Abstract: In a method for AI model optimization, an optimization device receives an original AI model and search configuration information that comprises a plurality of search items each indicating its search categories for performing optimization information search on the original AI model. The device obtains a plurality of search operators corresponding to the plurality of search items, and arranges the search operators in an operation sequence based on the search configuration information. The device then executes the search operators in the arranged operation sequence on the original AI model to obtain an optimized AI model. In the execution of the operation sequence, each search operator, except for the first search operator in the operation sequence, is executed utilizing operation results of a preceding search operator in the operation sequence, the operation results including generated network structures and search space information.

Abstract: An audio file associated with a user voice query may be received at a user device. The audio file may be compared to a plurality of references, such as cache entries, corresponding to a plurality of other voice queries. Based on a determination that the voice query corresponds to one of the references, an operation associated with the voice query may be executed. An indication may be received that the operation was not an intended operation associated with the voice query. Based on receiving this indication, the incorrectly identified operation, associated reference, e.g., voice query, may be disabled for the user or the device. However, the cache entry may remain enabled for one or more of a plurality of other devices.

Abstract: Provided is a method of detecting a semantics section in a video. The method includes extracting all video features by inputting an inputted video to a pre-trained first deep neural network algorithm, extracting a query sentence feature by inputting an inputted query sentence to a pre-trained second deep neural network algorithm, generating video-query relation integration feature information in which all of the video features and the query sentence feature have been integrated by inputting all of the video features and the query sentence feature to a plurality of scaled-dot product attention layers, and estimating a video segment corresponding to the query sentence in the video based on the video-query relation integration feature information.

Abstract: A system enables analysis of retroactively changing datasets and/or of various versions of logics. In an example, the system determines a first version of data and/or a first version of a logic and a second version of the data and/or a second version of the logic, wherein at least some of the second version of the data was retroactively added. The system determines two outputs each derived from one of a) applying the first version of the logic to the first version of the data, b) applying the second version of the logic to the first version of the data, c) applying the first version of the logic to the second version of the data, or d) applying the second version of the logic to the second version of the data. The system compares the outputs and determines one or more differences between the outputs.

Abstract: A data processing method and apparatus are described. The data processing apparatus obtains an input tensor corresponding to input data. The data processing apparatus determines M1 first-type tensor blocks and M2 second-type tensor blocks. P processing units in the data processing apparatus process the M tensor blocks concurrently. In a first time period, all of the tensor blocks that are processed concurrently by the P processing units are first-type tensor blocks. In a second time period, all of the tensor blocks that are processed concurrently by the P processing units are second-type tensor blocks.

Abstract: Upon completing replicating a file set from a source to destination, a snapshot taken of the file set is maintained at the source. The file set includes a namespace file having metadata and other files having content data. Verification is started on the file set replicated to the destination. While the verification is in-progress, a next replication cycle is started. Upon detecting corruption in the namespace file, the next replication cycle is paused and any changes to the file set are rolled back. The snapshot being maintained at the source is fetched. The namespace file having the corruption is replaced with the namespace file from the snapshot being maintained at the source. Other files in the file set having the content data are not replaced.

Abstract: A cascading search system includes an associative memory array, a similarity match processor and an exact match processor. The columns of the array store a plurality of multiportion data vectors and have a first section, for a first portion of a vector, a second section for storing a second portion of a vector and a match row. The similarity match processor performs a parallel similarity search of a similarity query in the first sections and stores a match bit indication in the match row of the column. Each match bit indication indicates if its column has a first portion which matches the similarity query. The exact match processor performs an exact search in parallel in the second section of each similarity matched column whose match bit indication indicates a match of its first section and outputs those similarity matched columns whose second portions match the exact query.

Abstract: Methods, apparatus and computer program product for generating tiered search index fields based on a divided group-based communication data corpus in a group-based communication platform are described herein. In some embodiments, the system provides for receiving a group-based communication data corpus, generating a retrieval score, and assigning each group-based communication data object associated with a retrieval score. Each group-based communication data object may meet or exceed a retrieval score threshold of a high retrieval probability corpus. Each group-based communication data object associated with a retrieval score below the retrieval score threshold may be assigned to a low retrieval probability corpus. High and a low retrieval probability search index fields may be generated and associated with the high and low retrieval probability corpus.

Abstract: A copy-on-write union filesystem. A computer system identifies a first filesystem namespace as a mutable layer, and a second filesystem namespace as an immutable layer. The computer system exposes a merged filesystem namespace as a union filesystem that merges the mutable layer and the immutable layer, and that overlays the first filesystem namespace over the second filesystem namespace. The computer system identifies an operation on the merged filesystem namespace that requests to read from a file within the merged filesystem namespace, the file being stored in the second filesystem namespace and not being stored in the first filesystem namespace. The computer system returns a handle referencing a proxy object stored in a system memory. The proxy object includes a mapping to a file object corresponding to the file as stored in the second filesystem namespace.

Abstract: Embodiments of the present invention are directed to facilitating data preprocessing for machine learning. In accordance with aspects of the present disclosure, a training set of data is accessed. A preprocessing query specifying a set of preprocessing parameter values that indicate a manner in which to preprocess the training set of data is received. Based on the preprocessing query, a preprocessing operation is performed to preprocess the training set of data in accordance with the set of preprocessing parameter values to obtain a set of preprocessed data. The set of preprocessed data can be provided for presentation as a preview. Based on an acceptance of the set of preprocessed data, the set of preprocessed data is used to train a machine learning model that can be subsequently used to predict data.

Abstract: Data is ingested from a source system. Ingesting the data includes determining corresponding chunk identifiers for a plurality of data chunks corresponding to the ingested data and for each of the plurality of data chunks, verifying whether the corresponding chunk identifier is included in a data structure tracking identifiers of data chunks that were already stored in a storage of a storage system before the data ingestion started and storing the data chunk in a storage based on the verification. After the ingesting is completed, deduplication of the ingested data chunks stored in the storage having a same chunk identifier is performed and the data structure is updated based on the deduplication.

Abstract: Methods for enhancing compatibility of a document of an entity with an organization's database on a computer server. Methods may include using a computer hardware processor to digitize a document from a first format into a digital format, such as bytes, when the first format may not be compatible with the database. Methods may further include using a computer hardware processor to convert the document from a digital format into a second format, where the second format of the document may be compatible with the organization's database. Methods may include using a computer hardware processor to populate the database on the computer server with data from the document in the second format. Methods may further include storing the populated database and the document in the second format on the computer server.

Abstract: A system includes hardware processors and a non-transitory memory storing instructions that, when executed by the hardware processors, causes the hardware processors to perform actions including receiving an input data set related to digital content, where the input data set includes a plurality of input entries. The actions also include matching one or more input words of each input entry of the plurality of input entries to one or more baseline words of one or more baseline entries of a baseline data set and generating an output data set including a plurality of output entries. Each respective input entry corresponds to a respective output entry of the plurality of output entries, and each respective output entry includes a baseline entry of the one or more baseline entries having a highest probability of matching the respective input entry and additional data associated with the respective input entry.

Abstract: A method and apparatus are provided for avoiding target database constraint violations in a database replication system. The database replication system includes a source database and a target database, wherein target database transactions automatically abort due to constraint violations. Each database has one or more tables and at least one of the target database tables has one or more constraints. The database replication system replicates source database transactions. For each target database transaction corresponding to a replicated source database transaction, changes made at the source database are replicated to the target database. When it is detected that a replicated change causes a constraint violation as a result of one or more of the target database tables having one or more constraints, the applying of the database change to the target database of the database change is deferred and the database change is posted to an electronic list.

Abstract: A system may include one or more tangible, non-transitory computer-readable media having stored thereon instructions, the instructions when executed to instruct one or more processors to: receive a platform-independent access binding data structure (a PIAB data structure), the PIAB data structure comprising a platform-independent representation of a subject class, a verb, an object, and a binding comprising a permission for the subject class to perform the verb on the object; convert the PIAB data structure to a first platform-specific binding for a first target platform; and cause the first platform-specific binding to be implemented on the first target platform.

Abstract: A system includes a controller, a database controlled by the controller, and a memory including an end-user executable code and a business software management code. The controller receives a user update to modify the end-user executable code, converts the user update to a machine language input, and, in response to the machine language input, modifies one or more of the end-user executable code, the database structure, and a documentation of the end-user executable code. In response to executing the program instructions associated with the modified end-user executable code on the end-user device, the controller generates a data collection GUI according to the modified end-user executable code and stores data in the database structure according to the modified end-user executable code.

Abstract: An information processing apparatus (100) according to the present disclosure includes an acquiring unit (132) that uses a trained encoder to acquire a feature value of each set of partial data forming first content; a calculating unit (133) that calculates a relative feature value that is a relative feature value between sets of partial data from the acquired feature values of the sets of partial data to calculate a relative feature value sequence indicating a feature of a structure of the first content; and a generating unit (134) that generates second content based on the relative feature value sequence of the first content and a feature value of optional data.

Abstract: In embodiments of field value search drill down, a search system exposes a search interface that displays one or more events returned as a search result set. A field-value pair can be emphasized in the field-value pairs of an event displayed in the search interface, and a menu is displayed with search options that are selectable to operate on the emphasized field-value pair of the event. The menu includes the search options to add search criteria of the emphasized field-value pair to a search command in a search bar of the search interface, exclude the search criteria of the emphasized field-value pair from a search, or create a new data search based on the emphasized field-value pair. A selection of one of the search options in the menu can be received, and the search command in the search bar is updated based on the search option that is selected.

Abstract: Providing an Artificial Intelligence (AI) and Internet of Things (IoT) based system and method that predicts the chronological requirements for various components of the file-being-stored, and then takes an appropriate storage action on each component based on the predicted chronological requirements.

Abstract: An embodiment of a system for determining the disposition of a hyperdocument and hyperdocument requests using a trained artificial neural network, including an information source, a requesting application, and a server containing a trained artificial neural network (ANN), the ANN being capable of evaluating the information, and providing results reflecting the evaluation to a requesting application is described.