19 results about "Data abstraction" patented technology

The invention discloses an equipment manufacturing energy consumption and state visual monitoring method based on an industrial internet of things, solves the problem of passive fixation of an industrial monitoring visualization technology in the prior art, and realizes autonomous and intelligent generation of an optimal visual view matched with a real-time working condition. The method comprises the following steps: collecting industrial field multi-source heterogeneous data in real time, abstracting a monitoring object as an entity node, and constructing a feature vector containing static and dynamic attributes; constructing and updating a dynamic knowledge graph based on nodes and features, wherein relation edges of the dynamic knowledge graph comprise dynamic edges generated by static connection and real-time analysis; inputting the atlas into a pre-trained multi-attention graph neural network to obtain a display priority score and a visual affinity matrix of each node; calculating coordinates of all nodes on an interface through a layout optimization algorithm in combination with the priority and the affinity matrix; and generating a visual drawing instruction according to the coordinates and the node attributes, and driving a display device to render a monitoring interface.

A system, method, and apparatus for generating a data visualization are disclosed. The method includes (i) obtaining a natural language query, (ii) determining an intent of the natural language query, (iii) generating one or more data requests to selected one or more data sources, the one or more data requests based at least in part on the intent, (iv) abstracting result data to obtain a data abstraction, and (iv) generating one or more data requests to selected one or more data sources based at least in part on the data abstraction. The result data is responsive to the one or more data requests, and (v) generating a visualization for the result data based at least in part on the data abstraction.

The invention discloses a big data management method and system based on semantic analysis, and relates to the technical field of big data management. A multilayer semantic abstract data organization structure comprising a base layer, a mode layer and a statistical layer is constructed, real-time semantic analysis is performed on input data, and semantic descriptors are generated; the high-frequency data instance in the base layer is stored to a corresponding level according to a central semantic index, a semantic folding operation is triggered based on a preset rule, the high-frequency data instance in the base layer is automatically abstracted into a modeled representation of a mode layer, and similar representations in the mode layer are further aggregated into an aggregated representation of a statistical layer; when the query is responded, the system analyzes a query intention and determines a target data abstraction level, high-level aggregation or modeling representation is restored into bottom-layer detailed data as required through reverse expansion operation, and a query result is generated; through the multi-layer semantic abstract data organization structure and the data folding operation, the storage efficiency of mass complex data and the performance of multi-granularity intelligent query are effectively improved.

A computer-implemented method according to one embodiment includes creating a data abstraction that includes a set of memory resources and variables associated with the memory resources. The computer-implemented method also includes implementing the data abstraction within a programming language environment. A system according to another embodiment, includes a processor. The system also includes logic that is integrated with the processor, executable by the processor, or integrated with and executable by the processor. Moreover, the logic is configured to: create a data abstraction that includes a set of memory resources and variables associated with the memory resources. The logic is further configured to: implement the data abstraction within a programming language environment.

The embodiment of the invention provides a recursive query processing method and a related device, and relates to the technical field of big data, the method comprises the following steps: when a statement to be processed is determined to be a recursive query statement, extracting a table name and a recursive condition in the recursive query statement; according to the table name and the recursion condition, constructing a directed graph, and based on the directed graph, obtaining the maximum depth of nodes and the total number of paths; splitting the directed graph into a plurality of sub-graphs according to the maximum depth and the total number of paths; and determining a link from a starting node to a target node according to the plurality of sub-graphs, and outputting the link. According to the method, the recursive query table data is abstracted into the directed graph, so that the success rate and efficiency of recursive query can be improved.

The invention relates to the technical field of processors, in particular to a tile block instruction set architecture and a processing method, and the tile block instruction set architecture comprises a tile block parallel processing system control instruction, a tile block descriptor register configuration instruction, a tile block data handling instruction, a tile block data copying instruction and a tile block arithmetic instruction. According to the tile block instruction set architecture processing method, one or more tile block operation units, one or more tile block storage management units and one or more tile block calculation task scheduling and synchronization units are combined into a tile block parallel processing system, so that tile block instructions are allowed to be processed in parallel, and tile block-based data operation is completed. The architecture aims at solving the problems that when an existing processing system processes two-dimensional data blocks, instruction overhead is too large, data abstraction is insufficient, and the coupling degree of software and hardware is too high.

The invention discloses a user consumption behavior prediction method under cross-tenant privacy protection and a related product. After a consumption prediction request of a tenant for a target user is received, interaction data and current geographic position data of the target user are firstly obtained. Converting the interaction data into a first semantic note for describing the consumption motivation of the user by using a large language model, and abstracting the geographic position data into a regional commercial attribute tag which does not contain specific street information to obtain a first tag; vectorizing and splicing the first semantic note and the first tag to form a target feature vector; and then a reference vector with high similarity with the target feature vector is retrieved in a public vector database, group features associated with the reference vector are combined, reasoning analysis is performed through a large language model, and finally a consumption intention prediction result of the target user is obtained. According to the invention, on the premise of ensuring the privacy security of the user, the cross-brand data can be effectively utilized to carry out precision marketing.

The invention relates to a platform data file distributed management method and system. The invention aims to solve the problems of load imbalance, low data reliability, difficulty in file management and the like in traditional distributed file storage. The management method is realized through a distributed development kit, and specifically comprises the following steps of: segmenting a file into a plurality of blocks, and storing the file blocks on different servers; performing multiple backups on each file block to enhance the reliability of the data; abstracting metadata of the file into a directory tree for management so as to simplify file management operation; a unified file management function is provided for the outside through the distributed development kit, and a user only needs to interact with the kit and does not need to care about bottom storage details. The method has the advantages of being high in fault tolerance, suitable for batch processing, suitable for storing large files and the like, can run on a low-cost machine, effectively reduces hardware cost, and meanwhile improves the concurrent access capacity of a system and the high availability of data. The management system comprises a file segmentation module, a file storage module, a backup module, a metadata management module, a distributed development kit and the like so as to realize the management method.

This invention relates to the field of multidimensional data processing technology. It discloses a method and system for constructing a multidimensional data model for thermal power equipment. The method includes: collecting basic and indicator data of the thermal power equipment; using machine learning algorithms to identify data relationships between different business domains and constructing a multi-level data abstract model; defining the hierarchical structure and specific attributes in the data model, as well as the relationships between each layer; and dynamically updating the data relationships to adapt to changes in the data source, forming a data model that supports the entire lifecycle management of the equipment. This method combines advanced machine learning algorithms to identify data relationships between different business domains to support comprehensive equipment management.

The application belongs to the fields of artificial intelligence, computer vision, natural language processing and the like, and is a document-level image-text review sentiment classification method fusing common sense knowledge. First, review data collected from the Internet is abstracted into structured data, and after data preprocessing work such as Chinese word segmentation and image normalization, the feature representation of the review data is obtained by sequentially passing through single-modal sentiment feature representation learning and multi-modal sentiment feature fusion, and finally the fusion features are input into a classifier or a multi-layer perceptron is directly used to obtain a sentiment probability distribution. The method can effectively improve the accuracy of sentiment classification on the one hand by fusing common sense knowledge to strengthen text semantic understanding, and on the other hand by introducing related image information to assist in improving text sentiment expression.

The application discloses a kind of software defect automatic positioning methods based on heterogeneous graph representation learning.The method is first obtained from the required data in project, by abstracting various data into different types of vertex and edge, constructs a heterogeneous graph;Then the information in heterogeneous graph is aggregated by three modules, and the feature representation of each method in source program is obtained;Finally, a prediction model is constructed based on convolutional neural networks (CNN) to calculate the probability of defects existing in each method, and the defect positioning result is obtained.At the same time, the application introduces a resampling strategy, which effectively overcomes the class imbalance problem in the problem.Through the method of the application, the effectiveness of feature learning and the accuracy of defect positioning can be effectively improved, which has very important practical significance for reducing software development and maintenance cost.

The application relates to the technical field of processors, in particular to a tile block instruction set architecture and a processing method, wherein the tile block instruction set architecture comprises a tile block parallel processing system control instruction, a tile block descriptor register configuration instruction, a tile block data carrying instruction, a tile block data copying instruction and a tile block arithmetic instruction; the tile block instruction set architecture processing method is to combine one or more tile block operation units, one or more tile block storage management units and one or more tile block computing task scheduling and synchronization units into a tile block parallel processing system, so as to allow parallel processing of tile block instructions and complete tile block-based data operation; the architecture aims to solve the problems of excessive instruction overhead, insufficient data abstraction and excessively high coupling degree of software and hardware of an existing processing system when processing a two-dimensional data block.

The invention discloses a distributed soft bus communication method and system based on multi-hop path matrix optimization, and belongs to the technical field of communication, and the method comprises the following steps: obtaining multi-dimensional link state information through a lightweight detection message, and forming structured link measurement data; abstracting link measurement data into a cost vector, and mapping the cost vector into a matrix element in a weighted combination manner, so as to obtain a computable multi-dimensional cost matrix; cost function weights in the multi-dimensional cost matrix are set according to different service scenes, an optimal path is calculated by using an improved Dijkstra algorithm or an incremental graph algorithm, a stability control mechanism is introduced, and a result is dynamically updated; and an optimal path result is injected into the SoftBus routing layer through an interface and is bound by the session management layer, and inter-application communication is established. According to the method, conversion from a single index to multi-target constraint is realized in a path selection link, and the communication efficiency and reliability of the distributed soft bus are improved on the premise of keeping transparency.

The present application relates to a kind of equipment time series data storage model construction method and system suitable for coal mine multi-disaster fusion analysis, belong to coal mine safety disaster analysis technical field, to solve the problem of data effectiveness, history deficiency in prior art.The method first abstracts equipment time series data into point type data mode and line type data mode, and constructs the data storage conceptual model including monitoring time series data model, alarm time series data model, personnel positioning monitoring data model, monitoring time series slice data model and alarm time series slice data model.Database system using NoSQL technology, design the database table into conceptual model, and create the corresponding wide table structure, take column-based storage mode, set the compression strategy of column, realize long-term online storage and efficient reading of data.The equipment time series data storage model constructed by the present application can effectively reduce the complexity of data access, reduce data load and operation amount.

The embodiment of the invention provides a robot learning data processing method and device based on human behavior data, and the method comprises the steps: converting multi-source heterogeneous human behavior data into a unified intermediate behavior representation which describes a human operation intention through a key point pose; based on the kinematic model and the kinetic model of the target robot body, the intermediate behavior representation is mapped into a robot action sequence, and the robot action sequence meets the joint limit, moment constraint and contact stability requirements of the target robot body. By abstracting human behavior data into intermediate behavior representation irrelevant to a specific acquisition mode, unified processing of multi-source heterogeneous data is realized, and bidirectional mapping from human intention to robot action is realized based on kinematics and dynamics models. Through decoupling data acquisition, representation and mapping processes, the problems of data splitting, domain difference and insufficient generalization ability caused by body difference between human behavior data and a robot body in the prior art are solved.

An unstructured data storage and retrieval method for a large-scale complex object belongs to the field of databases, and comprises the step of abstractly dividing unstructured data of heterogeneous objects into three types for storage. Extracting object retrievable information from the object native data based on a meta-information model, and performing data conversion and encoding to generate object retrieval data; carrying out database and table division according to object categories, dividing into a common attribute retrieval group, a plurality of hidden attribute retrieval groups, a keyword retrieval group and an auxiliary retrieval column, and establishing a corresponding index structure; semantic extension is carried out based on a standard SQL grammatical structure, and a mixed retrieval mode is achieved. According to the method, in-database integrated storage and retrieval based on the database are realized, unnecessary data migration and synchronization overhead are avoided, data consistency is improved, system deployment complexity and operation and maintenance cost are greatly reduced, meanwhile, efficient sharing and exchange of large-scale heterogeneous diversified data are supported, and the method is suitable for popularization and application. And the requirements of AI, digital twinning and other intelligent applications on a data surface are met.

The invention relates to the technical field of digital twinning, in particular to an implementation method of a digital twinning data modular interface, which comprises the following steps of: constructing a unified data abstraction layer and defining an access descriptor; dynamically loading the analysis module and standardizing the data flow; data distribution based on a standardized bus; and carrying out space-time alignment and fusion mapping on the multi-source heterogeneous data. The unified data abstraction layer and the pluggable analysis module are constructed, plug and play of the sensor and standardized processing of heterogeneous data are achieved, hardware strong coupling is thoroughly eliminated, and the development threshold is greatly reduced. Meanwhile, in combination with a space-time alignment and cascade fusion strategy, the problem of accurate matching of multi-modal data is solved, and the mapping accuracy and real-time performance of the twin are remarkably improved. In addition, the bus-based asynchronous decoupling architecture effectively enhances the expansibility of the system.

A system, method, and device for generating data visualizations are disclosed. The method includes (i) obtaining a natural language query, (ii) determining an intent for the natural language query, (iii) generating one or more data requests to one or more selected data sources, the one or more data requests being based at least in part on the intent, (iv) abstracting result data to obtain a data abstraction, the result data being responsive to the one or more data requests, and (v) generating a visualization for the result data based at least in part on the data abstraction.

The invention discloses an open world standardization processing method and system, and the method comprises the following steps: S1, dividing an open world scene into a plurality of logic tiles, abstracting all to-be-managed scene objects into a unified basic data structure, and forming tile data associated with the logic tiles; s2, in a working thread independent of a main thread, based on the tile data, executing hierarchical loading, unloading management and integrated data rejection for the logic tiles; and S3, synchronizing the rendering data processed by the working thread to a main thread, and calling a graphic engine interface by the main thread to instantiate and render. According to the application, the load of a main thread is remarkably reduced, a unified data abstraction and tile management framework is realized on the architecture, different modules such as physical modules, illumination modules and audio modules can be accessed in the same normal form, redundant codes are eliminated, and the system complexity is reduced.