77results about How to "Solve data silos" patented technology

The invention provides an edge computing-oriented federated learning indoor positioning privacy protection method. The method is based on federated learning and differential privacy protection technologies. According to the method, credible federation training of an indoor positioning model is carried out in an edge computing environment; in the training process, all participating users do not share training data, distributed training and credible aggregation of the indoor positioning model are carried out only by sharing positioning model parameters, meanwhile, the model parameters are updated in an end-cloud cooperation iteration mode, so that the indoor positioning model is continuously optimized, and privacy protection and cooperative benefits of multi-user positioning model training are achieved. Experimental results show that compared with a traditional centralized model training method and a federated learning-based model training method, the method provided by the invention notonly can provide proven privacy protection, but also ensures the positioning effect of the model under the condition of increasing extremely little calculation overhead.

The invention discloses a multidimensional data fusion analysis system and method based on a cloud platform and belongs to the field of communication. The multidimensional data fusion analysis system comprises a data source layer, an access transmission layer, a data sharing and exchanging center and a big data fusion analysis platform, wherein the data sharing and exchanging center comprises a data extracting module, a converting module and a cleaning module, the extracting module is used for fusing various data sources and synchronizing data to a system data center in a unified mode, the cleaning module is used for processing various types of data according to unified metadata specifications, the converting module is used for conducting unified format conversion on the data undergoing standardized processing of the cleaning module, and the big data fusion analysis platform comprises an index establishment module, an analysis mining module and a deployment and control early-warning module. The multidimensional data fusion analysis system can integrate multidimensional information in a cross-platform mode according to intelligent deep mining of an analysis model, form a valuable information link and achieve information sharing and seamless connection of different service systems, and the problems of data islands and system islands are effectively solved.

The invention provides a storage method and system for target characteristic data. The method comprises the following steps that: a plurality of data storage modules used for storing data, and corresponding storage controllers are arranged in advance; a client side sends a data storage request which carries data information to a dispatcher, wherein the data information comprises security classification information, shared information, user group information and access permission information, the data storage request is forwarded to the selected storage controller, and the selected storage controller is notified to the client side; and the client side is directly communicated with the storage controller, the storage controller provides a uniform data read-write interface for the client side, and data uploaded by the client side is stored to the data storage module corresponding to the storage controller. When the method is applied, data storage cost can be lowered, the data is uniformly managed and maintained, data safety is guaranteed, and the data can be conveniently shared and used.

The invention discloses a data acquiring and grading early warning system based on a WEBGIS. According to the system, a current signal acquired from an enterprise safety production system is converted into a code numerical value capable of being read by software through a basic hardware layer; a transmission layer carries out data transmission according to the OPC or Web service technology, and data layer warehouse management is carried out; big data integration, comprehensive analysis and digital modeling are carried out on a data model layer; the WEBGIS technology is combined, map showing is displayed on an application layer through a middleware layer according to the hazardous chemical substance characters and the safety production characteristics, and grading early warning is carried out; by means of GIS comprehensive display, major hazard source pre-plan management, integral communication, emergency resource management, monitoring implementing, early warning results, contingency plans, decision making aiding and an emergency disposal system are combined for emergency commanding of major hazard sources.

The invention discloses a medical pension data safe exchange proxy device and a construction method thereof. The medical pension data safe exchange proxy device of a third party is constructed between a medical data system and a pension data system, and the device is formed by three parts of a data exchange proxy device, an authentication proxy device and a safe data transmission module. A half-duplex service mode is used for two-way exchange of the data, and storage of non-authorized data is effectively isolated; a data exchange isolation belt in a medical pension heterogeneous system is formed by a data exchange proxy server end, a mapping relation is formed against different classified access rights of the heterogeneous system, and the flexibility of data exchange and interconnection of the medical pension data island system are enhanced; in each step of the process flow, technological means of encryption, authentication and classified authorization is comprehensively used. According to the medical pension data safe exchange proxy device and the construction method thereof, the heterogeneous medical and pension data systems are safely integrated, and positive meaning is provided for medical pension service application extension based on the wide area network.

The invention discloses a multi-source heterogeneous data fusion and model correction process control method, and belongs to the technical field of automation. The invention is directed to a multi-level complex engineering system. according to the characteristics of data, models and processes, the process is further subdivided into sub-processes; defining a process template of each sub-process; including a data template, a feature template, a method template and a process template, various flexible configurations of the templates are realized; establishment of adaptive multiple heterogeneous data, a link of data and information processed in the fusion process of the model and the fusion method is carried out; fusion and verification are carried out according to an optimal method obtained by matching of an inference engine; the test data fusion and model correction links are more efficient; and moreover, the precision of the model after multi-source heterogeneous data fusion correctioncan be improved, the engineering design efficiency of the multi-level complex engineering system is improved, the prediction precision of the multi-level complex engineering system applied to the corresponding engineering technical field is improved by utilizing the model after multi-source heterogeneous data fusion correction, and related engineering technical problems are solved.

The invention relates to a data analysis management system based on a storm surge prediction model, and the system comprises a storm surge condition information database, a visual management terminal and a data graphical analyzer, wherein the storm surge condition information database is used for storing the surge level, air pressure, wind force, wind direction and ocean current at a surge monitoring station, and employing a mode of SQL SERVER network database. The visual management terminal is used for carrying out the visual representation of related information of the system, brings convenience to an operator for management and data analysis, and mainly comprises a system management module, an observation station information module, a surge level information module, a typical storm surge information module, a surge prediction module, and a storm surge knowledge module. The data graphical analyzer is an analysis tool which is designed based on the GUI of MATLAB, has mathematical analysis algorithms of data clustering, regression analysis and deviation analysis and comprises a plurality of drawing modes. The system facilitates the data analysis and visual management.

The invention discloses an Internet of Things data transaction structure based on block chain, which comprises a local wireless sensing network, a data providing node, a data consuming node, a data storing node, a leaflet and transaction credentials. The invention also discloses a transaction security method based on the data transaction architecture of the Internet of Things, comprising the following steps: S1, a leaflet construction sending stage; S2, a data uploading stage; S3, a transaction preparation stage; S4, a transaction processing stage; The invention solves the problem that a largenumber of isolated islands of internet of things data exist due to centralized storage in the prior art, which greatly hinders the flow of internet of things data and wastes the value contained in the internet of things data.

The invention discloses an aviation service customization system and method based on a federated learning technology platform. The system comprises a federated learning technology platform, a data source module, a data processing module, a local training module and a cloud module. The method comprises the following steps of establishing own local models for airlines and partners by utilizing own source data respectively; and then uploading parameters and gradients of the model to a federated learning technology platform, realizing fusion on the federated learning technology platform so as to establish a new virtual model, performing prediction and updating by utilizing the virtual model, and finally establishing a federated personalized recommendation model. The model established through the method can provide personalized and accurate product recommendation for aviation customers, meanwhile, the modeling process is achieved through a federated learning technology platform, and safetyand privacy of data are guaranteed.

The invention discloses a cross-screen user identification method and device, and relates to the computer technical field; one specific embodiment mode of the method comprises the following steps: invoking a user account pin and a user identifier user_id; converting the data corresponding to the user identifier user_id into a uniform data structure; processing the user account pin and the converted uniform data structure so as to form from user relation data, and storing the formed user relation data, wherein the user relation data comprises the association relation between the user account pin and the user identifier user_id. The method can solve the data island problems, thus providing complete user data and strong insurance for developing businesses.

The invention discloses a tap water supply quality monitoring system. The system comprises a water quality data center module, a water quality monitoring management module and an emergency response module. By arranging the water quality data center module, a data sharing mechanism according to requirements is realized, the data input and output ways and formats are unified, and the problem of dataisland is solved. Through arranging the water quality monitoring management module, water quality data generated in the production process are utilized to perform real-time process water quality monitoring, water source body monitoring, water quality early warning, water quality analysis and the like. The analysis and monitoring result can be used for guiding production to synchronously carry outan emergency event processing flow according to the water quality monitoring and analysis result by arranging an emergency event response module, and providing the functions of technical guidance, progress query, data analysis, auxiliary progress promotion and the like of the water quality emergency event. The method is widely applied to the technical field of tap water data processing.

The invention discloses a medical search engine system based on a block chain. The block chain search engine system comprises an operating system and a processing system, the operating system comprises a search setting unit, a search information setting unit and a search information screening unit; wherein the search setting unit is connected with the search information setting unit, the search information setting unit is connected with the medical data processing unit, the medical data processing unit is connected with the search unit, the search unit is connected with the medical data resultdisplay unit, and the medical data result display unit is connected with the search information screening unit; the processing system comprises a medical data processing unit and a medical data result display unit. Comprehensive search can be carried out in the whole system through a simple search program, search information is accurate, search time is saved, and a search user can use the systemconveniently.

The invention discloses a distracted driving recognition method based on federated learning. Two machine learning technologies of federated learning and a shallow convolutional neural network are adopted, the federated learning mainly utilizes distributed data, a global statistical model is constructed through a deep neural network (DNN), and the recognition accuracy is improved, meanwhile, key parameters are uploaded under the homomorphic encryption condition, and data generated locally is not leaked; a convolutional neural network (CNN) mainly utilizes the advantages of image feature extraction and is responsible for extracting differentiated features of a user side, namely personalization of a local model; through federated learning and homomorphic encryption technologies, the problem of personal privacy protection in a cloud environment is solved, the problem of data islands is effectively solved, and the recognition efficiency is improved, so that the requirements of practical application are met.

The invention discloses a coal mine distributed collaborative one map system and a collaborative management method. The collaborative system comprises a service interface layer, a service collaboration layer and a data storage layer; the service interface layer comprises mine maps of different majors of coal mines and a service interface of attribute processing, the service collaboration layer corresponds to each professional service interface comprising a corresponding mine map and an interaction interface of attribute management, and comprises a user authentication module, a graphics and attribute processing cooperation module, a searching service module and a map service module; and the data storage layer is used for storing coal mine map data, and comprises an index database, a tile library, and a database. According to the technical scheme, a unified management platform for multi-major data stratification is established, dynamic real-time update and sharing of various types of data are realized, the consistency and timeliness of data in different majors and different users are insured, the digital efficiency and production efficiency of the mine are greatly improved, and a powerful and effective technical support for coal mine safety management is provided.

The invention provides an intelligent hospital information interaction system, and belongs to the technical field of data processing. A first user terminal specifically comprises a first data transmission module, a first data acquisition module, a search module, a consultation reservation module, a first interaction module and a payment module; a second user terminal specifically comprises a second data transmission module, an information acquisition module, a prescription creation module and a second interaction module; and a server specifically comprises a third data transmission module, a second data acquisition module, a consultation recommendation module, a consultation generation module, a cost calculation module and a communication establishment module. The technical scheme has theadvantages that the problem of data isolated island caused by a self-built system in each hospital is solved, so that doctors and patients can quickly and directly exchange information, the doctor-seeing efficiency of the patients is improved, the working intensity of the doctors is reduced, the doctor-seeing information of each hospital is counted, and the most suitable doctor-seeing scheme is recommended to the patients.

The embodiment of the invention provides a method and device for jointly training model based on multi-party private data, and the method comprises the steps that any first party in a plurality of first parties utilizes the first privacy data and first class labels of all objects in a local first object set and the first privacy data and first class labels of other first parties, based on a transverse federated learning mode, a first parameter set shared by a plurality of first parties corresponding to the first sub-model is acquired; a common object in which a local first object set and a second object set of a second party are overlapped with each other is determined; a first parameter set corresponding to a local first sub-model is updated based on a longitudinal federated learning modeby utilizing first privacy data corresponding to a local common object, and second privacy data and a second category label of a second party; and a second party obtains a second parameter set corresponding to a second sub-model of the second party. The prediction accuracy of the model obtained after training can be improved.

The invention relates to the technical field of databases, in particular to a system for achieving database synchronization and a method thereof. The method includes the steps of obtaining and analyzing the operations executed by a database in real time by monitoring logs in source data, according to a protocol, conducting data encapsulation on the executed operations, and pushing the data to a target server; after the target server receives the data, verifying the data; if an exception occurs, requesting for obtaining the data again, and if not, adding the data into a data queue to wait for being written into the database. The system for achieving the database synchronization and the method thereof solve the problem of data isolation in each application system, and can be applied to data synchronization among different application systems.

The invention relates to an NB-IOT-based micro-grid information acquisition and monitoring system and an implementation method thereof. The system comprises a field equipment layer, a network transmission layer and an application layer. The field equipment layer comprises a micro-grid environment information acquisition system and a micro-grid field equipment monitoring system, and is responsible for collecting environment data and field equipment operation data; the network transmission layer comprises a ZIGBEE wireless network, a MODBUS RTU bus network, an intelligent gateway and an NB-IOT network, and is responsible for data transmission and protocol conversion; the application layer comprises a micro-grid intelligent monitoring platform; and the micro-grid intelligent monitoring platform comprises a micro-grid environment safety evaluation system and a micro-grid field equipment monitoring system, and provides system monitoring and environment evaluation functions. The intelligent gateway is added into a traditional micro-grid, and remote transmission of micro-grid system data is realized through a protocol converter embedded in the gateway, so that remote monitoring of the system is realized; and by means of the micro-grid environment safety evaluation system, the possibility of accidental loss of field equipment is reduced.

The invention relates to a data storage and query method based on a block chain. The method comprises storage and data query, wherein the data storage is characterized by converting data which is notdynamically written at present; randomly extracting an algorithm from the encryption algorithm supported by the system; sorting the field names in the database table into a list, adding table names and library names, and sequentially adding aliases to the attribute names of the list to form an attribute alias mapping table; at the same time, inquiring the aliases in the attribute alias mapping table according to input data and field names, and then forming an alias data mapping table; writing the new attribute alias mapping table into a database, and returning to the primary key; writing the alias data mapping table and the primary key into a block chain to form block chain data; The data query is characterized by combining the obtained alias data mapping table and the attribute alias mapping table and decrypting the alias data mapping table and the attribute alias mapping table by using a corresponding algorithm to obtain an attribute and data mapping table and obtain data. Accordingto the method, data are hidden in a two-layer mapping relation, and the data security and sharing are ensured based on a block chain.

The invention belongs to the field of computer science, and discloses a distributed file storage sharing system based on an alliance chain. Through the alliance chain data interface module, the alliance file storage module, the alliance identity authentication module, the alliance authority control module and the node Web service module, distributed storage, file retrieval and file authorization of alliance files with identity access are realized, and safe sharing of files among alliance organizations is ensured. The invention provides an inter-alliance file distributed storage sharing solution with an identity authentication mechanism in order to overcome the defects of an existing block chain file sharing technology system, and provides development practice for inter-alliance data sharing, data barrier getting through and data island solving.

The invention discloses an internet financial gang fraud behavior detection method based on a knowledge graph. The method comprises the following steps: obtaining personal application information, operation behavior burying point data and blacklist data of users of a plurality of preset data sources; preprocessing application information and operation behavior buried point data, segmenting a training set and a test set, marking clients as fraudulent nodes and unmarked nodes according to blacklist hit conditions, solving similarity and affiliation factors between the fraudulent nodes and adjacent user nodes, and performing fraudulent risk assessment on the unmarked nodes to obtain a fraudulent risk assessment result; adopting a Neo4j graph database to construct a knowledge graph, testing averification set fraud risk assessment result, and detecting and processing real-time application user fraud behaviors. According to gang fraud behavior detection of the knowledge graph, the anti-fraud engine is constructed, suspicious group fraud risks are rapidly and efficiently recognized, financial risk control capacity is improved, and credit risks are reduced.

The invention discloses a method for solving the problems of network safety and data silos by using the Linux system. The method is characterized in that a Linux operating system platform is mounted on an X86 dual network interface card server, meanwhile respective IP address and gateway of dual network interface cards are arranged from the inside out according to respective network segments of internal and external networks, and an Oracle data base software is mounted on the Linux operating system platform; necessary network ports for non-Oracle data exchange in the Linux system are blocked; the monitoring program (TNS) of the well mounted Oracle software is set, and data visits to data bases of related production internal network and external network are performed; in the setting of the production internal network and the production external network, data base information is hunted through mirroring or information is obtained through writing programs to visit the data base. The method has the advantages as follows: under the premise of ensuring the safety of the production internal network, internal and external network data exchange visits can be achieved, the problems of data silos caused by safety factors on the production internal network can be solved, the quality is high and the price is low.

The invention belongs to the field of data identification, particularly relates to a distributed language relationship identification method, system and device based on federated learning, and aims tosolve the problem that a plurality of participants are difficult to perform joint modeling on the basis of not sharing data. The method comprises the steps that each local platform obtains to-be-recognized data and local training data with labels, a local classifier model is trained according to the local training data, the trained classifier models are integrated into a global classifier model,the local classifier model is initialized through the global classifier model, and to-be-identified data is identified by using the initialized local classification model to obtain a language relationship prediction result. According to the method, the local platform is combined with other platforms to perform data use and machine learning modeling on the basis of not sharing data, so that the training collaboration of the natural language recognition model is improved, the problem of data islands is solved, and the total amount and requirements of training data required for training the modelby each platform are reduced.

The embodiment of the invention discloses a line network comprehensive monitoring method and system based on a cloud platform. According to the technical scheme provided by the embodiment of the invention, the method comprises the steps: collecting and storing the equipment state data of each line in a monitoring platform and a data platform which are built on a cloud platform; enabling the monitoring platform to only store the latest equipment state data and perform data monitoring processing on the equipment state data to obtain the real-time monitoring data, enabling the monitoring platformto send corresponding real-time monitoring data to the client according to a data subscription demand of the client in the application layer, so that the client performs equipment real-time monitoring processing based on the real-time monitoring data; meanwhile, in response to a historical data application request sent by the client, enabling the data platform to provide equipment state data to the client as historical score unifying data, so that the client performs historical state monitoring processing based on the historical score unifying data, resource sharing and unified management andcontrol are realized, and the data island problem is solved.

The invention discloses a neurodegenerative disease modeling device based on federal transfer learning and related equipment, relates to the technical field of artificial intelligence, can be appliedto a smart hospital system, and comprises a data set acquisition unit used for acquiring a medical record data set A and a medical record data set B of a first type of neurodegenerative disease patients; an intersection selection unit which is used for selecting intersection data between the medical record data set A and the medical record data set B and taking the intersection data as a trainingset sample; a model training unit which is used for training an original neural network model based on the training set sample; and a federal migration unit which is used for performing federal migration learning on the original neural network model based on the training set samples of the multiple types of neurodegenerative diseases to obtain a final neural network model for predicting the disease degree of the neurodegenerative diseases. According to the method, the data island problem in the modeling process is solved, and the illness degree of a patient can be effectively predicted.

The invention discloses a distributed energy data fusion and application method based on multi-source isomerism. The method is multi-source isomerism data collection, data fusion and data application oriented to the natural gas energy industry. Wherein the data acquisition comprises natural gas station instrument data, molecular company information system data and filling data; the data acquisition layer takes OPC, HTTP, WEBSERVICE and TCP / IP as transmission channels, and completes data acquisition in a distributed task scheduling mode; after data collection is completed, integrity and validity verification is carried out on the data, meanwhile, the source, the service attribute and the context information of the data are established, and fusion and storage of the data are completed; data release adopts an authorization mode, and a unified data interface service is established in a form of transmitting JSON messages by an HTTP protocol. According to the invention, the problems of data islands, data redundancy and data exchange in enterprise management are solved, and application demonstration of information acquisition automation, data service standardization and decision management intelligence is established for enterprises.

The invention relates to a smart-city application deletion method and device based on container technology. The application deletion method includes: 1) acquiring an application deletion command, wherein the application deletion command includes a name of a to-be-deleted application and a smart-city application set of which the application needs to be deleted, the smart-city application set corresponds to a city application arrangement script based on a container file format, and the city application arrangement script is constructed through selecting needed application templates, instructionsof relationships among applications and city map graphics and dragging the application templates and icons of the instructions of the relationships among the applications onto a city map to realize association of the three parts; and 2) deleting the to-be-deleted application, a corresponding icon and related instructions of relationships among applications from the city application arrangement script according to the application deletion command. Compared with the prior art, the method can solve problems of data islands and management islands, and has the advantages of convenience, high efficiency and the like.

The invention discloses a privileged account threat analysis system, which comprises an intelligent threat auditing unit, a real-time threat monitoring unit and an overall configuration management unit which are connected with one another, and is characterized in that the intelligent threat auditing unit is used for auditing and monitoring account threat abnormal behaviors, and an administrator performs auditing analysis from different security dimensions and performs visual display. The privileged account threat analysis system can be compatible with structured and unstructured data of various operating systems, applications, databases, network equipment, security equipment and the like, can solve the data island problem existing in privilege threat analysis of a traditional system or tool, effectively carries out unified centralized management on external and internal threat behavior audits of the privileged account, realizes real-time omnibearing monitoring and association analysisof privileged account behaviors, can actively send the detected suspected privileged account threat behaviors to related personnel for real-time alarm, can improve the real-time protection capability,and can reduce further loss caused by the privileged account threat behaviors in an enterprise to the greatest extent.

The invention provides a big data processing method based on blockchain and equipment. The big data processing method comprises the following steps of S1, generating a user data processing account number, a password and authority for operating in the blockchain; S2, determining legality of carding data and transmitting reasonable data; S3, performing classified operation on the valid data; S4, inputting data in a stream format, and encrypting the data in a blockchain no-center storing mode; S5, performing backup of encrypted new data; and S6, performing data classified arrangement and re-storage through data preprocessing, data mining and machine learning manners. The big data processing method is a big data processing mode based on the blockchain, thereby reducing the number of times of acquiring the resource from a native resource database by a user terminal, and greatly reducing a data transmission load of a web server. Furthermore, in data encryption, the big data processing methodhas advantages of high strength, high speed, high attack resistance, etc. Furthermore the big data processing method has advantages of performing encryption and backup on the data, and ensuring highdata safety.