Abstract: The present disclosure relates to de-duplication. An example method includes acquiring mirror files, wherein each mirror file in the mirror files includes at least one layer. The method further includes generating a directory by de-duplicating layers of the mirror files. The method further includes redirecting the de-duplicated mirror files to the directory by using hard links. The method further includes packaging the redirected mirror files and the directory into a target package. Beneficially, the mirror files can be de-duplicated in a unit of layer, and the flexibility in downloading a part of the mirror files can further be ensured.

Abstract: An embodiment includes creating, by a system, a first dataset and a second dataset from a group of datasets by performing a skyline operation of a feature extracted from the group of datasets. The embodiment includes updating a pipeline metric in a pipeline store by orchestrating an automated modeler based on the first dataset where responsive to the updating an event notification message is transmitted. The embodiment includes detecting that the pipeline metric has been stored in the pipeline store based on the event notification message. The embodiment also includes responsive to the event notification message, creating a search space according to a criterion applied to the pipeline store and orchestrating the automated modeler according to the search space and the second dataset where the system models the first dataset and the second dataset individually and the second dataset is modelled based in part on the pipeline metric.

Abstract: Disclosed are a recommendation method, an apparatus, an electronic device, and a storage medium. The recommendation method includes: generating an item recommendation rank for a user based on a policy network in response to a user request at moment t; obtaining an experience at the moment t based on a feedback result from the user on the item recommendation rank; aggregating the experience of each user at the moment t to a global experience pool; and obtaining an updated policy network based on experiences in the global experience pool.

Abstract: A data processing service facilitates the creation and processing of data processing pipelines that process data processing jobs defined with respect to a set of tasks in a sequence and with data dependencies associated with each separate task such that the output from one task is used as input for a subsequent task. In various embodiments, the set of tasks include at least one cleanroom task that is executed in a cleanroom station and at least one non-cleanroom task executed in an execution environment of a user where each task is configured to read one or more input datasets and transform the one or more input datasets into one or more output datasets.

Abstract: In one aspect, a computerized method for heuristic-based classification of software discovered connected devices, comprising: implementing a discovery process on a set of devices and obtain a set of attributes of each device; identify a set of heuristics, wherein the set of heuristics combines an attribute from each discovered asset of the set of devices with an associated weight; comparing the attribute from each asset and the device catalog and using the weight to determine a partial match score; using the set of heuristics to reduce a search space used to identify the candidates for a friendly name of each device; and calculating a dot product of the weight of each of the heuristics and using the dot product to identify a set of candidate devices in the device catalog.

Abstract: A multi-layer federated learning method based on distributed clustering is provided, which comprises the following steps. Computing a feature distribution similarity for each of participating nodes with non-(non-independent and identically distributed) data sets, and grouping these nodes into plural clusters by the feature distribution similarity. Updating local model of nodes of each cluster by a federated learning algorithm, and inputting these nodes into a multi-layer aggregation mechanism. Terminating the operation of the multi-layer aggregation mechanism until the clustering result meets a required demand. Furthermore, we implement a blockchain-based multi-layer federated learning system, including model aggregation module, API module, time-series synchronization module, and IPFS, based on distributed clustering architecture. The learning performance is proven to effectively improved.

Abstract: The technology disclosed relates to splice site prediction and aberrant splicing detection. In particular, it relates to a splice site predictor that includes a convolutional neural network trained on training examples of donor splice sites, acceptor splice sites, and non-splicing sites. An input stage of the convolutional neural network feeds an input sequence of nucleotides for evaluation of target nucleotides in the input sequence. An output stage of the convolutional neural network translates analysis by the convolutional neural network into classification scores for likelihoods that each of the target nucleotides is a donor splice site, an acceptor splice site, and a non-splicing site.

Abstract: Methods, systems, and techniques for metadata determination and storage. A large language model that is implemented using at least one artificial neural network receives an initial prompt that includes a query related to the metadata. The metadata is in respect of data that is part of a dataset, and the initial prompt includes context for the query. The large language model determines the metadata in response to the query using the context. Once determined, the metadata is stored in the dataset such that the metadata is associated with the data to which it relates.

Abstract: The disclosure relates to methods and systems of match classification based on ensemble modeling. Match classification based on ensemble modeling is a prediction that two or more data records match or mismatch one another based on the output of two or more classifiers trained to make this prediction. A first match classifier may include an ensemble of decision trees, which may be trained via gradient boosting, and in particular using extreme gradient boosting to generate a first match classification. The first match classifier, which is itself an ensembled model, may be ensembled together with at least a second match classifier trained via weak supervision to generate a second match classification, which may be aggregated with the first match classification. The aggregated classification may be used to determine whether or not two or more match candidates match one another.

Abstract: A system and method for providing a configuration for data extraction from an automation system includes a signal selection agent configured to: receive a user selection of at least one module of the automation system; generate, for display to the user, a user interface identifying one or more selectable signals associated with the selected module and displaying one or more guidance elements comprising data mined from data sources pertaining to the automation system for guiding the user in the selection of relevant signals; receive a user selection of one or more of the selectable signals; and automatically generate the configuration for data extraction on the basis of the selected signals.

Abstract: An electronic device constituting a blockchain node included in a blockchain network is provided.

Abstract: An operation method of a node device connected to a non-random consensus body proof-based blockchain network according an embodiment of the present disclosure comprises acquiring new block data propagated through the blockchain network; and performing a neural consensus body proof-based block generation process corresponding to the new block data according to a preset condition, wherein the neural consensus body proof-based block generation process includes extracting validity verification data from the new block data; acquiring neural consensus body designation information of a next block generated on the basis of a random consensus body proof process according to a verification process of the validity verification data; and generating validity verification data of the next block by selectively driving a consensus body node function unit on the basis of the neural consensus body designation information of the next block.

Abstract: This disclosure relates to a system and method for data analysis. According to a first aspect, there is described a method, the method being performed using one or more processors, comprising: receiving one or more user inputs indicative of one or more relationships between data in a plurality of datasets; determining, based on the one or more user inputs, at least one object view for visualizing the data in the plurality of datasets; generating, based on the one or more user inputs, metadata comprising: an object graph indicative of the one or more relationships between two or more of the plurality of datasets; and information identifying the at least one object view; and in response to a query relating to the plurality of datasets, using the metadata to determine how response data responding to the query should be provided.

Abstract: A processing client may receive a dataset replacement indication that indicates an upcoming synchronized data replacement in which processing clients are to synchronously replace a first synchronized dataset with a second synchronized dataset as well as a request period during which the processing clients are to request, at randomized times, the second synchronized dataset. The processing client may transmit a query requesting the second synchronized dataset. The processing client may receive a response including the second synchronized dataset and an indication of an expiration time for the first synchronized dataset that indicates when the processing clients are to replace the first synchronized dataset. The processing client may store, at the processing client prior to the expiration time, the second synchronized dataset and the indication of the expiration time.

Abstract: A directory metadata operation method, including: in response to receiving a directory query request, obtaining a query path carried therein; obtaining a prebuilt prefix directory data table and a prebuilt full load directory data table, where a key of the prefix directory data table is a prefix path of the directory metadata, which is a subpath corresponding to a hierarchy other than a default hierarchy counting from end in a path corresponding to the directory metadata, a value of the prefix directory data table is the directory metadata, a key of the full load directory data table is identification information of the directory metadata and a value thereof is the directory metadata; determining a matching result between the query path and the key of the prefix directory data table; and determining a query result based on the matching result and the full load directory data table.

Abstract: Disclosed are some implementations of systems, apparatus, methods and computer program products for synchronizing data. A source device processes an update to data in a database. The source device transmits, via a message bus, a first event message pertaining to the update, the first event message having an associated indicator. A target device accessing the message bus detects the indicator. Responsive to detecting the indicator, the target device skips the first event message on the message bus and identifies a snapshot link in a second event message subsequent to the first event message. The target device accesses a snapshot event identified by the snapshot link, stores data of the snapshot event, and processes one or more event messages subsequent to the snapshot event.

Abstract: A system described herein may maintain a blockchain in conjunction with a plurality of other devices. The system may receive a first request to perform a first set of operations with respect to the blockchain, and may process the first request. Processing the first request may include performing the first set of operations to generate a first result set. Prior to completing processing of the first request, the system may receive a second request to perform a second set of operations with respect to the blockchain, and a third request to perform a third set of operations with respect to the blockchain. After completing processing of the first request, the system may batch process the second and third requests. Batch processing the second and third requests may include performing the second set of operations and the third set of operations to generate a second result set.

Abstract: A data organizer, optimizing, reconciliation system (DOORS)/method/program, provides capability to merge intuitively datasets created over users' digital lifetimes from multiple devices. Finds and gathers files from devices and may create master copies in Cloud or local storage. Enhanced functionality via artificial intelligence or machine learning provides more intuitive merging of datasets including file and folder name selection and may operate autonomously at times of low usage. Optimized by learning from training data, or legacy data, or learning in real time from user preferences, practices, and habits. Reducing loses of treasured photographs and important documents via preventing files getting marooned on inaccessible legacy devices. Because devices get lost, replaced, or may suffer ransomware attacks or may need to be re-set and suffer data loss for a variety of reasons.

Abstract: The present disclosure discloses the system and method for efficient file storage and management. The method includes collecting information on access of different files by different applications on a master storage. The collected information is then normalized within a predetermined range. Additionally, the method includes training a machine learning model with the normalized information to predict which of the files should be stored in a cache. The cache is ten implemented by storing predicted files in the cache. Consequently, the different applications then access one of the predicted files in the cache without making multiple copies of the predicted files.

Abstract: Disclosed are a method for improving identification accuracy of similar logistics tracking numbers, and a logistics query system. The method includes: acquiring the logistics tracking number, calculating a length of the logistics tracking number, screening the logistics channel with a same length of the logistics tracking number from a rule library according to the length of the logistics tracking number, recording all the logistics channels with the same length and calculating number N of the logistics channels; performing preliminary identification of the logistics tracking number according to the number N, the preliminary identification of the logistics tracking number includes: when N=1, using the logistics channel as an identification result of the logistics tracking number; when N?2, calculating matching precision of the logistics tracking number; and configuring the logistics channel with the highest matching precision value as an identification result of the logistics tracking number.