Abstract: Method and system for processing digital works, the method comprising the steps of identifying terms within each digital work of a plurality of digital works, wherein the terms are words and/or phrases. Determining a number of times that the identified terms occur within each digital work of the plurality of digital works. Generating a fingerprint for each digital work of the plurality of digital works, the generated fingerprint based on the identified terms and the number of times that the identified terms occur within each digital work. Using a neural network to find an encoding function, g, that encodes a higher dimensionality space, x, of each fingerprint into a lower dimensionality space, y. Applying the encoding function to each fingerprint of the plurality of digital works to reduce their dimensionality. Determining a similarity between a first fingerprint and one or more dimensionality reduced fingerprints.

Abstract: A schema associated with a type of document is received. A set of documents are queried for documents that match the schema, wherein the query is performed while bypassing user intervention. A subset of documents of the set of documents is identified that match the schema. The subset of documents that match the schema are transmitted to a document controller.

Abstract: Embodiments of the present invention include generating, at a first time, a first copy of a database including multiple tables, the first copy including a set of first structured text files. During a time period subsequent to the first time, queries that request updates to the tables during the time period are periodically generated and conveyed to a system managing the database. Responsively to the periodic queries, any updates to the tables and their respective update times are received and stored to a journal including one or more second structured text files. At a subsequent time, a request to generate a second copy of the database is received, the second copy including a state of the database at a specified time during the time period. The first copy of the database is updated with the updates whose update times are prior to the specified time, thereby generating the second copy.

Abstract: A system and method for managing queries including receiving a new query comprising a first plurality of conjoined terms, accessing a filtered view of a database from memory, the filtered view being filtered by the previously received query according to a filter represented by a second plurality of conjoined terms, at least one of the first plurality of conjoined terms or the second plurality of conjoined terms including at least one NULL value, determining that the filter of the new query implies a filter of the previously received query, and based on the determination that the filter of the new query implies the filter of the previously received query, executing the new query using the filtered view of the previously received query.

Abstract: A query directed at a table organized into a set of batch units is received. The query comprises a predicate for which values are unknown prior to runtime. A set of values for the predicate are determined based on the query. An index access plan is created based on the set of values. Based on the index access plan, the set of batch units are pruned using a pruning index associated with the table. The pruning index comprises a set of filters that index distinct values in each column of the table. The pruning of the set of batch units comprises identifying a subset of batch units to scan for data that satisfies the query. The subset of batch units of the table are scanned to identify data that satisfies the query.

Abstract: Embodiments for a storage classifier that provides recommendations to a backup server for storage targets among a plurality of disparate target storage types. The storage classifier receives metadata (name, type, size), and the Service Level Agreement with information such as: retention time, Recovery Point Objective, and Recovery Time Objective) from the backup software. The backup software itself receives policy recommendations from a data label rules engine based on certain file attributes. The storage classifier receives an initial recommendation for the storage type and location (e.g., on-premises deduplication storage or public-cloud object storage, etc.) from a data classifier. Based on these inputs, the storage classifier provides recommended specific storage targets to the backup software on a file-by-file basis for data stored in a backup operation.

Abstract: A system and method for detecting anomalies in very large datasets is disclosed. The method includes calculating statistics for data elements in a data set over a range of time periods. These statistics are arranged into a 2D array and analyzed using a machine learning algorithm to detect anomalous regions. The method also includes steps of analyzing time series of the data based on detected anomalous regions, correcting any errors in the datasets, and storing the corrected values in a separate database to maintain data integrity.

Abstract: A system includes a communications interface configured to receive a plurality of data series captured by a plurality of sensors. A first database includes at least a first data series of the plurality of data series. A second database includes metadata related to the first data series. One or more processors are configured to receive an instruction including an indication of a computation to perform and metadata filter criteria, query the second database to determine at least a portion of the first data series that satisfies the metadata filter criteria by reference to at least the metadata related to the first data series, access, from the first database and based on results of the query of the second database, the portion of the first data series, and perform the computation on the portion of first data series accessed from the first database.

Abstract: In an example, one or more leaf category specific unsupervised statistical language model (SLM) models are trained using sample item listings corresponding to each of one or more leaf categories and structured data about the one or more leaf categories, the training including calculating an expected perplexity and a standard deviation for item listing titles. A perplexity for a title of a particular item listing is calculated and a perplexity deviation signal is generated based on a difference between the perplexity for the title of the particular item listing and the expected perplexity for item listing titles in a leaf category of the particular item listing and based on the standard deviation for item listing titles in the leaf category of the particular item listing. A gradient boosting machine (GBM) fuses the perplexity deviation signal with one or more other signals to generate a miscategorization classification score corresponding to the particular item listing.

Abstract: Methods for rowgroup consolidation with delta accumulation and versioning in distributed systems are performed. The systems provide performant methods of row storage that enable versioned modifications of data while keeping and allowing access to older versions of the data for point-in-time transactions. The accumulation of valid rows, deletes, and modifications is maintained in blobs for rowgroups until a size threshold is reached, at which point the rows are moved into a columnar compressed form. Changes to data and associated metadata are stored locally and globally via appends, maintaining logical consistency. Metadata is stored in footers of files allowing faster access to the metadata and its associated data for transactions and instant rollback via metadata version flipping for aborted transactions, as well as lock-free reads of data.

Abstract: A pruning index is generated for a source table organized into a set of batch units. The source table comprises a column of semi-structured data. The pruning index comprises a set of filters that index distinct values in each column of the source table. Rather than reassembling an entire tree structure of the semi-structured data prior to indexing, the generating of the pruning index comprises traversing a reassembly hook object that represents a first portion of the semi-structured data that is subcolumnarized and traversing a residual object that represents a second portion of the semi-structured data that is not subcolumnarized. The reassembly hook object is traversed to identify values corresponding to the first portion of the semi-structured data and the residual object is traversed to identify values corresponding to the second portion. The pruning index is stored with an association with the source table.

Abstract: Computer-readable media, methods, and systems are disclosed for performing rollback recovery with data lineage capture for data pipelines. A middle operator receives ingested input events from a source operator reading data from an external input data source. The middle operator then logs information regarding middle input events to a middle operator input log, designating the logged middle input event information as incomplete. The middle operator then processes data associated with the middle input events and updates the middle input log entries setting them to a completed logging status designation for middle input events that were consumed to produce the one or more middle output events. The middle operator then transmits the middle output events to subsequent operators. Garbage collection is performed to remove completed entries from the middle operator output log. Finally, based on receiving a recovering message from a subsequent operator, corresponding middle output events are re-sent.

Abstract: A query directed at a source table organized into a set of batch units is received. The query includes a pattern matching predicate that specifies a search pattern. A set of N-grams are generated based on the search pattern. A pruning index associated with the source table is accessed. The pruning index comprises a set of filters that index distinct N-grams in each column of the source table. The pruning index is used to identify a subset of batch units to scan for matching data based on the set of N-grams generated for the search pattern. The query is processed by scanning the subset of batch units.

Abstract: A method and system for tracking the history of possession, activity, and changes related to an electronic document after its creation includes a document in a file system on an computing device having linked metadata which tracks hash values that update when a user performs a document action. Each hash value indicates a fixed document state. The metadata stores hash value identifiers associated with the initial document state and the document state right before the most recent document action and a list of identifiers associated with every state of the document in its history. The method and system allows for branching of multiple documents from a common state, identified by a common hash value and includes components to visualize a document's lineage. The method and system also includes the ability to store information in a document regarding similar, or near duplicate, documents that do not share the same lineage.

Abstract: Embodiments of the invention are directed toward a computer-implemented interactive diagnostic medical records system operated in conjunction with at least one user device. Interactive processing components provide collaborative interaction between a user and at least one database, the collaboration including deployment of a search module in response to a query associated with the user device, and a diagnostic inquiry module comprising an interactive sequence of at least two linked inquiries transmitted to the user device. In one embodiment, the diagnostic inquiry module is operable to assign a value or “weight” to the user responses to the interactive sequence of at least two linked inquiries. User interface tools may further be associated with the user device for accepting input from the diagnostic inquiry module and correspondingly displaying output as determined by the interactive processing components.

Abstract: A highly secure database platform is entity-centric and entity-controlled. Posting, query and retrieval of a data is bound to unique, unidirectional relationship identifiers that control a requesting entity's access to documents pertaining to a targeted entity. Data is immutably mapped to entities and relationships. The platform is domain-partitioned, schema-agnostic and order-preserving. The invention presents a trusted platform or service which is readily adapted to data concerning people, groups, businesses, devices, and/or microservices.

Abstract: Reducing backup data by providing a data protection target for storing content data from clients running operating system and applications, and a common data protection target (CDPT) separate from the data protection target for storing Gold image data for the operating system and applications. During backup, user content data is copied from the client to the data protection target, and Gold image data is copied to the CDPT. It is also referenced in the DP target to prevent redundant storage. During backup, the CDPT is queried to determine if the Gold image exists, and if so, the DP target does not store the Gold image data in the DP target, but uses the reference to indicate the location of the Gold image data corresponding to the backed up content data. During a restore, the restore stream is built from the DP target and CDPT to combine user content data and Gold image data.

Abstract: Reducing backup data by providing a data protection target for storing content data from clients running operating system and applications, and a common data protection target (CDPT) separate from the data protection target for storing Gold image data for the operating system and applications. Gold image data is stored by a user in a defined location or tagged with a defined tag. The stored or tagged Gold image data is identified by the backup software and copied to the CDPT. The identified Gold image data is cataloged. During backup, the CDPT is queried to determine if the Gold image exists, and if so, the DP target does not store the Gold image data in the DP target, but uses the reference to indicate the location of the Gold image data corresponding to the backed up content data. During a restore, the restore stream is built from the DP target and CDPT to combine user content data and Gold image data.

Abstract: Systems and method for improving query performance by querying an appropriate database engine based on the operation of the query request is provided. In one aspect, this approach involves querying a row-oriented database, querying a column-oriented database, or blacklisting the query request. In particular, updating the column-oriented database involves delete and insert operations. By maintaining updated databases and querying appropriate database engines, the response time of a query request may be improved.

Abstract: A method for global optimization is disclosed. The method may include receiving a search request that may include an input. The method may further determine an amount of rotations necessary to perform the search request with a Grover Search algorithm. Then, the method may include determining that the amount of rotations is less than a predefined amount. Further, the method may generate one or more quantum walks. The one or more quantum walks and the Grover Search algorithm may be used to generated a global optimization algorithm. The method may then execute the global optimization algorithm to identify the input.