A method, apparatus, equipment and medium for detecting enterprise anomalies
By fusing local access features into global features for anomaly detection in a distributed system, the problem of difficult cross-node data transmission is solved, detection efficiency is improved, and data privacy is protected.
Patent Information
- Application Number
- CN202610809213.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-05
- Publication Date
- 2026-11-17
AI Technical Summary
Existing technologies for enterprise anomaly detection suffer from low efficiency and high data processing requirements because the log formats and field definitions of each node are independent. This results in large amounts of raw log data being transmitted across nodes, difficulties in feature alignment, and independent log formats and field definitions.
In a distributed system, a central node sends feature acquisition requests to multiple assisting nodes. The assisting nodes extract local access features and feed them back to the central node for fusion to form global access features. Anomaly detection is then performed using an anomaly detection model to avoid cross-node raw data transmission.
It improves anomaly detection efficiency, reduces the data processing and transmission burden on the central node, and protects the local data privacy of assisting nodes.
Smart Images

Figure CN123122855A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of Internet technology, and in particular to a method, apparatus, equipment and medium for detecting enterprise anomalies. Background Technology
[0002] Enterprise anomaly detection is used to detect abnormal access behavior in an enterprise's business systems, such as unauthorized access or abnormal activity after an account has been deactivated. Currently, the main approach is to extract statistical features from historical enterprise access data and then use neural networks to identify abnormal access.
[0003] However, in actual deployments, enterprise access behavior is scattered across nodes of different business types, and the log formats and field definitions stored on each node are independent of each other. Traditional methods require aggregating the raw logs from each node into one place before extracting features uniformly. This results in a large amount of data being transmitted across raw logs from different nodes, and the heterogeneity of fields makes feature alignment difficult. It also places high demands on the computing power of the aggregation device, leading to low efficiency in enterprise anomaly detection. Summary of the Invention
[0004] This invention provides a method, apparatus, equipment, and medium for detecting enterprise anomalies, which can improve the efficiency of anomaly detection.
[0005] According to one aspect of the present invention, an embodiment of the present invention provides an enterprise anomaly detection method, characterized in that it is applied to the central node of a distributed system, the method comprising: The system sends the target enterprise's feature acquisition request to multiple assisting nodes with different business types; the distributed system includes: a central node and multiple assisting nodes. The system receives local access features from each of the assisting nodes and fuses these local access features to obtain global access features; wherein, the local access features are obtained by feature extraction from the access data of the target enterprise accessing the assisting nodes. Based on the global access characteristics, determine whether the target enterprise has any anomalies.
[0006] According to one aspect of the present invention, an embodiment of the present invention provides an enterprise anomaly detection method, characterized in that it is applied to a cooperating node in a distributed system, the method comprising: Received a request to obtain the target company's features; Based on the target enterprise's feature acquisition request, the local access features of the target enterprise are determined and fed back to the central node of the distributed system; the distributed system includes: a central node and multiple assisting nodes; wherein, the local access features are obtained by feature extraction from the access data of the target enterprise accessing the assisting nodes.
[0007] According to another aspect of the present invention, embodiments of the present invention also provide an enterprise anomaly detection device, characterized in that it is applied in the central node of a distributed system, the device comprising: The request sending module is used to send the target enterprise's feature acquisition request to multiple assisting nodes with different business types; the distributed system includes: a central node and multiple assisting nodes; The feature fusion module is used to receive local access features fed back by each of the assisting nodes and fuse the local access features to obtain global access features; wherein, the local access features are obtained by feature extraction of the access data of the target enterprise accessing the assisting nodes; An anomaly detection module is used to determine whether the target enterprise has any anomalies based on the global access characteristics.
[0008] According to another aspect of the present invention, embodiments of the present invention also provide an enterprise anomaly detection device, characterized in that it is applied in a collaborative node of a distributed system, the device comprising: The request receiving module is used to receive feature acquisition requests from the target enterprise. The feature extraction module is used to determine the local access features of the target enterprise based on the feature acquisition request of the target enterprise, and feed them back to the central node of the distributed system; the distributed system includes: a central node and multiple assisting nodes; wherein, the local access features are obtained by feature extraction from the access data of the target enterprise accessing the assisting nodes.
[0009] According to another aspect of the present invention, embodiments of the present invention also provide an enterprise anomaly detection device, the enterprise anomaly detection device comprising: At least one processor; and A memory that is communicatively connected to at least one processor; wherein, The memory stores a computer program that can be executed by at least one processor, such that the at least one processor is able to perform the enterprise anomaly detection method according to any embodiment of the present invention.
[0010] According to another aspect of the present invention, a computer-readable storage medium is provided, which stores computer instructions for causing a processor to execute and implement the enterprise anomaly detection method of any embodiment of the present invention.
[0011] According to another aspect of the present invention, a computer program product is provided, the computer program product comprising a computer program that, when executed by a processor, implements the enterprise anomaly detection method according to any embodiment of the present invention.
[0012] The technical solution of this invention obtains the characteristics of the target enterprise's access behavior in each assistance system by sending a feature acquisition request to the assistance system, and then integrates the features into global features for anomaly detection. This avoids the transmission of raw data across nodes, reduces the data processing and transmission burden of the central node, improves the efficiency of anomaly detection, and protects the privacy of local data on each assistance node.
[0013] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description
[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0015] Figure 1 This is a flowchart of an enterprise anomaly detection method provided by an embodiment of the present invention; Figure 2 This is a flowchart of an enterprise anomaly detection method provided by an embodiment of the present invention; Figure 3 This is a structural diagram of an enterprise anomaly detection device provided according to an embodiment of the present invention; Figure 4 This is a structural diagram of an enterprise anomaly detection device provided according to an embodiment of the present invention; Figure 5 This is a structural schematic diagram of an enterprise anomaly detection device provided in an embodiment of the present invention. Detailed Implementation
[0016] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0017] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0018] The acquisition, storage, and application of access data involved in the technical solutions of this invention comply with the provisions of relevant laws and regulations and do not violate public order and good morals.
[0019] Figure 1 This is a flowchart illustrating an enterprise anomaly detection method provided by an embodiment of the present invention. This embodiment is applicable to enterprise anomaly detection and can be applied to a central node in a distributed system. The method can be executed by an enterprise anomaly detection device, which can be implemented in hardware and / or software. This enterprise anomaly detection device can be configured in a server.
[0020] See Figure 1 The enterprise anomaly detection methods shown include: S101. Send the target enterprise's feature acquisition request to multiple assisting nodes with different business types; the distributed system includes: a central node and multiple assisting nodes.
[0021] In this context, a distributed system refers to a system where data is distributed across multiple nodes. The central node can be used for global data processing. Supporting nodes can be used for local data processing. The target enterprise can be the enterprise to be detected for anomalies. A feature acquisition request can be a request to obtain and access data features.
[0022] A distributed system includes multiple assisting nodes and a central node. The assisting nodes have different business types. For example, assisting nodes may include: an electricity payment system, a voucher issuance system, and an administrative management system, etc. The central node may include: a central processing system.
[0023] The central node generates a feature acquisition request based on information such as the target company's corporate identifier and sends the request to each assisting node. This enables the assisting nodes to obtain the target company's access data for a target time period on their respective nodes, extract features from the access data to obtain local access features, and then feed these local access features back to the central node.
[0024] S102. Receive the local access features fed back by each of the assisting nodes and fuse the local access features to obtain global access features; wherein, the local access features are obtained by feature extraction of the access data of the target enterprise accessing the assisting nodes.
[0025] Local access characteristics can be the access characteristics of access data for a specific assisting node. Global access characteristics can be the access characteristics of the target enterprise for all assisting nodes. Access data can be request and response data.
[0026] Multiple local access features fed back by each assisting node are fused, and the fused global access feature has the same feature dimension as the local access features. For example, if the local access feature is an n*m feature matrix, then the global access feature is also an n*m feature matrix.
[0027] S103. Based on the global access characteristics, determine whether the target enterprise is abnormal.
[0028] This includes anomaly detection of global access characteristics. By comparing the target enterprise's access behavior with normal patterns, and determining whether the target enterprise exhibits abnormal access behavior, an assessment is made to determine whether the target enterprise faces security, compliance, or operational risks. Anomalies to the target enterprise may include: account theft, system failure, and / or system configuration errors.
[0029] In an optional embodiment, determining whether the target enterprise is abnormal based on the global access features includes: inputting the global access features into an anomaly detection model deployed on the central node, performing anomaly detection on the target enterprise, and obtaining anomaly detection results; and determining whether the target enterprise is abnormal based on the anomaly detection results.
[0030] The anomaly detection result can be the outcome of anomaly detection. In some embodiments, the anomaly detection result can be the probability that the target enterprise exhibits anomalies. The anomaly detection model can be a model used to detect anomalies in the target enterprise. The anomaly detection model is deployed on each node of the distributed system.
[0031] The anomaly detection model, after pre-training, establishes a correlation between access features and enterprise status, enabling it to predict enterprise status based on access features. Global access features are input into the anomaly detection model deployed at the central node. Based on these global access features, the model predicts the status of the target enterprise, yielding the anomaly detection result.
[0032] If the anomaly detection result indicates the probability that the target company has an anomaly, then the magnitude of this probability is used to determine whether the target company has an anomaly.
[0033] It is evident that by inputting global access features into the model, anomaly detection of target enterprises can be performed based on the correlation between historical access patterns learned by the model and enterprise status, thereby improving the accuracy and efficiency of anomaly identification.
[0034] In an optional embodiment, after determining whether the target enterprise has an anomaly based on the anomaly detection results, the method further includes: when it is determined that the target enterprise has an anomaly, obtaining intermediate detection results fed back by each of the assisting nodes; the intermediate detection results are obtained by performing anomaly detection on the local access features through anomaly detection models deployed on the assisting nodes; and determining the business dimension of the target enterprise that has an anomaly based on the intermediate detection results.
[0035] Intermediate detection results can be the results obtained by each assisting node in detecting anomalies in the target enterprise.
[0036] Each assisting node acquires access data from the target enterprise to that node and inputs this data into an anomaly detection model deployed locally on that node. The anomaly detection model extracts features from the access data to obtain local access features, performs anomaly detection on these local features, and obtains intermediate detection results for the target enterprise. These intermediate detection results indicate whether the target enterprise's access behavior to a particular assisting node is abnormal, or the probability of such anomalies.
[0037] When an anomaly is identified in the target enterprise based on global access characteristics, anomaly detection is performed on each assisting node based on local access characteristics, and the intermediate detection results are analyzed. This determines which nodes exhibit abnormal access behavior of the target enterprise. For example, the intermediate detection results returned by each assisting system represent the probability that the target enterprise is abnormal. When the global access characteristics determine that the target enterprise is abnormal, the business types of the assisting nodes with a probability greater than 50% in the intermediate detection results are identified as abnormal business dimensions.
[0038] It is evident that by determining the abnormal business dimensions of the target enterprise based on the intermediate detection results fed back by each assisting node, it is easier to analyze the reasons for the abnormality of the target enterprise and improve the efficiency of enterprise risk control verification.
[0039] In an optional embodiment, the method further includes: generating global model parameters based on the local model parameters received from each of the assisting nodes; updating the anomaly detection model of the central node based on the global model parameters; distributing the global model parameters to each assisting node so that each assisting node updates the parameters of its loaded anomaly detection model based on the model parameters; and obtaining the local access features by extracting features from the access data of the target enterprise accessing the assisting node through the anomaly detection model deployed on the assisting node.
[0040] Local model parameters can be model parameters obtained from training the anomaly detection model on the assisting node. Global model parameters can be model parameters generated based on all local model parameters.
[0041] The central node aggregates all local model parameters uploaded by assisting nodes and performs aggregation calculations to generate global model parameters. The central node uses these integrated global model parameters to update the parameters of its anomaly detection model and then sends the unified global model parameters back to all assisting nodes. Each assisting node receives the global model parameters and synchronously updates its locally deployed anomaly detection model, ensuring consistent model parameters across the entire system.
[0042] In some embodiments, the average value of the local model parameters fed back by each assisting node can be calculated to obtain the global model parameters.
[0043] The local access characteristics of the assisting nodes are obtained by directly extracting the characteristics of the target enterprise's access data from the locally deployed anomaly detection model. The data is stored locally on each assisting node without the need for the original data to be transmitted across nodes, thus avoiding the risk of data leakage caused by the centralized uploading of the original business data.
[0044] As can be seen, by aggregating the local model parameters of each assisting node and integrating them to generate global model parameters, the model parameters of the center and each assisting node are updated, the model configuration of all nodes is unified, and the defects of inconsistent parameters and different detection standards caused by independent training of each node are avoided.
[0045] The technical solution of this invention obtains the characteristics of the target enterprise's access behavior in each assistance system by sending a feature acquisition request to the assistance system, and then integrates the features into global features for anomaly detection. This avoids the transmission of raw data across nodes, reduces the data processing and transmission burden of the central node, improves the efficiency of anomaly detection, and protects the privacy of local data on each assistance node.
[0046] Figure 2This is a flowchart illustrating an enterprise anomaly detection method provided by an embodiment of the present invention. This embodiment is applicable to enterprise anomaly detection and can be used as a cooperating node in a distributed system. The method can be executed by an enterprise anomaly detection device, which can be implemented in hardware and / or software. This enterprise anomaly detection device can be configured in a server.
[0047] It should be noted that for parts not described in detail in the embodiments of the present invention, please refer to the descriptions in other embodiments.
[0048] See Figure 2 The enterprise anomaly detection methods shown include: S201, Received a request to obtain the characteristics of the target company.
[0049] The assisting nodes and the central node communicate with each other. When the central node sends a feature acquisition request, the assisting nodes receive the feature acquisition request from the target enterprise.
[0050] S202. Based on the feature acquisition request of the target enterprise, determine the local access features of the target enterprise and feed them back to the central node of the distributed system; the distributed system includes: a central node and multiple assisting nodes; wherein, the local access features are obtained by feature extraction of the access data of the target enterprise accessing the assisting nodes.
[0051] This involves parsing the feature acquisition request to identify the target company. For the target company, local access characteristics are then determined.
[0052] In a distributed system—that is, a system consisting of a central node and multiple assisting nodes—when analyzing the access behavior of a target enterprise, the assisting nodes do not directly report the raw access logs to the central node. Instead, each assisting node generates local access features based on the data generated by the target enterprise's access to its own node, using feature extraction techniques. This feature reflects the target enterprise's access behavior pattern from the assisting node's perspective. Subsequently, each assisting node only feeds back this local access feature to the central node. This reduces the data processing and transmission burden on the central node, protects the privacy of local data on each assisting node, and enables the central node to integrate local features from multiple assisting nodes, thereby forming a holistic understanding of the target enterprise's access patterns.
[0053] Local access characteristics are used to describe the target enterprise's access behavior to a single assisting node. The dimensions of the access data acquired by the assisting node can include: access frequency, access time, and requested resource type. The dimensions of local access characteristics can include: data volume characteristics, including: average data volume per request, peak data volume, and total cumulative transmission traffic; error rate characteristics, including: the proportion of request failures and timeouts, and access concentration.
[0054] Optionally, the extracted local access features can be homomorphically encrypted, and the encrypted local access features can be fed back to the central node.
[0055] In an optional embodiment, the method further includes: acquiring sample data, the sample data including sample access data and standard enterprise states corresponding to the sample access data; inputting the sample data into an anomaly detection model, extracting features from the sample access data to obtain node access features; performing anomaly detection on the node access features to obtain state detection results; adjusting the model parameters of the anomaly detection model based on the difference between the state detection results and the standard enterprise states to obtain local model parameters; sending the local model parameters to the central node so that the central node generates global model parameters based on the local model parameters fed back by each of the assisting nodes; receiving the global model parameters sent by the central node, and updating the model parameters of the anomaly detection model based on the global model parameters.
[0056] Here, sample access data can be access data used for model training. Standard enterprise status can be an abnormal or normal state of the enterprise. Node access features can be features extracted by assisting nodes from the sample access data during the model training phase. Node access features are used to describe the enterprise's access behavior characteristics towards assisting nodes.
[0057] Each assisting node acquires local sample data. This sample data includes sample access data and the corresponding standard enterprise status, i.e., the real label. The assisting node inputs the sample data into its local anomaly detection model. First, it extracts features from the sample access data to obtain node access features; then, it performs anomaly detection on these features to output the status detection result. Next, it compares the status detection result with the standard enterprise status, using this difference to adjust the model parameters and obtain the local model parameters.
[0058] Each assisting node sends its local model parameters to the central node. The central node aggregates the local model parameters from all assisting nodes, generates global model parameters, and distributes them to all assisting nodes. Each assisting node then uses the global model parameters to update its local anomaly detection model.
[0059] Optionally, after each assisting node receives the global model parameters and updates its local anomaly detection model, it repeats the process of acquiring sample data for model training until the difference between the state detection results of each assisting node and the standard enterprise state is less than a preset difference value, at which point the model training is considered complete.
[0060] As can be seen, by using local data to train the model at the assisting nodes, sending the updated parameters after model training to the central node, and updating the local anomaly detection model of the assisting nodes based on the global parameters returned by the central node, the assisting nodes can collaboratively train a unified anomaly detection model without sharing the original access data. This not only protects the data privacy of each node, but also improves the model's ability to detect abnormal states of the entire enterprise.
[0061] The technical solution of this invention determines the local access characteristics of the target enterprise based on the received feature acquisition request of the target enterprise, and feeds it back to the central node. This eliminates the need for each assisting node to share the original access data, protects the privacy of local data on each assisting node, and improves the security of the system.
[0062] Figure 3 This is a schematic diagram of an enterprise anomaly detection device provided in an embodiment of the present invention. This embodiment of the present invention is applicable to situations where anomaly detection is performed on an enterprise to be detected. It is applied to the central node of a distributed system. This device can execute enterprise anomaly detection methods and can be implemented in hardware and / or software.
[0063] See Figure 3 The enterprise anomaly detection device shown includes: The request sending module 301 is used to send the target enterprise's feature acquisition request to multiple assisting nodes with different business types; the distributed system includes: a central node and multiple assisting nodes; The feature fusion module 302 is used to receive local access features fed back by each of the assisting nodes and fuse the local access features to obtain global access features; wherein, the local access features are obtained by feature extraction of the access data of the target enterprise accessing the assisting nodes; The anomaly detection module 303 is used to determine whether the target enterprise has any anomalies based on the global access characteristics.
[0064] The technical solution of this invention obtains the characteristics of the target enterprise's access behavior in each assistance system by sending a feature acquisition request to the assistance system, and then integrates the features into global features for anomaly detection. This avoids the transmission of raw data across nodes, reduces the data processing and transmission burden of the central node, improves the efficiency of anomaly detection, and protects the privacy of local data on each assistance node.
[0065] In an optional embodiment, the anomaly detection module 303 includes: The model input unit is used to input the global access features into the anomaly detection model deployed on the central node to perform anomaly detection on the target enterprise and obtain anomaly detection results; The result output unit is used to determine whether the target enterprise has any abnormalities based on the anomaly detection results.
[0066] In an optional embodiment, an anomaly dimension detection module is further included, comprising: The intermediate result acquisition unit is used to acquire intermediate detection results fed back by each of the assisting nodes when it is determined that the target enterprise has an anomaly; the intermediate detection results are obtained by performing anomaly detection on the local access features through an anomaly detection model deployed on the assisting nodes. An anomaly dimension determination unit is used to determine, based on the intermediate detection results, the business dimensions of the target enterprise that are abnormal.
[0067] In an optional embodiment, a parameter update module is further included, comprising: The parameter fusion unit is used to generate global model parameters based on the local model parameters sent by each of the assisting nodes. The first parameter update unit is used to update the anomaly detection model of the central node according to the global model parameters; The parameter distribution unit is used to distribute the global model parameters to each assisting node, so that each assisting node updates the parameters of its loaded anomaly detection model according to the model parameters; the local access features are obtained by extracting features from the access data of the target enterprise accessing the assisting node through the anomaly detection model deployed on the assisting node.
[0068] The enterprise anomaly detection device provided in this embodiment of the invention can execute the enterprise anomaly detection method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the enterprise anomaly detection method.
[0069] Figure 4 This is a schematic diagram of an enterprise anomaly detection device provided in an embodiment of the present invention. This embodiment of the present invention is applicable to situations where anomaly detection is performed on an enterprise to be detected, and is applied to a cooperating node in a distributed system. This device can execute enterprise anomaly detection methods and can be implemented in hardware and / or software.
[0070] See Figure 4 The enterprise anomaly detection device shown includes: The request receiving module 401 is used to receive the feature acquisition request of the target enterprise; The feature extraction module 402 is used to determine the local access features of the target enterprise based on the feature acquisition request of the target enterprise, and feed them back to the central node of the distributed system; the distributed system includes: a central node and multiple assisting nodes; wherein, the local access features are obtained by feature extraction of the access data of the target enterprise accessing the assisting nodes.
[0071] The technical solution of this invention determines the local access characteristics of the target enterprise based on the received feature acquisition request of the target enterprise, and feeds it back to the central node. This eliminates the need for each assisting node to share the original access data, protects the privacy of local data on each assisting node, and improves the security of the system.
[0072] In an optional embodiment, a model training module is also included, comprising: A sample acquisition unit is used to acquire sample data, the sample data including: sample access data and the standard enterprise status corresponding to the sample access data; The node feature extraction unit is used to input the sample data into the anomaly detection model, extract features from the sample access data, and obtain node access features. A state detection unit is used to perform anomaly detection on the node access characteristics and obtain a state detection result. The parameter adjustment unit is used to adjust the model parameters of the anomaly detection model according to the difference between the state detection result and the standard enterprise state, so as to obtain local model parameters; The parameter feedback unit is used to send the local model parameters to the central node, so that the central node generates global model parameters based on the local model parameters fed back by each of the assisting nodes. The second parameter update unit is used to receive the global model parameters sent by the central node and update the model parameters of the anomaly detection model according to the global model parameters.
[0073] The enterprise anomaly detection device provided in this embodiment of the invention can execute the enterprise anomaly detection method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the enterprise anomaly detection method.
[0074] Figure 5 A schematic diagram of the structure of an enterprise anomaly detection device 500 that can be used to implement embodiments of the present invention is shown.
[0075] like Figure 5As shown, the enterprise anomaly detection device 500 includes at least one processor 501 and a memory, such as a read-only memory 502 or a random access memory 503, communicatively connected to the at least one processor 501. The memory stores computer programs executable by the at least one processor. The processor 501 can perform various appropriate actions and processes based on the computer program stored in the read-only memory 502 or loaded from storage unit 508 into the random access memory 503. The random access memory 503 can also store various programs and data required for the operation of the enterprise anomaly detection device 500. The processor 501, read-only memory 502, and random access memory 503 are interconnected via a bus 504. An input / output interface 505 is also connected to the bus 504.
[0076] Multiple components in the enterprise anomaly detection device 500 are connected to an input / output interface 505, including: an input unit 506, such as a keyboard or mouse; an output unit 507, such as various types of monitors or speakers; a storage unit 508, such as a hard disk or optical disk; and a communication unit 509, such as a network interface card (NIC), modem, or wireless transceiver. The communication unit 509 allows the enterprise anomaly detection device 500 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.
[0077] Processor 501 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 501 include, but are not limited to, central processing units, graphics processing units, various special-purpose artificial intelligence computing chips, various processors running machine learning model algorithms, digital signal processors, and any suitable processor, controller, microcontroller, etc. Processor 501 performs the various methods and processes described above, such as enterprise anomaly detection methods.
[0078] In some embodiments, the enterprise anomaly detection method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed onto the enterprise anomaly detection device 500 via read-only memory 502 and / or communication unit 509. When the computer program is loaded into random access memory 503 and executed by processor 501, one or more steps of the enterprise anomaly detection method described above may be performed. Alternatively, in other embodiments, processor 501 may be configured to execute the enterprise anomaly detection method by any other suitable means (e.g., by means of firmware).
[0079] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays, application-specific integrated circuits (ASICs), application-specific standard products (ASICs), systems-on-a-chip (SoCs), complex programmable logic devices, computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.
[0080] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.
[0081] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory, flash memory, optical fiber, portable compact disk read-only memory, optical storage devices, magnetic storage devices, or any suitable combination thereof.
[0082] To provide user interaction, the systems and techniques described herein can be implemented on an operational detection device. This enterprise anomaly detection device includes: a display device (e.g., a cathode ray tube or liquid crystal monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the enterprise anomaly detection device. Other types of devices can also be used to provide user interaction; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).
[0083] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.
[0084] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product within the cloud computing service system. This addresses the shortcomings of traditional physical hosts and virtual private servers, such as high management difficulty and weak business scalability.
[0085] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.
[0086] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.
Claims
1. A method for detecting enterprise anomalies, characterized in that, When applied to the central node of a distributed system, the method includes: The system sends the target enterprise's feature acquisition request to multiple assisting nodes with different business types; the distributed system includes: a central node and multiple assisting nodes. The system receives local access features from each of the assisting nodes and fuses these local access features to obtain global access features; wherein, the local access features are obtained by feature extraction from the access data of the target enterprise accessing the assisting nodes. Based on the global access characteristics, determine whether the target enterprise has any anomalies.
2. The method according to claim 1, characterized in that, The step of determining whether the target enterprise has any anomalies based on the global access characteristics includes: The global access features are input into the anomaly detection model deployed on the central node to perform anomaly detection on the target enterprise and obtain anomaly detection results. Based on the anomaly detection results, it is determined whether the target company has any anomalies.
3. The method according to claim 2, characterized in that, After determining whether the target enterprise has any anomalies based on the anomaly detection results, the method further includes: When it is determined that the target enterprise is abnormal, the intermediate detection results fed back by each of the assisting nodes are obtained; the intermediate detection results are obtained by performing anomaly detection on the local access features through the anomaly detection model deployed on the assisting nodes. Based on the intermediate detection results, it was determined that the target company had abnormal business dimensions.
4. The method according to claim 2, characterized in that, Also includes: Generate global model parameters based on the local model parameters sent by each of the assisting nodes; The anomaly detection model of the central node is updated based on the global model parameters; The global model parameters are distributed to each assisting node so that each assisting node updates the parameters of its loaded anomaly detection model according to the model parameters; the local access features are obtained by extracting features from the access data of the target enterprise accessing the assisting node through the anomaly detection model deployed on the assisting node.
5. A method for detecting enterprise anomalies, characterized in that, The method, applied to assisting nodes in a distributed system, includes: Received a request to obtain the target company's features; Based on the target enterprise's feature acquisition request, the local access features of the target enterprise are determined and fed back to the central node of the distributed system; the distributed system includes: a central node and multiple assisting nodes; wherein, the local access features are obtained by feature extraction from the access data of the target enterprise accessing the assisting nodes.
6. The method according to claim 5, characterized in that, Also includes: Acquire sample data, which includes: sample access data and the standard enterprise status corresponding to the sample access data; The sample data is input into the anomaly detection model, and features are extracted from the sample access data to obtain node access features; Anomaly detection is performed on the node access characteristics to obtain the state detection result; Based on the difference between the state detection result and the standard enterprise state, the model parameters of the anomaly detection model are adjusted to obtain local model parameters; The local model parameters are sent to the central node so that the central node generates global model parameters based on the local model parameters fed back by each of the assisting nodes. The system receives global model parameters sent by the central node and updates the model parameters of the anomaly detection model based on the global model parameters.
7. An enterprise anomaly detection device, characterized in that, The device, used in the central node of a distributed system, includes: The request sending module is used to send the target enterprise's feature acquisition request to multiple assisting nodes with different business types; the distributed system includes: a central node and multiple assisting nodes; The feature fusion module is used to receive local access features fed back by each of the assisting nodes and fuse the local access features to obtain global access features; wherein, the local access features are obtained by feature extraction of the access data of the target enterprise accessing the assisting nodes; An anomaly detection module is used to determine whether the target enterprise has any anomalies based on the global access characteristics.
8. An enterprise anomaly detection device, characterized in that, The device, used in assisting nodes of a distributed system, includes: The request receiving module is used to receive feature acquisition requests from the target enterprise. The feature extraction module is used to determine the local access features of the target enterprise based on the feature acquisition request of the target enterprise, and feed them back to the central node of the distributed system; the distributed system includes: a central node and multiple assisting nodes; wherein, the local access features are obtained by feature extraction from the access data of the target enterprise accessing the assisting nodes.
9. An enterprise anomaly detection device, characterized in that, The enterprise anomaly detection equipment includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the enterprise anomaly detection method according to any one of claims 1-6.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the enterprise anomaly detection method according to any one of claims 1-6.
