed to proceed/is performed. Also described is handling automated (e.g., sized based) stream truncation requests with respect to data retention.

Abstract: Multi-modal differential search with real-time focus adaptation techniques are described that overcome the challenges of conventional techniques in a variety of ways. In one example, a model is trained to support a visually guided machine-learning embedding space that supports visual intuition as to “what” is represented by text. The visually guided language embedding space supported by the model, once trained, may then be used to support visual intuition as part of a variety of functionality. In one such example, the visually guided language embedding space as implemented by the model may be leveraged as part of a multi-modal differential search to support search of digital images and other digital content with real-time focus adaptation which overcomes the challenges of conventional techniques.

Abstract: A method for accelerating data operations across a plurality of nodes of one or more clusters of a distributed computing environment. Rack awareness information characterizing the plurality of nodes is retrieved and a non-volatile memory (NVM) capability of each node is determined. A write operation is received at a management node of the plurality of nodes and one or more of the rack awareness information and the NVM capability of the plurality of nodes are analyzed to select one or more nodes to receive at least a portion of the write operation, wherein at least one of the selected nodes has an NVM capability. A multicast group for the write operation is then generated wherein the selected nodes are subscribers of the multicast group, and the multicast group is used to perform hardware accelerated read or write operations at one or more of the selected nodes.

Abstract: An example method of determining geolocations of objects based on information retrieved from heterogeneous data sources comprises: receiving, from a first data source associated with an object by an ontology-defined relationship, a first dataset including a first data item specifying a first time identifier and a first geolocation associated with the object; receiving, from a second data source associated with an object by an ontology-defined relationship, a second dataset including a second data item specifying a second time identifier and a second geolocation associated with the object; and determining, by applying a rule set associated with the ontology to the first dataset and the second dataset, a geolocation of the object and a corresponding time identifier.

Abstract: Interaction output over a local computer-readable medium (CRM) generated based on user interaction with rendered content input representing a virtualized asset being is received at a virtualized asset local provisioning server. A manner to exploit the virtualized asset is determined from interaction output. The virtualized asset is exploited based on the determined manner to exploit the virtualized asset. A request for a portion of the virtualized asset generated in response to the exploiting the virtualized asset is intercepted. If it is determined that the portion of the virtualized asset is absent from the local storage, a request for the portion of the virtualized asset is sent to a virtualized asset delivery system over a non-local CRM; the portion of the virtualized asset retrieved by the virtualized asset delivery system is received over the non-local CRM; and the received portion of the virtualized asset is used in exploiting the virtualized asset.

Abstract: Methods, apparatus, systems, computing devices, computing entities, and/or the like for using machine-learning concepts (e.g., machine learning models) to determine predicted taxonomy-based classification scores for claims and dynamically update data fields based on the same.

Abstract: A URL and a categorization associated with the URL are received. A key associated with the received URL is determined. An operation is performed on a database using the determined key. Examples of such operations include inserting the categorization into the database, changing a value associated with the key in the database, removing a key-value pair from the database, and querying the database.

Abstract: In a distributed database, many nodes can store copies, or instances, of the same record. If the record is split on one node, it should be split on the other nodes to maintain consistency, concurrency, and correctness of the data in the distributed database. In some distributed databases, the records are locked during the update process to ensure data integrity. Unfortunately, locking the records can increase latency, especially for larger databases. But if the records aren't locked and a node fails as a record is being split and updated simultaneously, the split and update may not propagate throughout the distributed database, leading to a loss of data integrity. Exchanging messages about the status of record splitting and forwarding updates internally reduces the likelihood of a loss of data integrity due to a node failure.

Abstract: A database contains a corpus of incident reports, a machine learning (ML) model trained to calculate paragraph vectors of the incident reports, and a look-up set table that contains a list of paragraph vectors respectively associated with sets of the incident reports. A plurality of ML worker nodes each store the look-up set table and are configured to execute the ML model. An update thread is configured to: determine that the look-up set table has expired; update the look-up set table by: (i) adding a first set of incident reports received since a most recent update of the look-up set table, and (ii) removing a second set of incident reports containing timestamps that are no longer within a sliding time window; store, in the database, the look-up set table as updated; and transmit, to the ML worker nodes, respective indications that the look-up set table has been updated.

Abstract: The illustrative storage management system uses snapshot technologies and performs block-level analysis thereof and to synthesize full backups of a database, regardless of the database management system (DBMS) and content of the database and without involving native DBMS utilities. Synthetic full copies are created efficiently and use network transport resources sparingly. Successive snapshots of the database may be hardware snapshots and/or software snapshots. The illustrative storage management system tracks which data blocks in the database change over time after a baseline full copy is saved to secondary storage. Only the changed blocks are subsequently moved to secondary storage as incremental backups and are indexed for future reference. A synthetic full backup of the database, at a given point in time, may be created by synthesizing the original full backup with changed blocks from any number of later incremental backups.

Abstract: A method and system for accessing updated data from a database in response to a user query has been developed. First, multiple transaction logs are generated for a database. Each transaction log contains a record of actions executed by a database management system and referenced according to the specified date of the actions. Data updates are received and stored with the database. An incremental database partition is created for each data update. Each incremental database partition is stored with reference to a corresponding transaction log for the date of the data update. The updated data is accessed through the incremental database partition in response to an outdated user query. The outdated user query contains a data access request for a date earlier than the receipt of data updates.

Abstract: Performing quantum file copying is disclosed herein. In one example, upon receiving a request to copy a source quantum file comprising a plurality of source qubits, a quantum file manager accesses a quantum file registry record identifying the plurality of source qubits and a location of each of the plurality of source qubits. The quantum file manager next allocates a plurality of target qubits equal in number to the plurality of source qubits, and copies data stored by each of the source qubits into a corresponding target qubit. The quantum file manager then generates a target quantum file registry record that identifies the plurality of target qubits and their locations. In some examples, a quantum file move operation may be performed by deleting the source quantum file after the copy operation, and updating the target quantum file registry record with the same quantum file identifier as the source quantum file.

Abstract: A memory stores graph information representing a graph that includes nodes and inter-node edges. The nodes include a first plurality of nodes each associated with node information and a first node. Each of the inter-node edges has a weight. A processor extracts, in accordance with the node information, two or more nodes and transforms the two or more nodes into an aggregate node. The processor generates an aggregate inter-node edge between the aggregate node and the first node. The aggregate inter-node edge is associated with a weight based on two or more weights associated with two or more inter-node edges between the two or more nodes and the first node. The processor estimates first node information to be associated with the first node based on transformed graph information representing a transformed graph including the aggregate node and the aggregate inter-node edge.

Abstract: Methods, systems, and computer-readable storage media for receiving, by a current database system, a query plan file representative of a captured query plan from a source database system, receiving, by the current database system, a set of definitions including one or more definitions, each definition in the set of definitions corresponding to an object that is implicated by the query plan, the object being included in a set of objects, and determining, by the current database system, that each definition in the set of definitions is identical to a respective definition of a corresponding object within the current database system, and in response: executing the captured query plan in the current database system to provide a query result.