31results about "Database indexing" patented technology

A cloud computing system is described herein that enables fast processing of queries over massive amounts of stored data. The system is characterized by the ability to scan tens of billions of data items and to perform aggregate calculations like counts, sums, and averages in real-time (less than three seconds). Ad hoc queries are supported including grouping, sorting, and filtering without the need to predefine queries by providing highly efficient loading and processing of data items across an arbitrarily large number of processors. The system does not require any fixed schema, thus the system supports any type of data. Calculations made to satisfy a query may be distributed across a large number of processors to parallelize the work. In addition, an optimal blob size for storing multiple serialized data items is determined, and existing blobs that are too large or too small are proactively redistributed or coalesced to increase performance.

A message transfer system includes an interface to a send agent, arranged to facilitate an input of an at least one message to the message transfer system, an interface to a delivery agent, arranged to facilitate an output of the at least one message from the message transfer system, a message queue, arranged to facilitate a processing of the at least one message through the message transfer system between the interface to a send agent and the interface to a delivery agent, and a message transfer node, arranged to include the message queue such that the message queue is arranged as distributed.

An information storage medium is disclosed. The information storage medium includes one container file, wherein the container file includes a data box including in which a plurality of component files forming content are arranged; and a meta box including information on locations in the content and in the data box with respect to the plurality of component files and characteristic information on of the plurality of component files, and the plurality of component files are arranged in the data box according to the characteristic information.

The invention relates to a flight operation data fusion method and device, and belongs to the technical field of data fusion, the method can accurately fuse various flight operation data belonging tothe same flight plan, and a designed association primary key and the fusion method can well process data splitting easily caused by flight delays and other abnormal conditions; various flight operation data of an air traffic control production system can be received and processed in real time, a latest data fusion result is generated immediately after the production system sends out the data, andcentralized processing does not need to be carried out in the future; according to the invention, various flight operation data generated by a plurality of air traffic control production systems can be fused, the processed data is more comprehensive, and the application range of a fusion result is wider.

The embodiment of the invention discloses a data distribution method and device, equipment and a storage medium, and the method comprises: determining the number of slots and at least two hash valuesof a hash distribution table; allocating nodes to each hash value, and storing the number of slots and the distribution mapping relationship between each hash value in the hash distribution table andthe nodes, so as to distribute the data of each hash value in the corresponding node, the distribution mapping relationship being non-fixed; and after the newly added nodes are added, performing dataredistribution based on a preset redistribution mode and updating the corresponding distribution mapping relationship. By the adoption of the technical scheme, logic and physical separation of data distribution is achieved by storing the distribution mapping relation between the non-fixed hash values and the nodes in the MPP system, when new nodes are added into the MPP system, data redistributioncan be achieved more efficiently, and then rapid system expansion is achieved.

The invention relates to an automatic comparison method and device for RPS software versions of a nuclear power station, belongs to the technical field of industrial control software configuration, and solves the problems that an existing method is low in comparison efficiency and poor in accuracy. The method comprises the steps: obtaining two engineering files of different versions to be compared; respectively analyzing the two engineering files to obtain corresponding engineering information, wherein the engineering information comprises equipment configuration information, algorithm configuration information, variable configuration information and picture configuration information; and respectively comparing whether the equipment configuration information, the algorithm configuration information, the variable configuration information and the picture configuration information are correspondingly consistent or not, generating a report according to a comparison result, and outputtingthe report to a corresponding position. According to the invention, the comparison efficiency and accuracy between different versions of engineering files are improved.

The invention discloses a state data storage method and device and a data sending method and device, and the method comprises the steps: obtaining to-be-stored switch state data which comprises a target switch state set by a user and a user identifier; and storing the switch state data by adopting a bitmap, the offset of the bitmap pointing to the user identifier, and the bit where the offset is located being used for storing the target switch state. According to the method and the device, the storage space of the switch state data can be greatly reduced, and the larger the data size is, the more obvious the effect of saving the storage space is.

The invention provides an indexing method and system for a concurrent hash index data structure based on machine learning. The indexing method comprises the steps of M1, searching a group node where atarget key value pair is located for from a root node according to a key in a request; M2, in the group nodes obtained through positioning, calculating data positions according to a machine learningmodel of the group nodes; and M3, searching target data according to the data position, and carrying out corresponding operation according to the request type. Compared with an existing machine learning hash index, the system can retrain the machine learning model, hash expansion operation is carried out when the hash conflict probability exceeds a threshold value, the machine learning model is made to adapt to newly inserted data, and the high performance of the index is kept.

The embodiment of the invention provides a vending machine updating method and device and a vending machine, and belongs to the technical field of automatic vending. The updating method comprises the steps of identifying the newly-added commodities located on a fixed commodity channel automatically, and obtaining the feature data and a prediction category P of the newly-added commodities; obtaining a label category L of a newly added commodity which is manually labeled and is located on the fixed goods channel; and according to a comparison result of the annotation category L and the prediction category P, determining whether to update the annotation category L and the feature data of the newly added commodity to an index library. According to the above technical scheme, the image collection of the newly-added commodities and the acquisition of the feature data of the related newly-added commodities are directly carried out through the equipment of a vending machine or a goods selling shelf, the accuracy of the recognition result can be improved, and the method can be directly executed in a local mode without updating an index library after offline acquisition.

The invention discloses a data storage method and device, equipment and a medium, and belongs to the technical field of data processing. The method comprises the steps that under the condition that a storage mode subscription service is started, if it is monitored that a data consumer consumes target data, a target storage mode associated with the target data is selected from candidate storage modes; and storing the target data by adopting the target storage mode. By means of the technical scheme, differential storage of mass data is achieved.

Data processing apparatus, method, and non-transitory tangible machine-readable medium thereof for recursively structured data are provided. The data processing apparatus accesses a piece of recursively structured data having a Location Independent Structure and arranges the piece of recursively structured data in a one-dimensional memory space according to LIS:=e⁰∘e¹∘ . . . ∘e^p−1∘LIS⁰∘LIS¹∘ . . . ∘LIS^p−1. The notation eⁱ represents an element of the piece of recursively structured data. The notation LISⁱ represents a sub-LIS of the recursively structured data and corresponds to the element eⁱ. The operator ‘∘’ indicates “contiguous to” in the one-dimensional memory space. When the elements in all depths are of the element size and the LIS sizes of the sub-LISs within the same depth are the same, the piece of recursively structured data is a complete LIS.

The invention discloses a database dictionary generation method and device, equipment and a storage medium, and the method comprises the steps: receiving data source information configured by a user,and the data source information comprising at least two pieces of address information; establishing at least two connections with at least two databases according to the at least two pieces of addressinformation; wherein the types of the databases in the at least two databases are different; acquiring data structure information in the at least two databases through the at least two connections; and generating a database dictionary according to the data structure information. According to the method, the database fields can be generated for multiple different types of databases at the same time, and the generation efficiency of the database dictionary is improved.

The invention discloses a telemetric sample data definition monitoring method for a scheduling energy management system (EMS). The method comprises the following steps that 1, telemetric sampling devices arranged on all sampling sites in a grid system send telemetric data to the scheduling EMS, telemetric sample data are identified and a telemetric sample definition list is generated; 2, the scheduling EMS judges whether the telemetric sample data in the step 1 are defined; 3, the telemetric sample data which are judged to be identified in the step 2 are screened, and the defined or redundant sampling sites are removed; 4, the scheduling EMS system generates a new sampling site list file according to sampling sites corresponding to the removed undefined telemetric sample data in the step 3. By the method, a great amount of time required by manually searching the undefined sampling sites is saved, the method has high accuracy, and the practicability of filling leak of scheduling automation work is improved.

The invention discloses a data permission control method and device, and relates to the technical field of electronic information. The method comprises the steps of obtaining an SQL statement; segmenting the SQL statement with the defined data permission strategy to obtain a plurality of SQL statement segments; traversing the SQL statement segments one by one, and judging whether the SQL statementsegments starting from select are first elements after segmentation or not; if not, segmenting the SQL statement segment again; if yes, analyzing the table names of the SQL statement segments, and screening out the SQL statement segments of which the table names define the data permission strategy; replacing the table name of the SQL statement segment of which the table name defines the data permission strategy with the select statement to which the permission condition is added to obtain a plurality of changed SQL statement segments; and reconnecting the changed SQL statement segment to obtain a target SQL statement segment. According to the method, the SQL statement generated by Hibernate is intercepted and modified to achieve data authority management, authority management is carried out on a certain type of data, and a user can only inquire and manage user information of a department where the user is located.

The invention provides a data analysis visualization method, system and device based on big data artificial intelligence, and the method comprises the steps: building an Internet human resource big data talent demand index standard, and constructing a metadatabase according to the standard; collecting data from recruitment advertisements disclosed on the network according to the talent demand index standard established in the step S1, and storing the data by using a big data platform; cleaning and desensitizing the captured webpage data to form structured data with a standard format; constructing a machine learning algorithm to analyze and mine the data; and carrying out interactive real-time data visualization webpage display on the analysis result. The problems that an existing research institution mainly adopts a sampling survey mode for research on human resource requirements and posts of specific administrative divisions, and a data visualization display mode for human resource big data division and interactive real-time access based on the Internet is lacked are solved.

The invention discloses a streaming data processing device and method, and the device comprises: a data collection module which is used for collecting the streaming data of at least two data streams,and transmitting the streaming data of each data stream to a data generation module; a data generation module which is used for receiving the streaming data sent by the data acquisition module, integrating the streaming data of the at least two data streams into one or more streaming data tuples, and sending the streaming data tuples to the data output module; and a data output module which is used for outputting the streaming data according to the streaming data tuples integrated by the data generation module. According to the invention, the stream data of each data stream is collected, the stream data of all the data streams is integrated into one or more stream data tuples, and the stream data is output according to the stream data tuples, so that a data set with relatively aligned timecan be obtained, the problem of data time misalignment is solved, the stream data can be directly used for calculation, the calculation efficiency is improved, and therefore, a more rigorous calculation result can be obtained.

The invention provides a cluster metadata storage and reading method and device, and the method comprises the steps: obtaining cluster metadata which comprises a plurality of fields; segmenting the plurality of fields to obtain N groups of fields; storing the N groups of fields to M nodes in a distributed manner; selecting any node from the M nodes as a current node; receiving reading request information sent by a client through a current node, and sending the reading request information to other nodes except the current node; acquiring a field corresponding to the current node and fields corresponding to other nodes according to the reading request information; splicing the field corresponding to the current node with the fields corresponding to other nodes to obtain cluster metadata comprising N groups of fields; cutting and saving cluster metadata on each node in a distributed mode so that space can be saved. When the client needs to read the data, even if one node is in a non-working state, the client cannot be influenced to read the cluster metadata, and the user experience is improved.

The embodiment of the invention provides a data scheduling method, device and equipment and a storage medium, and the method is applied to an Internet of Things platform, and comprises the steps: receiving to-be-processed data reported by intelligent instrument equipment, and decoding the to-be-processed data to obtain an equipment instruction and Internet of Things equipment data; querying according to the equipment instruction to obtain at least one standard service function corresponding to the equipment instruction; querying according to the at least one standard service function to obtain a plurality of Internet of Things equipment data items corresponding to each standard service function; and scheduling the Internet of Things equipment data to the corresponding standard business function for business processing according to the plurality of Internet of Things equipment data items corresponding to each standard business function. According to the method provided by the embodiment of the invention, the intelligent instrument equipment produced by a manufacturer does not need to be subjected to secondary development, so that the code development frequency is reduced, the system access efficiency of the intelligent instrument equipment is improved, and the cost is reduced.

The invention relates to a learning resource management control method based on an online learning system. The method specifically comprises the following steps: S1, obtaining a resource type needing to be created, searching whether the same resource type exists in the system, and if not, creating the resource type; S2, configuring a basic field according to the created resource type, and acquiring a specific field of the current resource type to form a resource field of the resource type; and S3, acquiring corresponding resource data, inputting the resource data into the resource field, forming learning resources under the resource type, and generating a resource value as an identifier of the learning resources. Compared with the prior art, the method has the advantages that the convenience of an online learning system and the stability of function development are improved, the requirement that multiple persons use existing resources in the application is met, newly-added resources are not needed, resource sharing is achieved, and the operation cost, development cost and maintenance cost of enterprises are reduced.

The invention provides a risk identification method and device for a service request and processing equipment. The binary bitmap is used for storing the corresponding relation information between thefeatures and the user; based on the binary bitmap, the corresponding relationship between each feature in the service request and each user can be quickly determined; matching calculation is performedon the binary bitmaps of different feature parameter combinations by using a preset strategy, so that a risk user set corresponding to each feature in the service request can be determined, and the risk identification result of the service request can be further rapidly determined based on the risk user set. Characteristic user information is stored by using a binary bitmap, so that the storage cost can be greatly reduced, offline calculation is not needed, and an online database is not needed either.

The invention discloses a query method and device and a storage medium. The query method comprises the steps of in response to interface operation which is carried out by a user on an interactive interface and is related to a query task, displaying a tree-shaped graph corresponding to the query task on the interactive interface, wherein a node assembly of the tree-shaped graph is used for representing label information related to the query task; generating a tree-shaped query model corresponding to the tree-shaped graph, wherein nodes of the tree-shaped query model correspond to label information represented by node components of the tree-shaped graph; according to the label information corresponding to the nodes of the tree-shaped query model, carrying out query operation related to the query task, and determining data information corresponding to the query task.

The invention provides a PET detector energy correction method and system and a computer readable storage medium, and the method comprises: building an energy response function model and building a correction coefficient lookup table, and carrying out the parameterization of a coefficient in a mathematical model through building a first lookup table and a second lookup table, enabling the coefficient to be transmitted to a PET detector front-end circuit, and achieving the correction of the energy response function model and the correction coefficient lookup table. The method is used for real-time table look-up correction of the PET detector, the correction precision is high, the efficiency is high, and the method is suitable for all PET detectors using SiPM as photoelectric conversion devices; and the method is suitable for linear and nonlinear PET detectors at the same time.

The invention relates to the field of computers, and discloses a data platform migration method and system, and the method comprises the steps: copying the actual operation data of an old data platform; determining all old functions in the old script as a practical old function set according to the old function table; finding new functions with the name similarity with the old functions larger than a preset threshold in the new function table as candidate substitute new functions with the old functions; replacing the old functions with candidate replaced new functions as corresponding new scripts for training; respectively executing each old script and the corresponding new script for training based on the training data, comparing execution results of the old scripts and the corresponding new scripts for training, and establishing a mapping relationship; and converting an old function in the old script into a corresponding new function to serve as a formal new script of the new data platform. Efficiency of data platform migration from the old data platform to a new data platform can be remarkably improved, and the manpower consumption of developers is reduced.

The invention discloses a rule matching model generation method, a rule matching method, a rule matching system, a rule matching device and a medium. The rule matching model generation method comprises the following steps: acquiring business data and first business rule data; extracting features of the business data according to the type of the business data; splitting the first service rule data to obtain second service rule data and a relationship between the second service rule data; and generating a business rule matching model by adopting a depth-first traversal algorithm and a backtracking algorithm based on the characteristics of the business data, the second business rule data and the relationship among the second business rule data. According to the invention, the service rule matching efficiency is improved.