59 results about "Associative processing" patented technology

The invention provides a network security risk early warning method and system based on multi-source data fusion, and relates to the technical field of network security, and the method comprises the steps: collecting multi-source security data; performing entity identification and association processing on the multi-source security data to obtain entity association information; performing space-time alignment fusion processing on the entity association information, and constructing a threat fusion matrix; performing risk analysis and threat identification on the threat fusion matrix based on rule engine matching, a behavior anomaly detection AI model and a graph neural network; and performing graded early warning based on a risk analysis and threat identification result. According to the network security risk early warning method and system based on multi-source data fusion, the accuracy and the real-time performance of network security risk early warning are remarkably improved through multi-source data fusion and multi-dimensional analysis.

The invention discloses a multi-modal fusion bank receipt intelligent processing system and method based on vision and NLP, and the method comprises the steps: receiving a bank receipt picture or a PDF document, and completing the text detection, direction correction and character recognition through a visual processing engine; a three-level receipt independent segmentation mechanism is applied, and independent receipt records are divided through spatial clustering analysis, semantic analysis, visual verification and cross-page association processing; jointly extracting text features and layout features of each receipt through a double-flow multi-modal fusion model, and fusing the text features and the layout features; field-level data extraction is executed through a field extraction engine, and financial data verification including account number, amount, date and abnormity quadruple verification is carried out; and generating structured JSON output to obtain a bank receipt processing result. According to the invention, the innovative five-layer processing architecture realizes high-precision analysis of the bank receipts through an independently researched and developed receipt independent segmentation engine, a vision-semantic fusion model and a financial data verification system.

The embodiment of the invention provides a multi-modal data processing method and system based on a knowledge graph, and aims to solve the problems of semantic association deficiency and potential association mining difficulty in multi-source information processing. The method comprises the following steps: firstly, acquiring a multi-source information set containing different expression form information units and source identifiers, and performing entity mapping on the multi-source information set and a preset semantic association network to generate an information unit set with semantic association tags; performing cross-source feature association processing to obtain a comprehensive feature set; then, a semantic reasoning rule is called for association extension, and an extension feature set is generated; and finally, based on the extended feature set, generating an information processing result conforming to an application scene and feeding back the information processing result to a corresponding interface, thereby effectively integrating multi-source information and mining potential association.

A method for concurrently performing multiple computations in an associative processing unit (APU) includes having data in two matrices, representing data in two portions of a memory array of the APU, creating a Tartan matrix by computing an outer product between a first bit vector indicating selected rows and a second bit vector indicating selected columns, the Tartan matrix representing data stored in a third portion of the memory array wherein all cells having a value 1 in the Tartan matrix indicate selected cells, concurrently activating all cells of the matrices and storing a result of Boolean operations therebetween in one of the two matrices, wherein a new value is obtained on cells located at a same row and a same column as the selected cells in the Tartan matrix and an original value remains on other cells.

This invention belongs to the field of intelligent maritime surveillance technology, specifically a smart waterway monitoring and management system based on light beacons. It synchronously acquires multi-source heterogeneous information about vessels through integrated light beacon equipment, performs spatiotemporal registration and target association processing on this information, generates dynamic vessel trajectories, and extracts static attributes, motion behavior characteristics, visual activity characteristics, and environmental interaction characteristics. Utilizing a multi-evidence chain collaborative decision-making mechanism, it accurately identifies four fundamentally different vessel states: normal anchoring, illegal anchoring, active navigation, and uncontrolled dragging of anchor. This achieves a leap from state perception to behavioral cognition, significantly improving the accuracy of vessel anchoring identification and its robustness in complex scenarios. Based on the identification results, it triggers differentiated alarm strategies matching the state, generating handling instructions with reasons for the state determination, thereby greatly optimizing the allocation of surveillance resources and improving emergency response efficiency.

A calculator for calculating a plurality of square roots includes a memory array, at least two registers, a bit subtractor, and a controller. The memory array is organized into columns and rows. The registers store a fixed value. The controller, operatively coupled to the other components, allocates a first set of rows to test variables and a second set to result variables, and initially stores each of a plurality of radicands in a separate column. For multiple iterations, the controller concurrently activates selections of rows to form current operands and current guesses for each column. The controller then instructs the bit subtractor to perform subtraction operations. For each column with a positive subtraction result, the controller selectively writes a new bit of the square root and selectively overwrites values with a new remainder derived from the subtraction result, without performing an explicit data-shifting operation.

The invention provides a maintenance fund data management method and device, equipment, a medium and a product. Relates to the technical field of data processing. The method comprises the following steps: acquiring index design demand data and maintenance fund data of a plurality of cells; determining at least three index dimensions according to the index design demand data, and determining a main body and a to-be-combined attribute of each index dimension; storing the maintenance fund data of each cell as an independent attribute table in an attribute value manner according to the main body of each index dimension and the to-be-combined attribute; obtaining user query request data, and determining a plurality of to-be-combined attributes of the at least three index dimensions according to the user query request data; generating corresponding index query conditions according to the to-be-combined attributes of the at least three index dimensions; and carrying out association processing on the attribute table of the index dimension corresponding to each to-be-combined attribute and the business process data to obtain an index result meeting the index query condition. The method has the technical effect of improving the flexibility and real-time performance of maintenance fund data processing.

The invention relates to the technical field of large language models, and discloses an intelligent geographical name and address data matching method based on a large language model, which comprises the following steps: collecting geographical name and address related data and corresponding spatial graphic data, and carrying out fusion association processing on the geographical name and address related data and the corresponding spatial graphic data to obtain processed information; carrying out textualization and vectorization conversion on the processed information to form a place name address knowledge base adaptive to a large language model; based on a place name and address knowledge base and a big language model enhanced retrieval and cue word fine tuning technology, obtaining a place name and address matching model with address hierarchy matching result evaluation; receiving a to-be-matched place name address text, performing retrieval based on the place name address knowledge base after preprocessing, and obtaining a candidate matching result set; and calling a place name and address matching model to perform hierarchical evaluation on the candidate matching result set, and outputting an optimal matching result and a matching hierarchy thereof.

The invention discloses an artificial intelligence security operation central system, and particularly relates to the technical field of network security operation management. The system comprises a data fusion layer, an intelligent analysis layer, a decision response layer and a digital twinborn layer. The data fusion layer is responsible for performing real-time receiving, standardization and association processing on the multi-source heterogeneous security data to generate a security event sequence in a unified format; the intelligent analysis layer is used for identifying security threats and detecting abnormal behaviors through a multi-mode neural network and a behavior mode analysis module; the decision response layer automatically matches a response strategy according to a dynamic risk assessment result; the digital twinborn layer constructs a system virtual mirror image and supports attack and defense deduction and strategy pre-verification. According to the system, intelligence and automation of the whole safety operation process are realized, and the threat detection precision, the response speed and the overall safety defense capability are effectively improved.

This invention relates to the technical field of robot training. Specifically, it relates to a robot adaptive training method, apparatus, and medium based on reinforcement learning. The method includes generating task sub-objective information using a high-level policy network, inputting the task sub-objective information into a low-level execution network, generating action control commands based on the task sub-objective information, collecting interaction feedback information during execution, calculating a reward value based on the interaction feedback information, associating the reward value with scene complexity parameters, performing dynamic reward shaping operations to generate an adjusted reward signal, generating a policy model optimized by meta-learning based on the adjusted reward signal, loading the policy model into a simulation environment to generate a target policy model optimized by simulation training, loading the target policy model into the robot, and controlling the robot to perform task operations in an actual interactive scenario. This invention achieves the effect of adaptive task learning for robots in multi-interaction scenarios.

The application discloses a power distribution Internet of Things data real-time processing system and method based on lake warehouse integration, the system comprises an Internet of Things pre-acquisition module and a lake warehouse integration core processing module deployed by using a Docker container; the Internet of Things pre-acquisition module is used for acquiring power distribution Internet of Things data and sending the data to the lake warehouse integration core processing module; the lake warehouse integration core processing module comprises a data processing unit and a multidimensional analysis unit; the data processing unit comprises an original data processing unit, a dimension model processing unit, a fact data processing unit, an aggregation association processing unit and an application data processing unit; the application can effectively process and store massive heterogeneous data, thereby improving the confusion of data calculation analysis and statistics, reducing repeated calculation, realizing reuse of calculation results, improving the rapid retrieval efficiency of data, simultaneously providing multidimensional data support for the application side, and ensuring the response performance during large-scale complex data calculation.

The invention provides a voucher processing method which can be applied to the technical field of big data and artificial intelligence. The method comprises the following steps: based on a field mapping rule, carrying out mode mapping on preprocessed historical certificates from different service systems according to a standard field unified mode; performing multi-dimensional association processing on the historical vouchers after mode mapping and the multi-source real-time information associated with the historical vouchers to obtain a unified voucher data model; in response to the service request, preprocessing target service associated information associated with the service request, and generating a target service voucher by using the preprocessed target service associated information and the unified voucher data model; and carrying out anomaly detection on the service request by using the preprocessed feature clustering result of the historical voucher, and carrying out compliance check on the generation process of the target service voucher by using a compliance rule determined by the unified voucher data model. The invention further provides a voucher processing system, voucher processing equipment, a storage medium and a program product.

This invention provides a method and system for abnormal event association processing based on an operation and maintenance knowledge graph, belonging to the field of operation and maintenance management technology. The method includes: taking each abnormal event entity in a structured abnormal event sequence as an input node set; based on the topological mapping position of the input node set in a preset operation and maintenance knowledge graph, aggregating the k-order adjacency topological features and business semantic embedding features of each input node to obtain a graph event representation vector; performing pairwise comparison and temporal alignment of the feature vectors representing different abnormal event entities in the graph event representation vector; iteratively calculating the temporal lag correlation index and state transition probability matrix between each abnormal event entity; and converging to obtain a candidate abnormal event association subgraph based on the state transition probability matrix and a preset association confidence threshold. This invention realizes abnormal event association and handling instruction generation, improving operation and maintenance efficiency and the accuracy of fault root cause location.

The application relates to the technical field of data processing, in particular to a business data processing method and device, equipment and a storage medium. The method comprises the following steps: obtaining a to-be-processed data source, the to-be-processed data source comprising user information and business data; performing association processing on the business data and a data lake table based on the user information to obtain a first association result, the data lake table being used for indicating a table of label information corresponding to the business data; performing screening processing on the first association result according to a rule broadcast stream to obtain a target association result, the rule broadcast stream being used for indicating a screening rule corresponding to the business data; and determining corresponding target business data according to the target association result and performing push processing on the target business data. The method realizes real-time computing and processing of data and push processing, improves the efficiency and accuracy of real-time data processing, realizes accurate push processing of data, and provides real-time, efficient and accurate data processing services for users.

The invention discloses a bill information processing method and device, equipment and a medium, and is characterized in that the bill information processing method comprises the steps: receiving a multi-modal bill and bill reason description inputted by a user; performing information extraction processing on the multi-modal bill by executing the bill processing request to obtain bill key field information; performing context semantic understanding on the bill cause description to obtain cause key elements; performing association processing on the bill key field information and the cause key elements, and determining cause information corresponding to the multi-modal bill; and filling the cause information into a cause field corresponding to the associated multi-modal bill.

The invention provides a structure disease automatic identification and early warning method and system based on a BIM + Internet of Things technology, and the system comprises a BIM model platform which creates and manages a digital three-dimensional model of a building structure or infrastructure based on a BIM model, and the BIM model comprises detailed component information; the Internet of Things sensor network comprises a plurality of sensors; the data acquisition module is used for acquiring physical data of a building structure or infrastructure in real time through an Internet of Things sensor network to obtain structure state data, and transmitting the structure state data through a wireless or wired network; the data processing and analyzing module is used for receiving the structure state data, performing association processing based on the component information and the structure state data, and analyzing and judging the type and degree of the structure disease through an algorithm; and the disease identification and early warning module performs disease identification by using big data analysis, machine learning and a physical model, automatically sends out early warning information after the disease is identified, and generates a report.

This application is directed to an in-memory associative processing system. An apparatus can receive a set of instructions indicating a first vector and a second vector as operands for a compute operation. The apparatus can select a vector mapping scheme from a set of vector mapping schemes for performing the compute operation using associative processing. The apparatus can write the first vector and the second vector to a set of planes each including an array of content-addressable memory cells based on the selected vector mapping scheme.

An associative graph search system includes a KNN graph determiner to determine in advance W neighbors of each item in a dataset and to store each item and its neighbors in a KNN graph, a reduced dimension vector finder implemented on an associative processing unit (APU) to find a first number of first nearest neighbors of a query vector, the APU operating in a constant complexity irrespective of the size of the number, a result expander to find for each first nearest neighbor, W second nearest neighbors using the KNN graph thereby creating a group of neighbors, and a KNN full dimension vector re-ranker to find a final number of full dimension nearest neighbors of the full dimension query vector from the group of neighbors.

The invention relates to the technical field of data information processing, in particular to an information system integrated service system, which comprises an association model building end, a field data processing end, an association processing end and an information generation end, and is characterized in that historical construction data is processed to obtain a data association model, and meanwhile, field data acquired in real time is subjected to inertial processing to obtain a data association model; obtaining abnormal data according to an abnormal processing result, taking the abnormal data as main associated data, taking the rest of field data as auxiliary associated data in sequence, performing calculation in the data association model to obtain a construction association value, judging the construction association value to obtain a related data type, and obtaining target transmission personnel according to the related data type; and the information generation end automatically generates the information notification file and transmits the information notification file to the terminal equipment of the target transmission personnel, so that the data information difference occurring in the construction process is reduced, and the multi-dimensional data integration service is realized.

The present disclosure provides a software development whole-process measurement method and system, relating to the technical field of computers, wherein the method comprises: obtaining development process data; wherein the development process data comprises at least one of code repository data, merge request information, project management data, task information and unstructured document data related to the project; based on a preset association rule and the merge request information, performing cross-source association processing on the code repository data, the project management data, the task information and the unstructured document data related to the project to generate fusion data; and calculating a composite measurement index based on the fusion data. Through automatic association and unified analysis, all-round measurement covering process efficiency, collaboration quality and knowledge inheritance is realized, and natural language processing technology is used to intelligently identify risks and knowledge gaps from unstructured documents, provide early warning and accurate problem tracing, and improve development transparency and decision-making efficiency.

A system for N-gram classification in a field of interest via hyperdimensional computing includes an associative memory array and a controller. The associative memory array stores hyperdimensional vectors in rows of the array. The hyperdimensional vectors represent symbols in the field of interest and the array includes bit-line processors along portions of bit-lines of the array. The controller activates rows of the array to perform XNOR, permute, and add operations on the hyperdimensional vectors with the bit-line processors, to encode N-grams, having N symbols therein, to generate fingerprints of a portion of the field of interest from the N-grams, to store the fingerprints within the associative memory array, and to match an input sequence to one of the stored fingerprints.

The invention provides a network intrusion real-time detection method and system based on deep learning, and the method comprises the steps: activating a traceability evolution collaborative architecture, synchronously capturing network traffic and implicit association information thereof, generating traceability basic data and an implicit association feature set, and constructing an intrusion feature traceability map. And driving the feature extraction network topology of the deep learning detection model to autonomously reconstruct by using the graph node association intensity, and generating a model after topology evolution. And performing deep association processing on the traceability basic data and the implicit association feature set through the traceability detection collaboration layer to generate a preliminary detection signal. A weight self-calibration link is further constructed, weight parameters of all levels of the model are adjusted in a linkage mode, the model after full-link calibration is generated, and accurate detection is achieved. And finally, inputting a precise detection result into an intrusion mode self-updating module, driving the initial intrusion mode library to iteratively generate a dynamic evolution intrusion mode library, and generating a network intrusion interception instruction. According to the invention, the accuracy and real-time performance of network intrusion detection are improved.

The invention discloses a short message recommendation method and device, equipment and a storage medium, and relates to the technical field of big data, and the method comprises the steps: carrying out the feature extraction of multi-source behavior data of a user, and obtaining a document preference dimension; generating candidate short message copywriting matched with the big language model and the cue word template through the big language model and the cue word template; and performing multi-target multi-classification prediction on the user, and performing association processing on the candidate short message copywriting through a double-flow recommendation model according to the obtained customer group segmentation result and the multi-source behavior data to obtain a short message copywriting recommendation result. According to the method, the candidate short message copywriting matched with the copywriting preference dimension of the user is automatically generated through the large language model, and then the multi-target multi-classification prediction is performed on the user to obtain the customer group segmentation result with different emphasis on multiple targets, so that the customer group segmentation result with different emphasis on multiple targets can be obtained even in various user data. And short message copywriting recommendation results matched with user behaviors can be accurately generated, the diversity of customer groups can be accurately considered, and the marketing conversion rate of short messages is effectively improved.

This invention belongs to the field of assembly production and discloses a production line workstation-level AR modeling method and system for auxiliary assembly. The method includes S1, obtaining a 3D design model of the product from a design data source and structured assembly process data from a process data source; performing lightweight processing on the 3D design model and parsing and associating the structured assembly process data to generate preprocessed data; S2, based on the preprocessed data, generating quantitative semantic data describing the strength of assembly relationships and the difficulty of assembly steps through quantitative calculation, and constructing a unified assembly semantic intermediate model based on this data. This invention, through quantitative calculation, extracts machine-understandable and computable semantic parameters representing the inherent characteristics (strength) of assembly relationships and the cognitive load (difficulty) of operators from raw, discrete data, realizing automated integration from source data to AR logic.

The embodiment of the invention provides an insurance service recommendation method and device, electronic equipment and a storage medium, and belongs to the field of financial science and technology. The method comprises the following steps: acquiring multi-modal data of a target customer; constructing a product knowledge graph according to the multi-modal data and an insurance knowledge base; generating a demand strength vector with a timestamp according to the product knowledge graph and the historical insurance buying sequence of the target customer; generating a demand prediction curve with a confidence interval according to the multi-modal data and the demand intensity vector; performing product association processing according to the demand prediction curve and the product knowledge graph through a graph neural network model to obtain a product association degree matrix; performing real-time recommendation strategy optimization processing according to the product association degree matrix and the real-time interaction data flow of the target customer through a reinforcement learning model to obtain a target product recommendation strategy; and performing insurance service recommendation on the target customer based on the target product recommendation strategy. Based on the method, the accuracy of recommending the insurance product combination to the customer can be remarkably improved.

The application discloses a land space planning data monitoring and evaluation system, and belongs to the technical field of data monitoring and evaluation. The system comprises a data collection module, a data division module, a planning fusion module and a visual evaluation module. The data collection module uses an information processing platform to acquire remote sensing image data, monitoring data and historical planning data, and classifies, summarizes and arranges the data into a data set. The data division module performs dynamic knowledge graph processing on the data set, and performs entity relationship correlation processing on land attributes, ecological elements and basic setting data to obtain remote sensing image features. The planning fusion module is used for performing neural network processing on the remote sensing image features, obtaining three-dimensional space dynamic analysis by cross-level planning hidden conflict, and obtaining a planning decision scheme. The visual evaluation module performs feasibility evaluation on the planning decision scheme and obtains an evaluation conclusion. The system uses an index system and a calculation model to quantitatively analyze and real-time monitor land resource utilization and development status, and provides technical support for the whole process of land space planning implementation monitoring, evaluation and early warning.

The application relates to the field of Internet, and discloses a SQL statement processing method, device, equipment and storage medium. The method comprises the following steps: obtaining data information from a database table, extracting a target field used for creating a data model from the data information and performing marking processing on the target field to obtain marked data; receiving a first SQL statement, obtaining first parameter information and execution time consumption of the first SQL statement, matching the first parameter information with the marked data, obtaining a matching field, and performing association processing on the matching field and the execution time consumption; within a first preset time, counting association processing results whose execution time consumption exceeds a preset time threshold and performing marking processing on the corresponding association processing results to obtain a marked association relationship; when a second SQL statement is received, obtaining second parameter information of the second SQL statement, and optimizing an execution strategy of the second SQL statement according to the second parameter information and the marked association relationship. Through the above method, the execution efficiency and analysis efficiency of the SQL statement can be improved.

The application discloses a task processing method based on a quantum-classical hybrid architecture, a server for the quantum-classical hybrid architecture and a task processing system based on the quantum-classical hybrid architecture. The method comprises the following steps: establishing a session connection with a client through a preset communication protocol, and setting a session identifier; receiving a first task processing request sent by the client; determining session computing power resources according to the first task processing request; associating the session computing power resources with the session identifier; and processing the first task processing request according to the association between the session identifier and the session computing power resources to obtain a first task processing result. In this way, the session computing power resources are associated with the set session identifier, so that the session computing power resources cannot be used by other clients before the session connection is ended, and are only used for processing the task request sent by the client, thereby reducing the queuing time of the to-be-processed tasks and improving the overall efficiency of the quantum-classical hybrid architecture.

A non-transitory computer readable medium stores a program causing a computer to execute: processing of progressing a predetermined game; processing of determining a performance pattern to occur during the predetermined game from among a plurality of performance patterns, the plurality of performance patterns being associated with pieces of performance content different from each other and update conditions for a parameter different from each other; processing of performing a performance based on the determined performance pattern; processing of updating the parameter in accordance with the update condition corresponding to the determined performance pattern; processing of enabling a performance corresponding to the performance pattern having occurred during the predetermined game to be performed at a timing desired by a player; and processing of enabling a performance corresponding to the performance pattern having not occurred during the predetermined game to be performed at the desired timing by spending a predetermined item.

This invention discloses a method and apparatus for associating multi-source heterogeneous data for system evaluation. The method includes: acquiring a multi-source heterogeneous data set for system evaluation; the multi-source heterogeneous data set includes a subset of multi-source heterogeneous data for each technical indicator measured by the equipment system; the subset of multi-source heterogeneous data includes image data sequences, text data sequences, and test data sequences; preprocessing the multi-source heterogeneous data set to obtain a preprocessed dataset; and performing heterogeneous data association processing on the preprocessed dataset to obtain a set of multi-source heterogeneous data information for the same equipment system.