43 results about "Black list" patented technology

The invention relates to the technical field of artificial intelligence, and discloses a method and a system for realizing multi-round questions and answers under security consultation based on a large model, and the method comprises the steps: receiving security consultation data of a target user to generate a dynamic user portrait of the target user, analyzing a first-level intention label and a first-level intention confidence coefficient of the security consultation data, and generating a second-level intention confidence coefficient; determining a second-level intention label of the target user; outputting structured query parameters of the target user; outputting dual-channel data of the security consultation data, and identifying a keyword blacklist, implicit commitments and undisclosed risks of the dual-channel data to filter the dual-channel data to obtain filtered multi-modal knowledge fragments; and generating a multi-dimensional answer of the target user to generate a multi-strategy model and a multi-strategy adjustment parameter and a user portrait update parameter of the dynamic user portrait, and executing multi-round questions and answers of the target user based on the multi-strategy adjustment parameter and the user portrait update parameter. According to the invention, the suggestion accuracy and personalization under security consultation can be improved.

The invention discloses a method and system for reducing the outbound complaint rate, and the method comprises the steps: (1), selecting an outbound number according to the priority of customer attribution, provincial meeting, in-province and business city in a pre-call stage, firstly filtering the number marked as promotion / fraud, and then taking the number with the highest historical call completing rate; determining whether to call out or not according to historical complaint portraits of clients, business blacklists and a seven-day dynamic frequency algorithm; (2) an in-speech stage: collecting speech in real time, fusing and judging complaint risks by utilizing four-dimensional acoustic features and a BERT semantic model finely adjusted by an outbound corpus, and automatically hanging up once a risk index is greater than or equal to 160; and (3) in a post-conversation stage, performing double scoring on conversation texts and acoustic features, if the total score is greater than or equal to 70, writing clients into a blacklist, setting a 180-day cold and quiet period, and automatically removing complaints without newly added complaints at the end of the period. According to the method, the call-out mode of processing before and after calls is adopted, the risk judgment model and the AI technology are fully utilized, the complaint risk is successfully reduced, and the cost is saved.

The application discloses a hot ticket card blacklist generation method and device and a medium, relates to the technical field of computers, and comprises the following steps: determining a quantity threshold of a hot ticket card blacklist; sorting each initial ticket card according to the time attribute of each initial ticket card to obtain a first sorting result, and screening a first candidate ticket card from the first sorting result; if the first ticket card quantity of the first candidate ticket card is not less than the quantity threshold, determining a target ticket card from the first candidate ticket card according to a priority rule; if the first ticket card quantity is less than the quantity threshold, sorting each initial ticket card according to the balance information in each initial ticket card to obtain a second sorting result, and screening a second candidate ticket card from the second sorting result; screening the target ticket card from the first candidate ticket card and the second candidate ticket card, and taking the target ticket card as the hot ticket card blacklist. The application can accurately obtain a hot ticket card blacklist with higher activity, and can reduce the memory resource demand on the memory database of a gate.

The embodiment of the invention discloses an article circulation alarm information generation method and device, electronic equipment and a readable medium. A specific embodiment of the method comprises the following steps: in response to received missed scanning prompt information of unmarked personnel, obtaining a missed scanning personnel video frame sequence and article information, and executing the following generation steps: obtaining a personnel feature vector data set; pedestrian re-identification processing is carried out on the video frame sequence of the scanning-missing persons to obtain human body feature vector information of the scanning-missing persons; performing matching and binding processing on the human body feature vector information of the missing scanning personnel to obtain information of the missing scanning personnel; adding the scanning missing person information to a preset blacklist library; acquiring a person face image of the visited authorized user; performing facial feature extraction processing on the person facial image to obtain facial feature vector information; generating risk level information; and generating alarm information and sending the alarm information to a preset receiving end. According to the embodiment, the user experience is improved.

The embodiment of the invention provides a digital key access authentication method and device, terminal equipment and a storage medium. According to the method, a digital key is generated through a cloud end, key certification information is generated at the same time, when a vehicle machine end receives a digital key access request of an equipment end, if it is judged that the digital key is not an accessed effective key, the vehicle machine end requests the equipment end to provide the key certification information, and the digital key is verified according to the key certification information; and after the verification is passed, notifying the equipment end to delete the locally stored key certification information. Compared with the prior art that the in-vehicle terminal cannot intercept the vehicle use permission of the key which does not fail in time, by adding the key first access verification mechanism, the deleted key can be effectively prevented from obtaining the vehicle use permission again, the safety of the vehicle is improved, and compared with a blacklist mechanism, more storage resources do not need to be occupied.

The application relates to the technical field of software development, in particular to a non-invasive dynamic burying point system for an application, an implementation method and a storage medium. A starting class is defined for a burying point application, the starting class is connected to a process after the burying point application is started, and a transformer module encapsulated into a java agent is loaded. Compared with an existing annotation-based invasive burying point scheme and a static instrumentation burying point scheme, the application is effective for any jvm application, meanwhile, targeted burying point optimization of common framework and tool class libraries and progressive burying point processing of specific scenes are considered; a classification process is generally mainly aimed at methods of the application itself, and existing classification can be generally adopted for widely used library classes. When the application needs to change or cancel burying point logic, only the black and white list classification and the processor associated with the white list need to be changed, and the server side is re-calculated according to the black list to obtain the bloom filter parameter, so that the operation is simple, the burying point is flexible, and the application is not aware of the burying point.

The invention relates to the technical field of computers, in particular to a Java back-end log real-time analysis and abnormity early warning system and method.The system comprises a plurality of modules for cooperative work: a log collection module obtains and structuralizes Java application log data in real time and stores the data in a distributed system; the trend analysis module constructs a trend chart based on a time sequence; the threshold management module dynamically calculates early warning, abnormal and serious abnormal thresholds of each interface; the anomaly detection module determines an anomaly level by comparing the trend change value and the ratio with a threshold value; the feature management module extracts abnormal log feature values and compares the abnormal log feature values with a blacklist library to filter false alarms; the alarm notification module executes graded alarm according to the abnormal grade and the preset authority, and collects and feeds back to the optimization system; according to the system, through multi-dimensional trend analysis and dynamic threshold management, the false alarm rate and the missing report rate are effectively reduced, the anomaly detection accuracy is improved, and stable operation of a Java back-end application is guaranteed.

The invention discloses a software uninstalling method and device of a Windows operating system, a storage medium and a terminal, relates to the technical field of software engineering, and mainly aims to solve the problem of low software uninstalling efficiency under the existing Windows operating system. The method mainly comprises the following steps: monitoring a system event in real time to capture process event information of a software installation process; in response to a process ending event of the software installation process, performing multi-dimensional feature matching on the process event information and software matching information of blacklist software, so as to take the installed software as target software under the condition of the installed software corresponding to the process event information; and performing reverse cleaning operation on the target software according to the software matching information and the process event information of the target software so as to unload the target software. The method is mainly used for automatically uninstalling software.

The invention provides an information system red-black list intelligent evaluation method and system based on multi-dimensional data. The method comprises the following steps: acquiring data, namely acquiring multi-dimensional data including performance indexes, log data, user behavior data and business data from a target information system in real time; preprocessing the data, and cleaning the collected multi-dimensional data, including abnormal value processing, missing value completion and data standardization; feature engineering is carried out, evaluation features are extracted and screened from the preprocessed data, and feature weights are dynamically distributed based on business scenes; performing intelligent model evaluation, classifying the information system by utilizing a trained machine learning classification model and combining the dynamically distributed feature weight, and outputting a red list representing a high-performance system and a black list representing a low-efficiency system; and optimization execution: according to the red list and the black list, automatically triggering optimization processing for the black list. According to the method, comprehensiveness, dynamic self-adaptability and automation of the evaluation process are realized.

The invention relates to the technical field of artificial intelligence, can be applied to the fields of financial science and technology and medical health, and discloses an anti-fraud method, device, equipment and medium based on voiceprint recognition, and the method comprises the steps: obtaining a call voice signal, and processing the call voice signal to obtain a multi-dimensional voiceprint feature; inputting the multi-dimensional voiceprint features into a deep learning model for optimization to obtain optimized voiceprint features; detecting whether voiceprint early warning conditions are met according to the optimized voiceprint features and a constructed voiceprint template library; and if the voiceprint early warning condition is met, performing fraud judgment through a decision tree algorithm according to the call voice signal, the optimized voiceprint feature and a fraud blacklist voiceprint library to obtain a fraud judgment result. And the anti-fraud accuracy is improved.

The invention provides an operation activity multi-account cheating identification method and system based on graph propagation. The method comprises the steps that a client side collects user behavior data and reports the data, a server side constructs a heterogeneous user relation graph based on the data and gives a cheating risk weight to an edge, the cheating probability is calculated by using a PageRank algorithm based on a seed blacklist user and is subjected to hierarchical disposal, and the client side dynamically adjusts operation activity interaction logic according to a hierarchical result; according to the system, the method is realized through cooperation of the client SDK module, the server graph construction and propagation module, the hierarchical decision module and the client anti-cheating execution module. According to the method, the defects that an existing supervision model is high in labeling cost, large in recognition delay and incapable of achieving front-end real-time intervention are effectively overcome, and the method is scientific, efficient and good in transverse transplantation capacity.

This invention discloses an intelligent generation method and apparatus for well logging acquisition construction design. The method includes: determining a matching template from a pre-established well logging acquisition construction design template library based on the well logging task category, well type, and well shape in the well logging acquisition task notification; using a well logging acquisition construction information knowledge graph based on a pre-established well logging team information library to screen qualified well logging teams that match the well logging task category, well type, and well shape in the notification and whose relevant information is not on a set blacklist; and using the knowledge graph to extract the information of qualified well logging teams required for the template from the well logging task notification and the well logging team information library, and filling this information into the template to generate the well logging acquisition construction design. This method enables intelligent requirements analysis for well logging acquisition construction design, improving the quality and timeliness of design preparation.

The invention provides a dynamic adaptive multilevel federated learning backdoor attack defense method, which comprises the following steps of: 1, feature extraction and clustering: extracting client model update features from cosine distance and normalized update energy dimensions, and clustering by using a DBSCAN algorithm; 2, model output classification: marking clients as possible malicious or benign clients according to model output; 3, dynamic weight aggregation: scoring the screened client model according to the classification statistics and the client aggregation result, and determining the weight of the client model in the global model according to the score; 4, directional rollback detoxification: after the global model is aggregated, analyzing the parameter update quantity of the client to identify abnormal update, and adaptively adjusting towards the direction of the initial model to eliminate the back door influence; and 5, self-adaption blacklist: quantifying historical malicious behaviors of the client through the scoring dictionary, and pre-filtering model parameters transmitted by the client in an aggregation preparation stage. According to the method, a multi-level defense system is constructed, and the confrontation adaptability and the security defense capability of the federal learning framework are remarkably improved.

The invention relates to the technical field of matrix multiplication, in particular to a sparse matrix-dense matrix multiplication acceleration system and method. The system comprises a sparse structure analyzer, a structure weighting scheduler, a plurality of PE rows and a runtime feedback controller. The sparse structure analyzer extracts features of each row of the sparse matrix and calculates a structure cost score; the structure weighting scheduler is combined with the PE row real-time load to distribute the PE row with the minimum comprehensive cost for the matrix row; executing scalar-vector multiply-accumulate by the PE array in parallel; and during operation, the feedback controller generates a blacklist bitmap and a splitting signal through performance monitoring, and dynamically adjusts task allocation. Parallel data reading and load balancing processing and calculation are carried out through a three-stage assembly line. According to the method, the calculation efficiency and the energy efficiency ratio of sparse-dense matrix multiplication are effectively improved, and then the data processing speed of the CPU / GPU is improved.

The invention discloses a financial business process management system and method, a storage medium and electronic equipment, and the method comprises the steps: building a product recommendation model based on a bank admission rule base; the dynamic allocation engine is used for allocating signing coordination personnel by adopting a multi-strategy rotation algorithm; the desensitization data reconciliation module supports fuzzy matching of various field combinations; the post-loan risk control platform integrates voucher auditing, channel risk grading and merchant blacklist linkage functions; the examination supervision module is used for realizing screen switching prevention monitoring and abnormal behavior detection; the activity management unit is used for completing closed-loop management of issuing and cancelling after verification of the immediate reduction money; the channel evaluation matrix is used for calculating a five-dimensional risk index including the expenditure rate deviation degree and reject ratio estimation; the input tracking system updates the bank approval progress in real time and synchronizes the bank approval progress to the business end; and the data center stores customer portraits, the approval knowledge base and third-party credit investigation data in a unified manner, and the financial business process management system greatly reduces account checking time consumption.

The system and method can assess the merchant risk level on a more continuous metric scale rather than a binary classification. This can produce a continuous risk score that is proportional to the likelihood that the merchant is risky, effectively solving the gray-shade problem encountered by traditional blacklist methods. The continuous risk score feature provides greater flexibility because it allows the payment network to make dynamic pricing decisions (known as interchange optimization) based on the merchant risk level. Using collective intelligence from transactions across payment networks, the system and method can assess the merchant risk level with high accuracy. The system and method may be particularly beneficial for small merchants with lower transaction volumes because even a few fraudulent transactions can easily place them in the high-risk merchant category. In addition, the system and method can help payment processing networks make better decisions on cross-border transactions.

The invention discloses a gate ticket checking method and device and a medium, and relates to the technical field of computers, and the method comprises the steps: determining a number threshold value of a hot ticket card blacklist; sorting the initial ticket cards according to the time attributes of the initial ticket cards to obtain a first sorting result, and screening out a first standby ticket card from the first sorting result; if the number of the first ticket cards of the first standby ticket cards is not smaller than a number threshold value, determining a target ticket card from the first standby ticket cards according to a priority rule; if the number of the first ticket cards is smaller than the number threshold value, sorting the initial ticket cards according to balance information in the initial ticket cards to obtain a second sorting result, and screening out a second standby ticket card from the second sorting result; and screening out a target ticket card from the first standby vote card and the second standby vote card, and taking the target ticket card as a hot ticket card blacklist. According to the method and the device, the hot ticket card blacklist with relatively high activeness can be accurately obtained, and the memory resource demand on the gate memory database can be reduced.

The present disclosure provides a blacklist generation method and device, electronic equipment and storage medium, relates to the technical field of computers, and particularly relates to the technical field of artificial intelligence such as automatic driving, data processing and simulation testing. The specific implementation scheme is: in response to a simulation task initiation request, the current target blacklist is called; machines in the cluster located in the target blacklist are shielded; the simulation task is submitted to the cluster, and running result data of the simulation task is obtained; the running result data is mined to obtain a fault machine that has a problem in the simulation task; and the target blacklist is updated based on the fault machine. According to the scheme of the present disclosure, the blacklist can be automatically generated in resource scheduling, and then the machines in the current latest target blacklist can be shielded for next time resource scheduling, so that the stability and efficiency of the simulation task running are improved.

The embodiment of the invention discloses a screen projection control method and device, electronic equipment and a computer readable storage medium, and relates to the technical field of screen projection, and the method comprises the steps that screen projection equipment receives screen projection information sent by screen projection equipment, establishes screen projection connection with the screen projection equipment according to the screen projection information, and displays a user interface for the screen projection connection, and determining a target connection state of the projection screen connection according to an interface operation for the user interface, and controlling projection screen display according to the target connection state. According to the invention, the screen-projected device can complete the screen projection connection with the screen projection device without waiting for the interface operation of the user interface, so that the screen projection device is not in a connection timeout state due to the refusal, long-time non-processing or shielding of the screen-projected device, and the screen-projected device is prevented from being added to a blacklist of the screen projection device. The screen projection device can still search the screen-projected device during next screen projection, the convenience of screen projection of the device is improved, and the user experience is improved.

This invention provides a digital key access authentication method, apparatus, terminal device, and storage medium. The method generates key verification information simultaneously with the digital key generated in the cloud. When the vehicle's infotainment system receives a digital key access request from the device, if it determines that the digital key is not a valid key already accessed, it requests the device to provide key verification information and verifies the digital key based on this information. Upon successful verification, the system notifies the device to delete the locally stored key verification information. Compared to existing technologies where the vehicle's infotainment system cannot intercept vehicle access permissions from keys that have not expired in time, this method, by adding a key initial access verification mechanism, effectively prevents deleted keys from regaining vehicle access, improving vehicle security. Furthermore, it does not require additional storage resources compared to a blacklist mechanism.

The present invention relates to systems and methods for query-based analysis of optimization of a supply chain. In some embodiments, a query is received within a trusted environment of an optimization system, wherein the query includes a natural language expression. The query is supplied to an Artificial Intelligence (AI) transformation system, wherein the AI transformation system converts the query into a function and parameter statement. The function and parameter statement are translated into at least one API call and function, which are executed to generate results. The results are then outputted. The AI transformation system is locally trained on the optimization system. In some embodiments, the query may be augmented with information. The information includes at least one of query source information and query intent information. Additionally, sensitive data in the query may be identified. The sensitive information is identified by blacklist, whitelist or semantic analysis.

The invention provides a multi-source task-oriented material flow synthesis method and device and a storage medium, and relates to the technical field of automatic material management. The method comprises the following steps: monitoring a towed form and verifying a drop point, correcting a node connection line, and receiving node attribute configuration information; a supplier blacklist is monitored based on a Rete algorithm, budget is monitored by means of a Boolean algebraic optimization algorithm, and an alarm is generated when abnormity occurs; based on the request import form, receiving configuration information, and updating and storing the configuration information; calling data connector configuration parameters, and synchronizing the form data and the parameters to a material platform. According to the method, a full-process automatic material management system is constructed, form node visual arrangement is realized, the operation threshold is reduced, the risk is accurately identified and early warning is performed, the data barrier is broken through, the construction efficiency is improved, the risk control capability and cross-platform collaboration are improved, and the problems of much manual intervention, slow response and data island in the traditional process are solved.

The application provides a cloud disk creation request scheduling method and device, electronic equipment and storage medium, and belongs to the technical field of cloud storage. The method comprises the following steps: determining a target disk creation type to which a target cloud disk creation request belongs based on historical disk creation records of a user account, different disk creation types have different disk creation rejection water levels, the disk creation rejection water level is the maximum water level when a cluster performs a cloud disk creation operation, and the water level is determined according to real-time data link QPS capability and snapshot loading capability of the cluster; obtaining a plurality of candidate clusters based on a target disk creation rejection water level corresponding to the target disk creation type, current water levels of the plurality of clusters and a cluster blacklist; determining a target cluster from the plurality of candidate clusters according to a cluster allocation mode corresponding to the target disk creation type; and sending the target cloud disk creation request to the target cluster. The application improves the performance of the cluster in the distributed storage system, avoids problems such as read-write lag of the cluster, and improves the success rate of cloud disk creation.

The invention discloses an ash silo logistics anomaly detection and anti-cheating management and control method, and relates to the technical field of logistics safety management and control, and the method comprises the steps: collecting the historical business data of a vehicle in an ash silo logistics system, and constructing a logistics association map with a cargo owner, a vehicle and a driver as vertexes; inputting the association graph into a graph neural network model to obtain a vehicle node anomaly association degree score and generate a cross-factory dynamic blacklist; performing risk scoring on the vehicle requesting to enter the parking lot and comparing with the blacklist to realize access control release or rejection control; the process is predicted and early warning or control is executed based on a digital twinborn model before loading operation; and aggregating each factory model through federal learning to generate a global risk scoring model and dynamically updating the blacklist. According to the invention, the abnormity identification accuracy can be improved, and cross-factory collaborative anti-cheating management and control are realized.

The present invention discloses a cross-institutional blacklist sharing method and system based on reusable obfuscation circuits, relating to the field of privacy computing. Specifically, the cross-institutional blacklist sharing method based on reusable obfuscation circuits involves uploading multi-party blacklist data to a blockchain trusted environment in ciphertext form. Three obfuscation techniques are then used to convert the circuits into reusable obfuscated circuits, enabling the obfuscated circuits to be reused, significantly reducing communication overhead. Finally, the data encryption and oblivious transmission process of the reusable obfuscation circuits are utilized to ensure the reliable and confidential execution of computing tasks. The cross-institutional blacklist sharing method based on reusable obfuscation circuits effectively protects data privacy from abuse and enhances the feasibility and effectiveness of cross-institutional blacklist data applications.

The invention discloses an identity verification method and device based on an SIM card, equipment, a medium and a product, and the method comprises the steps that a called SIM card sends an identity verification request message to a calling SIM card by receiving a call request of the calling SIM card; wherein the identity verification request message at least comprises the identity information of the called SIM card; receiving and analyzing an identity verification response message sent by the calling SIM card to obtain identity information of the calling SIM card; wherein the identity verification response message is sent by the calling SIM card after the identity information verification of the called SIM card is passed; and verifying the identity information of the calling SIM card. By adopting the embodiment of the invention, the problem of error verification of a traditional blacklist mechanism can be avoided, and the verification effect is improved.

The invention discloses a blacklist text matching method and device based on a twin network. The blacklist text matching method and device are used for improving matching effectiveness for blacklist texts. According to the scheme, the method comprises the steps of obtaining a to-be-matched text; performing local feature extraction on the to-be-matched text through a convolutional neural network module of the twin network to obtain local features of the to-be-matched text; performing global feature extraction on the local features of the to-be-matched text through a multi-head self-attention module of the twin network to obtain global features of the to-be-matched text; and determining a blacklist text meeting a preset similarity condition with the global feature of the to-be-matched text from a preset blacklist text feature library, the preset blacklist text feature library comprising a plurality of blacklist text features, and any blacklist text feature being obtained by performing feature extraction on the preset blacklist text through a twin network.

The application discloses a credit anti-fraud blacklist area identification system and method based on grid division, which comprises a data acquisition and preprocessing module for acquiring and preprocessing GPS coordinate data and face recognition images to obtain preprocessed GPS coordinate data and preprocessed image features, and acquiring business data; a gridding division and coding module for performing spatial discretization and grid coding according to the preprocessed GPS coordinate data to obtain GPS grids, and according to the business data, counting a plurality of indexes in each GPS grid; a feature engineering construction module for determining a plurality of feature vectors based on the plurality of indexes and the preprocessed image features; and a black grid identification model training and execution module for inputting the plurality of feature vectors into a trained black grid identification model to obtain a black grid identification result. The application can improve the accuracy of identifying credit anti-fraud blacklist areas.

The invention discloses an instruction monitoring method and device, equipment, a storage medium and a program product. The method comprises the steps of obtaining a user instruction in response to an instruction input operation of a user, and collecting biological characteristics of the user; identifying an emotional state of the user based on the biological characteristics, and determining a first risk level corresponding to the user instruction according to the emotional state and a preset risk mapping relationship; according to the reversibility of the system operation indicated by the user instruction, the influence range of the system operation indicated by the user instruction, the authority of the user and the instruction blacklist, performing security assessment on the user instruction, and determining a second risk level corresponding to the user instruction; and determining a risk level of the user instruction according to the first risk level and the second risk level, and executing a system operation corresponding to the risk level. By adopting the method, the risk level of the instruction can be determined from the two aspects of the instruction and the user emotion, and the accuracy of an instruction risk level judgment result is effectively improved.

An object of a system according to an embodiment is to effectively prevent telephone fraud for elderly people.SOLUTION: A system includes a context analysis unit, a black list filter unit, and an in-terminal AI processing unit. The context analysis unit analyzes a context. The blacklist filtering unit performs filtering using an existing telephone number blacklist. The in-terminal AI processing unit performs AI processing in the terminals.SELECTED DRAWING: Figure 1