Abstract: Methods, systems, and computer program products for active learning for natural language question answering are provided herein. A computer-implemented method includes generating a semantic signature of a natural language query; generating a SQL signature for a SQL query corresponding to the natural language query; determining whether a set of mappings includes a semantic signature matching the generated semantic signature, wherein each mapping in the set is between (i) a given semantic signature and (ii) a SQL signature representing a class of SQL queries corresponding to the given semantic signature; and in response to determining that the set of mappings does not include the generated semantic signature, adding a mapping between the generated semantic signature and the generated SQL signature to the set of mappings.

Abstract: A method for storing video data includes, when receiving the I-frame data to be stored, detecting whether the written data exists in the video cache space; when detecting that the written data exists in the video cache space, reading a target writing position of the I-frame data to be stored and determining whether the target writing position is located within a position range corresponding to the written data in the first cache space; when determining the target writing position is located within the position range, writing, based on the target writing position, the I-frame data to be stored to the first cache space for caching and detecting whether the first cache space is full; and when detecting that the first cache space is full, writing all the video data in the video cache space to a memory space of the terminal device for storage and emptying the video cache space.

Abstract: A method at a computing device for data management between a publisher and at least one subscriber, the method including receiving, at a system element, memory requirements from the publisher; creating a memory allocation of a pool of data objects for the publisher based on the received memory requirements; receiving, at the system element, consumption criteria from each of the at least one subscriber; and adjusting the memory allocation of the pool of data objects based on the consumption criteria received from the at least one subscriber.

Abstract: A database system with a transactional commit protocol based on a safe conjunction of majorities is disclosed. In various embodiments, a directed acyclic graph comprising, for each of a plurality of database transactions, data identifying the transaction, data indicating one or more prior database transactions on which the database transaction depends, and data indicating whether the transaction has been committed is stored. A leading edge comprising a set of committed leaf nodes of the directed acyclic graph is determined. Data representing the leading edge is provided to a transaction coordinator configured to use the leading edge to determine based at least in part on the leading edge to commit a transaction.

Abstract: A method, apparatus, and non-transitory computer-readable storage medium for presenting information identifying relationships in a database. Entity labels identifying a plurality of different entity types modeled in the database are presented in a user interface. Input identifying one or more selected entity types is received. A plurality of entity instance relationship sets is accessed. Each entity instance relationship set identifies a relationship between entity instances of at least two different entity types. Attribute data of the entity instances identified in only those entity instance relationship sets that identify a relationship between entity instances of the one or more selected entity types are presented, such that attribute data for each unique relationship between entity instances of the one or more selected entity types are presented only once.

Abstract: This application relates to systems and methods for providing search results based on a primary intent. In some examples, a disclosed system includes a memory resource storing instructions; and one or more processors coupled to the memory resource. The one or more processors are configured to execute the instructions to: receive, from a user, a search query including a plurality of words, identify a plurality of intention terms from the words of the search query, compute, for each of the plurality of intention terms, a compatibility score between the intention term and a query context associated with the intention term, determine, from the plurality of intention terms, a primary intention term having a maximum compatibility score among the plurality of intention terms, and generate, based on the primary intention term, a set of intent-based search results in response to the search query, the set of intent-based search results identifying a set of items associated with the primary intention term.

Abstract: Techniques are disclosed relating to database schemas. A computer system may receive a metadata document that describes an updated schema for a database that stores a plurality of records. The computer system may instantiate a set of processes to perform an upgrade routine to upgrade ones of the plurality of records to comply with the updated schema. While the set of processes are upgrading the records, the computer system may receive a request to perform an operation on one of the plurality of records. The computer system may detect that the record corresponds to an earlier schema version than a version of the updated schema of the metadata document and may upgrade the record to comply with that updated schema. After upgrading the record, the computer system may perform the requested operation on the record.

Abstract: Embodiments of the present disclosure provide a data protection method, an electronic device, and a computer program product. The method includes determining an object feature for each protection object in a set of protection objects that generate protected data, the set of protection objects including at least one protection object configured with a predetermined data protection strategy. The method further includes determining a set of candidate objects belonging to the same class as the at least one protection object from the set of protection objects according to the determined object features. The method further includes configuring the predetermined data protection strategy to at least one candidate object in the set of candidate objects.

Abstract: An indication to remove one or more data in a time series database is received. A metadata index associated with the time series database is updated to indicate a soft removal of each data of the one or more data. A data hole index associated with the time series database is updated to indicate a data hole at a location of each data of the one or more data in the time series database. Responsive to an input/output rate for the time series database being below a threshold, the data hole of each data of the one or more data is optimized.

Abstract: A method is disclosed. The method includes receiving, by a second entity computer associated with a second user from a first entity computer associated with a first user, a push transaction message comprising a value and link data for a transaction. The method also comprises transmitting, by the second entity computer to a directory computer in communication with a database storing supplemental data associated with the transaction and other transactions, a request for the supplemental data associated with the transaction, the request comprising link data. The method also includes receiving, by the second entity computer, the supplemental data from the directory computer, and providing, by the second entity computer to the second user via a second user device, the supplemental data along with a notification of an action caused by the push transaction message.

Abstract: The present disclosure provides a method for graph computing, an electronic device and a non-transitory computer-readable storage medium. An execution engine for managing execution of the graph computing is configured in a CPU. One or more interface functions running in a GPU for processing parameters of the graph computing are configured. During the execution of the graph computing, the one or more interface functions are called by the execution engine through a graph computing interface. The interface functions are executed in the GPU in parallel through multiple parallel threads. The interface functions are configured to process a plurality of graph vertexes in parallel. The multiple parallel threads are configured to feedback respective execution results of the interface functions to the execution engine. The graph computing is completed by the execution engine according to the execution results.

Abstract: Disambiguation of search keywords to expert systems is performed through the application of a semantic ranking function to determine a classification category with confidence for a given query. The semantic ranking function is executed through an application of a semantic distance distribution profile score through analysis of user problem statement (UPS) with nodes in the taxonomy in the same vector space to create a search that combines a user problem statement and a taxonomical node in the search engine index. The distribution profile score calculation is provenance influenced to derive geographic, cultural, and demographic provenance for a user problem statement.

Abstract: Systems and methods for image searching are described. The systems and methods include receiving a search query comprising user input for a reference image; converting the user input for the reference image to a preference statement using a machine learning model; encoding the preference statement in an embedding space to obtain an encoded preference statement; combining the encoded preference statement with an encoded reference image representing the reference image in the embedding space to obtain a multi-modal search encoding; and performing a search operation using the multi-modal search encoding to retrieve a second image, wherein the second image differs from the reference image based on the user input for the reference image.

Abstract: The present disclosure describes systems and methods for information retrieval. Embodiments of the disclosure provide a color embedding network trained using machine learning techniques to generate embedded color representations for color terms included in a text search query. For example, techniques described herein are used to represent color text in a same space as color embeddings (e.g., an embedding space created by determining a histogram of LAB based colors in a three-dimensional (3D) space). Further, techniques are described for indexing color palettes for all the searchable images in the search space. Accordingly, color terms in a text query are directly converted into a color palette and an image search system can return one or more search images with corresponding color palettes that are relevant to (e.g., within a threshold distance from) the color palette of the text query.

Abstract: Certain aspects of the present disclosure provide techniques for enumerating relationships in an object graph using traversal context objects that are decoupled from the implementation details of an object graph and the objects stored therein. An example method generally includes receiving a request to traverse an object graph, the object graph including a plurality of objects. A root context object is generated, and a traversal context object associated with a root node of the object graph is generated. The traversal context object is linked to the root context object and added to a traversal data structure. The graph is traversed by adding and removing traversal context objects to and from the traversal data structure until a terminating event occurs.

Abstract: A computer-implemented method provides a requested medical data record to a receiving entity. The method includes receiving a set of medical data records and receiving or initiating a joint data index and, for each medical data record of the set, applying a plurality of hash functions to a patient identifier corresponding to the medical data record to determine a hash vector, the patient identifier corresponding to the medical data record identifying the patient being subject of the medical data record, and updating the joint data index based on the hash vector. Furthermore, it includes providing the joint data index to the receiving entity and receiving a request for a record, corresponding to a request patient identifier and being based on the joint data index, from the receiving entity. The method includes providing the requested medical data record to the receiving entity via a secure communication channel.

Abstract: A method and an apparatus for generating a video are provided. The method may include performing, by a server, semantic analysis on an original video according to a timing characteristic of the original video, and segmenting the original video to obtain video segments with semantic information; obtaining, by the server, a video generation sequence model with a timing characteristic based on at least one previously configured video generation sequence model with a timing characteristic according to preference video information obtained from a client; and reorganizing, by the server, the video segments with the semantic information according to the video generation sequence model with the timing characteristic to obtain a target video of the client.

Abstract: Aspects described herein relate to maintaining a dataset with approximate constraints including determining, for a dataset, a constraint collection of tuples that satisfy a constraint and an exception collection of tuples that are an exception to the constraint, constructing, for the dataset, a sharded bitmap of bits, wherein each bit in the sharded bitmap indicates whether a tuple in the dataset is in the exception collection of tuples, wherein the sharded bitmap includes, for each shard of multiple shards, a bitmap of bits and a starting bit location index within the sharded bitmap of bits for the shard, and processing a query on the dataset including processing the constraint collection of tuples and the exception collection of tuples based on the sharded bitmap.

Abstract: Examples of the present disclosure describe systems and methods for query execution across multiple graphs. In an example, a graph or isolated collection may be split into multiple subparts, such that each subpart may store information of the isolated collection. Cross-collection reference resources may be used to reference resources that are stored by other isolated collection subparts. A breadth-first search of an isolated collection subpart may be performed in order to identify matches or potential matches in an isolated collection subpart. In an example, a potential match may comprise a cross-collection reference resource, which may reference a resource in another isolated collection subpart. Once query execution has completed in the isolated collection subpart, query execution may be paused and transferred to another isolated collection subpart that comprises a resource referenced by a cross-collection resource reference.

Abstract: A method and system for persisting data are provided. Batch data is periodically extracted via a computer system from at least one primary data source. Batch data is transformed via the computer system. The batch data is loaded, via the computer system, into a database. Real-time data is continuously received from at least one secondary data source. The received real-time data is indexed in an index. Business intelligence reports are generated from both the real-time data in the index and the batch data in the database.