Abstract: Techniques and solutions are described for compressing data and facilitating access to compressed data. Compression can be applied to proper data subsets of a data set, such as to columns of a table. Using various methods, the proper data subsets can be evaluated to be included in a group of proper data subsets to be compressed using a first compression technique, where unselected proper data subsets are not compressed using the first compression technique. Data in the data set can be reordered based on a reordering sequence for the proper data subsets. Reordering data in the data set can improve compression when at least a portion of the proper data subsets are compressed. A data structure is provided that facilitates accessing specified data stored in a compressed format.

Abstract: Systems and methods are disclosed herein for a recommendations engine that generates content recommendations using a trained model that is personalized based on the information corresponding to content consumption. The disclosed techniques herein provide a trained model to provide content recommendations. The trained model may have been trained using a predefined set of training data agnostic of a particular user profile. A system receives information corresponding to content consumption. The system may associate the information corresponding to content consumption with a profile. The system generates a personalized model based on the information corresponding to content consumption and on the trained model. The personalized model may be associated with the user profile. The system generates the content recommendations using the personalized model. The system then causes to be provided the content recommendations.

Abstract: A method and apparatus for estimating the cardinality of graph pattern queries using graph statistics and metadata is presented. In various embodiments, node and edge labels are used to compute estimates for graph patterns (bi-grams) and the estimates for these patterns as composed to provide cardinality estimates of longer paths. The computation of bi-grams is low cost and requires only minimal changes to the existing query planner in the existing database machinery to be useful. The resulting estimates are used during query execution to minimize the amount of work needed to complete a query and return a faithful and accurate answer to the user. The apparatus includes modules containing computer-executable instructions to perform the above method.

Abstract: A method includes obtaining, at a device, a hierarchical structure representing a graphical layout of content items of an electronic document, the content items including at least text. The method also includes generating a word embedding representing a word of the electronic document. The method further includes determining position information of a location of the word in the electronic document. The method also includes determining a descriptor that indicates a relationship of the location to the hierarchical structure. The method further includes providing input data to a machine learning model to generate a semantic region category label of a semantic region of the electronic document. The semantic region includes the word. The input data includes the word embedding, the position information, and the descriptor.

Abstract: Access to data representations of relatively dynamic actual usage and patterns of spectrum usage across channels, geographies (locations), and times can be advantageous in planning for resource utilization, and in structuring pricing for spectrum resources. Actual usage of spectrum resources can vary dynamically. Significant opportunities to advantageously utilize spectrum resources can be identified from relatively dynamic representations of actual usage.

Abstract: A database system and method supporting parallel update includes: a concurrency control protocol, which caches max Read Timestamp, max Write Timestamp and max Normal Write Timestamp for each data item in the database; a write-ahead log module, which records normal logs and delta logs; a cache module, which caches normal records with the largest version number of the data item and all subsequent delta records; a persistence module, which saves the records in the cache to a persistent medium; a write control module, which controls the write operation, and which, if transaction timestamp is greater than the max Read Timestamp, determine and execute normal write and incremental write respectively and generate normal records and delta records, and update the max Write Timestamp and the max Normal Write Timestamp; and a read control module maintaining a dependency graph according to dependency between the write transaction and the read transaction.

Abstract: Example implementations relate to provisioning a deduplication data store. An example includes receiving inputs indicating an expected data amount and at least one retention period for the deduplication data store; performing a comparison of the at least one retention period to a stored data structure; determining at least one storage multiplier based on the comparison of the at least one retention period to the stored data structure; and determining an amount of storage to provision for the deduplication data store based on the expected data amount and the at least one storage multiplier.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer-readable media that utilize machine learning models to generate identifier embeddings from digital content identifiers and then leverage these identifier embeddings to determine digital connections between digital content items. In particular, the disclosed systems can utilize an embedding machine-learning model that comprises a character-level embedding machine-learning model and a word-level embedding machine-learning model. For example, the disclosed systems can combine a character embedding from the character-level embedding machine-learning model and a token embedding from the word-level embedding machine-learning model. The disclosed systems can determine digital connections between the plurality of digital content items by processing these identifier embeddings for a plurality of digital content items utilizing a content management model.

Abstract: A computerized search method includes receiving first input designating a first location of a first analytic. The method includes, in response to a scheduling event, obtaining a first document from the first location, identifying a first predefined label within the first document, obtaining first and second data associated with the first predefined label, storing the first datum into a value index as a current value of the first analytic, and storing the second datum into a text index as a textual description of the first analytic. The method includes presenting a search interface and, in response to receiving a search query from a user: identifying a set of result analytics relevant to the search query based on the text index and presenting, for each of the result analytics, a textual description of the analytic from the text index and a most recent value of the analytic from the value index.

Abstract: A method is provided including presenting a user interface to allow a user of a social networking service to create a new collection for collecting a set of content items for the user, and upon a user request to create the new collection, adding the new collection to a list of collections associated with a user profile of the user, and presenting, in the user interface, the new collection as part of the list of collections associated with the user profile at the social networking service.

Abstract: Multi-site distributed storage systems and computer-implemented methods are described for improving a resumption time of input/output (I/O) operations during a common snapshot process for storage objects. A computer-implemented method comprises performing a baseline transfer from at least one storage object of a first storage node to at least one replicated storage object of a second storage node, starting the common snapshot process including stop processing of I/O operations, performing a snapshot create operation on the primary storage site for the at least one storage object of the first storage node, resuming processing of I/O operations, and assigning a new universal unique identifier (UUID) to the at least one storage object of the second storage node after resuming processing of I/O operations with the new UUID to identify when file system contents are different than the baseline transfer.

Abstract: Methods, systems, apparatuses, and computer program products are described. A database system may store data logs tracking performance data for multiple tenants of the database system. The database system may receive, via a user interface of a user device, a user input indicating a request to perform an analysis of performance data associated with one of the multiple tenants. The request may indicate a time period and a performance metric for the analysis. The database system may execute a query based on the analysis. The query may execute on a subset of the data logs associated with the tenant and the performance metric. The system may analyze a result of the query to determine one or more entities that affected the performance metric during the time period. The system may transmit an indication of the one or more entities for display at the user interface.

Abstract: Embodiments automate the timely remediation of issues by matching failed operations to a script that is configured to fix the problem or problems without causing additional side effects, thus minimizing unavailability of services/data and loss of data. A method detects a backup operation alert message sent from a data manager or storage system that encodes one or more attributes affected by the backup operation. A script library is scanned to identify scripts that match the alert message. A component generates a score for each script, where the score factors the relevance of the script to the backup operation, and executes the script with the highest score, as the script most likely to remediate the issue causing the alert message.

Abstract: In one implementation, a method for providing a user with an assessment of website quality includes receiving, at a computer system, identification of a website; retrieving, by the computer system, content that comprises the website; and automatically analyzing, by the computer system, the content. The method can further include determining, by the computer system, (i) a search engine optimization (SEO) sub-score for the website, (ii) an accessibility sub-score for the website, and (iii) a quality assurance score for the website based on the analysis of the content. The method can additionally include determining, by the computer system, a quality score for the website based on a combination of the SEO sub-score, the accessibility sub-score, and the quality assurance sub-score; and transmitting, by the computer system, the quality score to a client computing device for presentation in a user interface.

Abstract: At a logging service, a representation of a transaction requested by a client is received. The transaction includes at least a first write operation directed to a data object at a first data store. Based at least in part on the contents of the representation and contents of a persistent log of the logging service, a determination is made that the requested transaction is to be committed. A log record corresponding to the first write operation is inserted into the persistent log. The first write is propagated to the first data store after the record has been inserted in the persistent log.

Abstract: The described embodiments set forth techniques for preserving clone relationships between files at a computing device. In particular, the techniques involve identifying clone relationships between files in conjunction with performing operations on the files where it can be beneficial to preserve the clone relationships. The operations can include, for example, preserving clone relationships between files that are being copied from a source storage device (that supports file cloning) to a destination storage device that supports file cloning. Additionally, the operations can include preserving clone relationships when backing up and restoring files between a source storage device (that supports file cloning) and a destination storage device that does not support file cloning. In this manner, the various benefits afforded by the clone relationships between files can be retained even as the files are propagated to destination storage devices that may or may not support file cloning.

Abstract: Systems and methods are described for distributed processing a query in a first query language utilizing a query execution engine intended for single-device execution. While distributed processing provides numerous benefits over single-device processing, distributed query execution engines can be significantly more difficult to develop that single-device engines. Embodiments of this disclosure enable the use of a single-device engine to support distributed processing, by dividing a query into multiple stages, each of which can be executed by multiple, concurrent executions of a single-device engine. Between stages, data can be shuffled between executions of the engine, such that individual executions of the engine are provided with a complete set of records needed to implement an individual stage. Because single-device engines can be significantly less difficult to develop, use of the techniques described herein can enable a distributed system to rapidly support multiple query languages.

Abstract: A computing system receives a plurality of payloads from a plurality of computing devices, each of the plurality of payloads comprising one or more events for a pageview corresponding to a respective website. For every predefined time period, the computing system collects a plurality of payloads received within the predefined time period, stores the collected plurality of payloads in a single file in a datastore, generates metadata associated with the single file, and stores the generated metadata in a database comprising metadata for a plurality of single files. When the computing system receives a request for a pageview of a website, the computing system accesses the database to determine a plurality of payloads within one or more single files comprising payloads for the pageview of the website. The computing system generates the pageview by assembling the plurality of payloads for the pageview of the website.

Abstract: An apparatus comprises at least one processing device that includes a processor coupled to a memory. The processing device is configured to store configuration information, for utilization by software of a host device, in at least one file of a first database of a first type, and in conjunction with an upgrade of at least a portion of the software, to export the configuration information from the at least one file of the first database to one or more files in an intermediate format, to remove the at least one file of the first database, to export the configuration information from the one or more files in the intermediate format to at least one file of a second database of a second type different than the first type, and to configure the upgraded software to utilize the configuration information of the at least one file of the second database.

Abstract: Technologies for performing random sparse lifting and Procrustean orthogonal sparse hashing using column read-enabled memory include a device that has a memory that is column addressable and circuitry connected to the memory. The circuitry is configured to add a set of input data vectors to the memory as a set of binary dimensionally expanded vectors, including multiplying each input data vector with a projection matrix. The circuitry is also configured to produce a search hash code from a search data vector, including multiplying the search data vector with the projection matrix. Further, the circuitry is configured to determine a Hamming distance between the search hash code and each of the binary dimensionally expanded vectors.