677 results about "User privacy" patented technology

The invention discloses a dynamic portrait construction method and system fusing a large model and user behavior data, belongs to the technical field of artificial intelligence and big data analysis, and aims to solve the problems of insufficient real-time performance, difficulty in multi-source data integration and high privacy risk in the traditional technology. The method comprises the following steps: collecting basic attributes, behavior sequences and unstructured data by burying points, and processing through an Apache Flink session window and a dynamic watermark; multi-modal feature extraction (discrete feature embedding, bidirectional LSTM coding behavior sequence and BERT coding text) is carried out, and joint embedding is generated through cross-modal contrast learning; generating three types of labels, namely a static label (rule engine), a dynamic label (1.3 B parameter quantity Nano-vLLM) and a predictive label (XGBoost), and dynamically adjusting weights; and realizing global model updating through federated learning and differential privacy. The system comprises a data acquisition layer, a feature extraction layer and a label generation updating layer. The real-time performance and accuracy of the portrait are improved, the privacy of the user is protected, and the commercial value in e-commerce, finance and other scenes is remarkable.

The invention provides a credible data space construction method and system based on data sharing, and relates to the technical field of data processing.The method comprises the steps that S1, in response to a request of a user for sharing target data, credible preprocessing is conducted on the target data to obtain credible data; s2, establishing a decision-making capability model for a user to autonomously perform credible authorization on the credible data; s3, analyzing a multi-dimensional sharing risk of the trusted data; s4, planning an optimal authorization auxiliary unit sequence based on the multi-dimensional sharing risk and the decision ability model; s5, in the process that the user autonomously performs credible authorization on the credible data, sequentially executing the authorization auxiliary units in the optimal authorization auxiliary unit sequence; and S6, aggregating the credible data subjected to credible authorization to construct a credible data space. According to the method, the credibility and the security in the data sharing process are remarkably improved, and the quality of the shared data and the protection of user privacy are ensured.

The invention discloses a method for predicting an abnormal behavior state of an elderly living at home based on digital twinning, and the method comprises the steps: obtaining a social relation network data flow and a non-intrusive energy consumption data flow of the elderly living at home, and constructing a time sequence dynamic social relation graph based on the social relation network data flow, meanwhile, a semantic activity event chain is generated based on energy consumption data flow overload decomposition, cross-modal fusion processing is carried out on the social relation graph and the activity event chain, a composite state vector is generated, and based on a time sequence of the composite state vector, the social relation graph and the activity event chain are subjected to cross-modal fusion processing; and predicting the probability of entering a preset abnormal behavior state (such as social isolation, depression tendency or decline of self-care ability) of the home elderly. According to the invention, the privacy of the user is protected through non-intrusive data acquisition, and the social communication and exclusive activity information of the elderly are fused, so that the transformation from post-event alarm to pre-event early warning is realized, a precious time window is won for guardians or service institutions to take preventive intervention measures, and remarkable social and humanistic values are achieved.

The invention discloses a voice privacy leakage detection and desensitization system and method based on an end-to-end model, and belongs to the field of voice recognition and privacy protection. The problems that semantic desensitization is not thorough, the voiceprint protection effect is poor, and full-process auditing is lacked are solved. According to the technical principle part, a voice recognition module, a semantic desensitization module, a voiceprint desensitization module and a log auditing module are included, original audio is input, and Chinese voice transcription is completed through a pruned and optimized Whisper-medium model; and the semantic desensitization module and the voiceprint desensitization module regenerate a new audio, the voiceprint similarity is ensured to be less than or equal to 0.3 through verification, and the identity traceability of the speaker is cut off. Log auditing records the processing process in real time, and abnormal behavior early warning is conducted. According to the method, a multi-level privacy leakage detection defense line is formed, abnormal behaviors can be found in time, and user privacy safety is guaranteed.

The embodiment of the invention provides an image processing method and device based on computer vision and artificial intelligence, and the method and device achieve the privacy protection of data through innovatively designing a multi-dimensional data safety collection mechanism and through feature branch isolation and space-time coding. A hierarchical feature fusion network is constructed, and a safe and reliable target tracking system is established in combination with environmental constraints and physical laws. A dynamic weight adaptive mechanism is introduced, and the accuracy of a processing result and user privacy are ensured through multi-scale feature extraction and prediction correction. According to the method, the data security is protected, and meanwhile, the defects of the traditional technology in the aspects of data fusion, feature extraction, prediction constraint and the like are effectively overcome.

The invention discloses an electronic ticket ordering and payment application system based on a block chain, relates to the technical field of block chains, and is used for ticket business whole-process digital management, and a user interaction module supports multi-terminal operation and information management; the block chain order management module is responsible for order chaining and state updating; the multi-mode payment module supports double-track payment and encrypted transmission; the comprehensive matching module is linked with the block chain order management module and comprises a meal subsystem, a tourism subsystem, a hotel reservation subsystem and the like, and the passenger ticket verification module realizes multi-mode verification; the data storage module stores key data; and the intelligent contract scheduling module automatically executes the pre-deployed contract. The problems that an existing system is poor in data safety, low in efficiency and insufficient in adaptability are solved, ticket transaction safety and process automation are improved, data synchronization is guaranteed, multiple languages and multiple currencies are supported, user privacy is protected, multiple scenes are adapted, and travel and supervision requirements are met.

The invention relates to the technical field of artificial intelligence and education technology cross fusion, and particularly discloses a new media course intelligent learning resource generation system based on generative AI, and the system obtains the multi-dimensional state data of a learner through the means of eye movement tracking, electroencephalogram collection, interaction behavior analysis and the like, and constructs a high-dimensional state space through fusion; differential induction logic programming and a depth feature extraction network are cooperatively compressed into a low-dimensional semantic meta-state space, and the model interpretability and decision-making efficiency are improved; dynamic matching of personalized resource types and difficulty levels is realized based on a graph neural network and reinforcement learning; calling a multi-modal generation model to generate texts, videos or interactive contents as required, and introducing a cross-modal consistency verification mechanism to guarantee the output quality; the system also has an online adaptive optimization capability, combines a symbol rule recombination mechanism and a differential privacy protection strategy, and takes account of user privacy security while ensuring long-term stable operation of the system.

This disclosure provides techniques to utilize decentralized identifiers (DIDs) and verifiable credentials for secure, privacy-preserving transactions. In one aspect, a method is provided that includes: receiving user information; determining a DID based on the information; providing the DID to a user device; verifying the user's identity by validating the DID and associated verifiable credentials; and performing a transaction based on the verified DID. Other aspects are provided, such as generating a public-private key pair for the user, associating the DID with the public key, and / or creating a DID document stored on a distributed ledger accessible to authorized entities. Further aspects include processing transactions through smart contracts on a blockchain network, which may involve converting central bank digital currency to fiat currency while maintaining user privacy, applying transaction limits based on verified identity attributes, and providing zero-knowledge proofs to auditors to verify compliance without accessing underlying transaction details.

The invention discloses a neural network model reasoning method for improving intermediate data security, and relates to the field of neural networks, in the method, in the neural network model reasoning process of an NPU, intermediate feature map data generated by each operator layer is encrypted in real time and stored in an end-side SoC memory in an encrypted form, and the security of the intermediate data is improved. Intermediate data leakage caused by analysis of an attacker on memory data can be effectively prevented, the intermediate data output by the neural network model is protected against cracking, threats in privacy leakage aspects such as member inference attacks can be effectively resisted, user privacy data is ensured not to be stolen, and user experience is improved. The data security of the neural network model in the reasoning process is greatly improved, and a powerful guarantee is provided for user information security.

The invention discloses a decentralized identity authentication method and related equipment in combination with a Merkel tree and zero-knowledge proof. The method comprises the following steps: generating an initial Merkel proof through an identity provider; zero-knowledge proof is generated through a zero-knowledge proof module; requesting to obtain an initial Merkel proof corresponding to a user from an identity provider through an authenticator; the authenticator performs first identity authentication on the user according to the initial Merkel proof, and if the first identity authentication of the user is successful, the authenticator performs second identity authentication on the user according to the zero-knowledge proof; and if the second identity authentication of the user is successful, generating an identity authentication success result, and returning the identity authentication success result to the user. According to the method, the integrity and verifiability of data can be improved, the risks of fraud and identity embezzlement are reduced, a novel attack mode is effectively resisted, higher-level data security characteristics are realized, user privacy protection is greatly enhanced, and the method can be widely applied to the technical field of network security.

A Monitored Learning System captures a user's multi-modal interactions across applications and devices—including GUI events, keystrokes, pointer movements, screen content, and audio—to train models that predict and perform user-consistent actions. A User Interaction Monitor records event, visual, and audio data; a Data Processing Unit aggregates and enriches the data via time alignment, optical character recognition, GUI element recognition, and speech-to-text to produce structured training datasets; an AI Training Engine learns policies that generalize the user's workflows; and an AI Simulation & Deployment module executes predicted actions on target applications, optionally with scheduling and feedback logging for continual improvement. The system enables a personalized automation agent that adapts to variations in content and interface layout while respecting privacy through configurable redaction.

A method comprises:receiving, by a user device, from a verifier device, a request for user data;retrieving a first cryptogram and a decryption key;sending and, to a server, the first cryptogram;retrieving a random and a second cryptogram generated using reference user authentication data concatenated with the random;sending, to the verifier device, the second cryptogram and the random;storing the reference random;sending, to the user device, the second cryptogram;decrypting the second cryptogram using the decryption key;extracting the reference user authentication data and the random;providing, the user device, with user authentication data;verifying that it matches the reference user authentication data;providing, the verifier device, with the random;verifying that it matches the reference random; andauthenticating the user data.

The invention discloses a new media advertisement putting and monitoring integrated system, and the system comprises a user portrait construction module which obtains the browsing behavior, geographic position and social interaction data of a user at a mobile terminal, a PC terminal and an intelligent terminal through a multi-source data collection unit, and generates a dynamically updated user interest tag; the intelligent advertisement putting engine is used for matching advertisement contents with user portraits in real time based on a deep reinforcement learning algorithm and supporting programmed cross-platform bidding and private market direct putting; according to the invention, through multi-dimensional effect evaluation, emotion analysis and user life cycle value prediction are introduced, and a comprehensive evaluation system is constructed; data security is enhanced, and through homomorphic encryption and block chain technologies, user privacy is guaranteed and a non-tampering effect auditing voucher is provided. The customer obtaining cost of advertisers can be obviously reduced, and the rate of return on investment is improved.

The invention particularly relates to a shared economic credit evaluation and evidence storage method based on a block chain. According to the shared economic credit evaluation and evidence storage method based on the block chain, during first registration and login, initialization processing is performed on a user identity, service behavior data is acquired in real time, preprocessing is performed, feature engineering processing is performed, and credit factors are extracted and normalized; constructing a multi-dimensional credit model, and realizing dynamic credit score and threshold adaptive calibration; generating a multi-dimensional abstract, and storing an evidence on a chain; the target sharing economy platform obtains the credit voucher of the user in a cross-platform manner and performs verification and synchronization; multiple security and compliance strategies are provided, and privacy protection, compliance supervision and black and white list feedback closed-loop management are realized. According to the shared economy credit assessment and evidence storage method based on the block chain, credit credibility and traceability are improved, migration and verification of user credit among multiple shared economy platforms are realized, user privacy and compliance are guaranteed, and multi-platform joint management and risk early warning are realized.

The invention discloses a smart home control method and device, electronic equipment and a storage medium, and relates to the technical field of transmission and bearing, and the method recognizes feature information based on a radar module, and controls a smart home system to execute a target program according to the feature information. Wherein the feature information comprises at least one of action information, quantity information, movement track information, breathing and heartbeat information and posture information of the target feature, and the target program and the feature information have an association relationship. Through application of the technical scheme of the invention, perception without external terminal or voice dependence is realized through the radar module, multifunctional control of the smart home can be realized, the control difficulty of the smart home is reduced, sensitive information such as a human face does not need to be collected, convenience, functional diversity and user privacy are considered, and thus the use experience is improved.

The invention discloses a staggered layer occlusion structure capable of forming a privacy space. The staggered layer occlusion structure comprises an upper layer structure and a lower layer structure which are built in the vertical direction. The lower-layer structure is provided with an occlusion space which is surrounded by the upper-layer structure in the horizontal direction and / or built in the vertical direction; the lower-layer structure is provided with a first privacy space, the upper-layer structure is provided with a second privacy space, the first privacy space and the second privacy space are physically isolated through spatial arrangement and / or barrier arrangement, and the first privacy space is located below the second privacy space. The first privacy space and the second privacy space are physically isolated to guarantee the privacy and safety of the user, the first privacy space is located below the second privacy space, and the first privacy space and / or the second privacy space are / is located in the south direction of the staggered layer occlusion structure. The upper-layer structure and the lower-layer structure are house types which are through in the south-north direction or through in the east-west direction. And a good ventilation effect is achieved while house lighting is guaranteed.

The invention relates to the technical field of smart home, discloses a smart home equipment state monitoring and fault self-healing optimization method and system, and aims to solve the problems of low operation and maintenance efficiency and untimely fault processing of smart home equipment in the prior art. The method comprises the following steps: continuously acquiring equipment operation state data, communication interaction data and environment context data through a multi-source sensor and a communication interface; preprocessing and feature engineering are carried out on the collected original data, and association weights of cross-device features are calculated based on an attention mechanism; constructing a fault prediction network by using a gating cycle unit and a causal graph model, and outputting fault probability distribution and a root cause analysis result; and generating a self-adaptive repair strategy according to the fault type, the root cause chain and the equipment operation context. The system comprises a multi-source heterogeneous data acquisition module, an equipment state feature extraction and fusion module, a fault prediction and root cause analysis engine, a self-adaptive repair strategy generator and a knowledge migration and model evolution module. Through adoption of the technical scheme, fine-grained and continuous sensing of the equipment state can be realized, accurate root cause analysis and multi-granularity early warning are provided, manual intervention requirements are reduced, the self-healing efficiency of the system is improved, and the adaptability and robustness of the system are enhanced on the premise of guaranteeing user privacy.

The embodiment of the invention discloses an action processing method, device, system and equipment, a storage medium and a computer program product. The method comprises the following steps: acquiring point cloud data corresponding to each target object in a target space; recognizing the action of the target object based on the point cloud data to obtain the action corresponding to the target object; if the action is the target action, indicating the movable equipment to move to an area corresponding to the target object, and obtaining feedback information obtained by interaction between the movable equipment and the target object; and processing an event corresponding to the target action based on the target action and the feedback information. According to the method, action recognition is carried out by acquiring the point cloud data of the target object, a user does not need to wear any equipment, user privacy leakage is avoided while the action recognition accuracy is improved, meanwhile, the mobile equipment is linked to interact with the user, the event is processed according to the feedback information obtained through interaction, and the user experience is improved. And the action processing efficiency and the event processing capability are improved.

The invention discloses a multi-agent deep reinforcement learning method considering user-power distribution network cooperation under topology reconstruction, and the method comprises the steps: carrying out the modeling of an independent interaction user agent through a multi-agent reinforcement learning MADDPG algorithm, enabling a user to make a decision independently according to the state of the user, enabling the user to interact with a power distribution network, and carrying out the topology reconstruction through the power distribution network. A virtual power flow method is adopted to maintain radial operation of the power distribution network, power and voltage balance of the power distribution network under topology reconstruction is considered, and a new method for interactive operation of independent users and the power distribution network under the extreme disaster condition is provided. According to the method, the current practical situations of frequent occurrence of extreme events with high influence and low probability and improvement of distributed photovoltaic permeability are considered, and through independent scheduling of independent users and reconstruction of the topology of the power distribution network, the power supply reliability and safety of the power distribution network under extreme disasters are guaranteed, and the privacy of the users is protected.

The application provides a user positioning method and device for low-orbit satellite communication and satellite network equipment, relates to the technical field of satellite communication, and the device comprises a receiving unit configured to receive a transmission delay and a Doppler frequency shift reported by a terminal device; a first processing unit configured to control the satellite network equipment to perform the following first operation: calculating a first included angle from the terminal device to a subsatellite point via a satellite according to the transmission delay and an orbital height of the satellite; calculating a second included angle from the subsatellite point to the terminal device in a satellite moving direction via the subsatellite point from a satellite overhead angle according to the first included angle, the Doppler frequency shift, a carrier frequency and a satellite moving speed; and determining position information of the terminal device according to the first included angle and the second included angle. The application can automatically estimate the position information of the terminal device without the terminal device reporting the position information, ensure the service continuity of a communication architecture depending on the terminal device information, and protect the user privacy and safety.

The invention discloses a privacy protection method in multi-modal data fusion, and the method comprises the following steps: S1, obtaining multi-modal original data of at least two modal types, carrying out the standardization preprocessing of each modal data, eliminating the differences of different modal data in dimensions and value ranges, and obtaining standardized multi-modal input data; s2, a privacy-semantic coupling model is constructed, the model comprises a privacy feature space and a semantic feature space, and the invention relates to the technical field of data security. According to the privacy protection method in multi-modal data fusion, a privacy-semantic coupling model is constructed, a privacy variable is used as an implicit interference factor to be embedded into a multi-modal fusion process, privacy noise in multi-modal data is automatically corrected and counteracted through a data processing model, and under the premise that original sensitive data is not explicitly exposed, the privacy of the multi-modal data is effectively protected. The privacy protection in the multi-modal data fusion process is realized, and the risk of user privacy disclosure is greatly reduced.

The invention provides a variable frequency air conditioner cluster regulation and control system based on privacy protection and a distributed communication network, and the system comprises a model building module which is used for building an air conditioner comfort model, building an air conditioner operation model, and substituting the air conditioner comfort model into the air conditioner operation model to obtain an air conditioner regulation and control model, designing a mask function according to the encryption amplitude and the encryption offset; the communication masking module is used for masking the comfort level and the comfort level change rate according to a masking function and disclosing the masked comfort level and the masked comfort level change rate between adjacent air conditioners in the distributed communication network; and the air conditioner regulation and control module is used for controlling each air conditioner in the distributed communication network to adjust the actual power change rate of the air conditioner based on the masked comfort degree and the masked comfort degree change rate of the adjacent air conditioner of the air conditioner. According to the invention, on the premise of ensuring the privacy security of the user, accurate power grid power response and comfortable fair distribution are realized.

The invention relates to the technical field of computers, in particular to a beverage recommendation method and device, communication equipment and a storage medium. The method comprises the following steps: performing feature extraction on a biological feature set to obtain an extracted feature set; the feature set is encrypted, an encrypted feature vector is sent to a server, and the encrypted feature vector is used for indicating the server to obtain beverage recommendation information corresponding to the terminal equipment according to the encrypted feature vector and a season factor and sending the beverage recommendation information to the terminal equipment; and obtaining and displaying the beverage recommendation information sent by the server. By adopting the beverage recommendation method and device, an intelligent beverage recommendation mode with innovative visual experience can be provided through data interaction under the condition of protecting the privacy of the user, and the accuracy and intelligence of beverage recommendation are improved.

The invention discloses a multi-dimensional data privacy calculation method and device, equipment, a medium and a product, and relates to the field of data security and privacy protection. The method comprises the following steps: a user terminal obtains a user privacy level, performs personalized privacy budget allocation based on the user privacy level, and determines a privacy budget allocation result; the user terminal adaptively selects a privacy protection disturbance mechanism according to the adaptive privacy boundary and the privacy budget distribution result so as to perform local disturbance, and a disturbance data set formed by user multi-dimensional data disturbance is obtained; and the aggregation server side performs multi-dimensional data joint distribution estimation by using the disturbance data set to obtain a joint probability distribution estimation result. According to the method, user privacy protection, data utility precision control and system expandability are considered, and efficient and controllable multi-dimensional data privacy calculation is realized.

The embodiment of the application relates to a defecation state recognition method and device and an intelligent defecation system, the method comprising: obtaining weight sensing data collected by a user in a defecation process from a weight sensor; judging whether the user leaves the defecation device based on the weight sensing data; if the user leaves the defecation device, obtaining defecation sensing data collected for the current defecation; and recognizing the defecation state of the user based on the defecation sensing data to obtain defecation state information of the user. The embodiment of the application greatly improves the convenience and efficiency of detecting the defecation state of the user. The defecation sensing data is collected when the user leaves the defecation device, so that the privacy information of the user can be avoided in the collected information, and the degree of privacy protection of the user is improved.

A computing machine stores supervised computing resources and private computing resources. The computing machine facilitates tracking, by a supervision service, of activity of the computing machine with respect to a supervised computing resource of the supervised resources that is executing on the computing machine. The computing machine supports user privacy, by the supervision service, with respect to a private computing resource of the private computing resources that is executing on the computing machine.

The invention provides a model updating method, equipment and device and a storage medium. The method comprises the steps that a data analysis function entity of a terminal sends a first data collection request to a data collection function entity of the terminal; the data analysis function entity of the terminal receives first data provided by the data collection function entity of the terminal based on the first data collection request; and the data analysis function entity of the terminal performs reasoning based on the first data, and updates a local model based on a reasoning result. The local model of the terminal can have enough training data support, so that the problem that the reasoning performance of the model is influenced due to insufficient training data is avoided, and the reasoning performance of the model can be effectively improved. Moreover, local analysis and reasoning are carried out on the terminal side, so that the terminal is more intelligent, dependence on a network core is reduced, and privacy protection and data security of a user are enhanced.

The embodiment of the invention provides a risk behavior recognition method and device, equipment, a storage medium and a program product, and relates to the field of big data. The method comprises the following steps: receiving encrypted entity relationship data uploaded by a plurality of mechanisms; using a federated learning algorithm to generate fusion features of a plurality of pairs of transaction entities according to the encrypted data uploaded by the plurality of mechanisms; constructing a dynamic graph according to the fusion features of the plurality of pairs of transaction entities; dividing the dynamic graph into a plurality of sub-graphs; and identifying risk behaviors existing in each sub-graph. According to the method provided by the invention, by using the federated learning algorithm, the encrypted entity relationship data of the multiple mechanisms are fused, so that cross-mechanism risk behavior identification is effectively realized on the basis of guaranteeing the privacy of the user.

The application discloses a collaborative filtering recommendation method and system based on adaptive noise adding privacy protection, and the method comprises the following steps: suspected abnormal users are screened out by using a DBSCAN algorithm, and true abnormal users are determined by using a box chart, so that abnormal user data in a user data set is screened out; the similarity between each user is calculated, different sizes of noise are added according to the similarity value, and a user similarity noise matrix is constructed; an initial centroid is selected according to the user similarity noise matrix by using a k-means algorithm, and iterative updating is performed, so that user data clustering is realized; a recommendation list is obtained according to a neighbor set of a target user; and items are recommended to the target user according to the scores of the items in the recommendation list. Through the technical scheme of the application, the accuracy of the recommendation result is improved, the user privacy is protected, the problem that the clustering centroid deviates greatly is avoided, and the problem that noise is continuously accumulated in the iteration process to cause the final recommendation result to be inaccurate is avoided to a certain extent.