Multi-dimensional index anomaly detection method, device, equipment, medium and product

Abstract

The invention discloses a multi-dimensional index anomaly detection method and device, equipment, a medium and a product, and relates to the technical field of intelligent operation and maintenance, and the method comprises the steps: determining a plurality of key elements associated with an index anomaly root cause based on the index anomaly root cause in a target scene; performing full permutation and combination on the plurality of key elements to obtain a multi-dimensional combination list; the multi-dimensional combination list is associated with a pre-constructed data model; in the process of traversing the multi-dimensional combination list, screening a pre-constructed data model associated with the multi-dimensional combination list based on the exception rule to obtain a target multi-dimensional combination meeting the exception rule; and obtaining a dimension value corresponding to the target multi-dimension combination from a pre-constructed data model, and generating a dimension value combination sequence and a corresponding index abnormal value based on the dimension value. According to the method, multiple dimension factors can be comprehensively considered, and the root cause of the abnormal index can be accurately positioned.

Description

Technical Field

[0001] The present application relates to the technical field of intelligent operation and maintenance, and particularly to a multi-dimensional index anomaly detection method, device, equipment, medium and product. Background Art

[0002] Currently, for data analysis and fault diagnosis in intelligent operation and maintenance, relatively single analysis and diagnosis methods are usually adopted; for example, only whether the index of a certain device exceeds a preset fixed threshold within a specific time period is concerned. If the index exceeds the preset fixed threshold, it is determined as an abnormal index. This method is prone to ignoring some potential risk indicators, unable to comprehensively consider multiple dimensions of factors and their mutual relationships, and thus unable to accurately locate the root causes of all abnormal indicators. Therefore, there is an urgent need for a method that can comprehensively consider multiple dimensions of factors and locate the root causes of abnormal indicators. Summary of the Invention

[0003] In view of the above-mentioned defects or deficiencies in the related art, the purpose of the present application is to provide a multi-dimensional index anomaly detection method, device, equipment, medium and product, which can comprehensively consider multiple dimensions of factors and accurately locate the root causes of abnormal indicators.

[0004] To achieve the above purpose, the present application provides the following solutions: In a first aspect, the present application provides a multi-dimensional index anomaly detection method, including: determining multiple key elements associated with the root cause of index anomalies based on the root cause of index anomalies in a target scenario; performing full permutation and combination on the multiple key elements to obtain a multi-dimensional combination list; the multi-dimensional combination list includes multiple dimension combinations; each dimension combination includes the multiple key elements; the multi-dimensional combination list is associated with a pre-constructed data model; during the process of traversing the multi-dimensional combination list, screening the pre-constructed data model associated with the multi-dimensional combination list based on an anomaly rule to obtain a target multi-dimensional combination that meets the anomaly rule; obtaining dimension values corresponding to the target multi-dimensional combination from the pre-constructed data model, and generating a dimension value combination sequence and a corresponding index anomaly value based on the dimension values.

[0005] Optionally, after performing a full permutation and combination of the multiple key elements to obtain a multi-dimensional combination list, the method further includes: screening the multi-dimensional combination list based on a preset screening rule to obtain a multi-dimensional combination list to be analyzed; during the process of traversing the multi-dimensional combination list, screening the pre-constructed data model associated with the multi-dimensional combination list based on an exception rule to obtain a target multi-dimensional combination that meets the exception rule, including: during the process of traversing the multi-dimensional combination list to be analyzed, screening the pre-constructed data model associated with the multi-dimensional combination list based on an exception rule to obtain the target multi-dimensional combination that meets the exception rule.

[0006] Optionally, during the process of traversing the multi-dimensional combination list to be analyzed, screening the pre-constructed data model associated with the multi-dimensional combination list based on an exception rule to obtain the target multi-dimensional combination that meets the exception rule, including: traversing the multi-dimensional combination list to be analyzed to obtain the number of dimension combinations included in the multi-dimensional combination list to be analyzed; analyzing and counting each multi-dimensional combination in the multi-dimensional combination list to be analyzed in sequence according to the order of the multi-dimensional combination list to be analyzed; during the process of analyzing the multi-dimensional combination list to be analyzed, determining whether the current multi-dimensional combination in the multi-dimensional combination list to be analyzed meets the exception rule; if it meets the exception rule, determining the current multi-dimensional combination as the target multi-dimensional combination, and saving the target multi-dimensional combination and the dimension values corresponding to the target multi-dimensional combination; if it does not meet the exception rule, skipping the current multi-dimensional combination and analyzing the next multi-dimensional combination in the multi-dimensional combination list to be analyzed until the count value is equal to the number of dimension combinations, ending the traversal.

[0007] Optionally, the step of if it does not meet the exception rule, skipping the current multi-dimensional combination and analyzing the next multi-dimensional combination in the multi-dimensional combination list to be analyzed includes: if it does not meet the exception rule, skipping the current multi-dimensional combination and determining whether the traversal of the multi-dimensional combination list to be analyzed is completed; if the traversal is not completed, analyzing the next multi-dimensional combination in the multi-dimensional combination list to be analyzed; if the traversal is completed, ending the analysis.

[0008] Optionally, the method further includes: if no multi-dimensional combination that meets the exception rule is screened out after traversing the multi-dimensional combination list, returning an empty set or the previous historical dimension value combination sequence and the corresponding index anomaly value.

[0009] Optionally, the process of constructing the pre-built data model includes: obtaining all data associated with the root cause of metric anomalies in the target scenario to generate a fact table; using multiple key elements in the target scenario as dimensions to generate a dimension table; and associating the fact table with the dimension table based on a preset algorithm to generate a star or constellation data model.

[0010] In a second aspect, the present application provides a multi-dimensional metric anomaly detection device, including: A determination module, configured to determine multiple key elements associated with the root cause of metric anomalies based on the root cause of metric anomalies in the target scenario; A combination module, configured to perform a full permutation and combination of the multiple key elements to obtain a multi-dimensional combination list; the multi-dimensional combination list includes multiple dimension combinations; each dimension combination contains the multiple key elements; the multi-dimensional combination list is associated with a pre-built data model; An analysis module, configured to, during the process of traversing the multi-dimensional combination list, screen the pre-built data model associated with the multi-dimensional combination list based on anomaly rules to obtain a target multi-dimensional combination that meets the anomaly rules; A generation module, configured to obtain dimension values corresponding to the target multi-dimensional combination from the pre-built data model, and generate a dimension value combination sequence and corresponding metric anomaly values based on the dimension values.

[0011] In a third aspect, the present application provides a computer device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the computer program to implement the steps of the multi-dimensional metric anomaly detection method described in any one of the above.

[0012] In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the multi-dimensional metric anomaly detection method described in any one of the above are implemented.

[0013] In a fifth aspect, the present application provides a computer program product, including a computer program, and when the computer program is executed by a processor, the steps of the multi-dimensional metric anomaly detection method described in any one of the above are implemented.

[0014] According to the specific embodiments provided by the present application, the following technical effects are disclosed in the present application: The present application provides a multi - dimensional index anomaly detection method, apparatus, device, medium and product. It determines multiple key elements associated with the root cause of index anomalies in a target scenario; obtains a multi - dimensional combination list by performing full permutations and combinations on the multiple key elements; during the process of traversing the multi - dimensional combination list, filters a pre - constructed data model associated with the multi - dimensional combination list through anomaly rules to obtain a target multi - dimensional combination that meets the anomaly rules; obtains dimension values corresponding to the target multi - dimensional combination through the pre - constructed data model, and generates a dimension value combination sequence and corresponding index anomaly values based on the dimension values. On the one hand, by constructing a multi - dimensional combination list, it can comprehensively cover the dimensions associated with the root cause of index anomalies, and then comprehensively filter the pre - constructed data model to obtain a dimension value combination sequence and index anomaly values, achieving the purpose of analyzing the root cause of index anomalies from multiple dimensions and effectively improving the accuracy of the analysis results. On the other hand, by traversing the multi - dimensional combination list through the pre - constructed analysis model and anomaly rules, it can effectively improve the analysis efficiency and the accuracy of the analysis results. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] To more clearly illustrate the technical solutions in the embodiments of the present application or in the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0016] Figure 1 It is an application environment diagram of a multi - dimensional index anomaly detection method in an embodiment of the present application; Figure 2 It is a flowchart of a multi - dimensional index anomaly detection method provided in an embodiment of the present application; Figure 3 It is a flowchart of a multi - dimensional index anomaly detection method provided in another embodiment of the present application; Figure 4 It is a schematic diagram of the functional modules of a multi - dimensional index anomaly detection apparatus provided in an embodiment of the present application; Figure 5 It is a schematic diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0018] To make the above objects, features, and advantages of the present application more obvious and understandable, the technical solutions claimed in the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] The multi-dimensional index anomaly detection method provided in the embodiments of the present application can be applied to an application environment as Figure 1 shown. Among them, the terminal 102 communicates with the server 104 through a network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, placed in the cloud or on other servers. The terminal 102 can send the root cause of the index anomaly in the target scenario to the server 104. After receiving the root cause of the index anomaly in the target scenario, for the root cause of the index anomaly in the target scenario, the server 104 determines multiple key elements associated with the root cause of the index anomaly based on the root cause of the index anomaly in the target scenario; performs a full permutation and combination of the multiple key elements to obtain a multi-dimensional combination list; the multi-dimensional combination list includes multiple dimension combinations; each dimension combination contains multiple key elements; the multi-dimensional combination list is associated with a pre-constructed data model; during the process of traversing the multi-dimensional combination list, filters the pre-constructed data model associated with the multi-dimensional combination list based on the anomaly rule to obtain a target multi-dimensional combination that meets the anomaly rule; obtains the dimension values corresponding to the target multi-dimensional combination from the pre-constructed data model, and generates a dimension value combination sequence and the corresponding index anomaly value based on the dimension values. In addition, in some embodiments, the multi-dimensional index anomaly detection method can also be implemented separately by the server 104 or the terminal 102. For example, the terminal 102 can directly perform multi-dimensional index anomaly detection on the root cause of the index anomaly in the target scenario, or the server 104 can obtain the root cause of the index anomaly in the target scenario from the data storage system and perform multi-dimensional index anomaly detection on the root cause of the index anomaly in the target scenario.

[0020] Among them, the terminal 102 can be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, Internet of Things devices, and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. The portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers, and can also be a cloud server.

[0021] In an exemplary embodiment, as Figure 2 shown, a multi-dimensional metric anomaly detection method is provided. This method is executed by a computer device, and specifically, it can be executed alone by a computer device such as a terminal or a server, or jointly executed by a terminal and a server. In the embodiments of the present application, taking this method applied to Figure 1 the server 104 in as an example, the method includes the following steps S201 to S204. Among them:

[0022] In the example embodiment, the user inputs the root cause of metric anomaly in the target scenario through the terminal, and the server obtains the target metric in the target scenario and selects multiple key elements in multiple dimensions according to the target metric. Among them, the root cause of metric anomaly in the target scenario may include: abnormal public transportation flow, root cause analysis of equipment failures, production process optimization, abnormal user growth, abnormal inventory turnover rate, fraud transaction detection, credit risk detection, abnormal energy consumption detection, medical metric anomaly monitoring, adverse drug reaction analysis, etc.

[0023] Key elements. For example, for abnormal public transportation flow, the key elements may include: date (weekday / weekend), time period (morning and evening rush hours), line type (urban line / suburban line), event (concert / holiday), transportation type (subway / vehicle); for root cause analysis of equipment failures, the key elements may include: time (quarter / month / time period), equipment type (compressor / elevator / generator), supplier, maintenance cycle, operating parameters (temperature / pressure); for production process optimization, the key elements may include: shift, equipment model, raw material batch, process parameters, operator; for abnormal user growth, the key elements may include: time (week / promotion period), user source (app store / social media), user profile (age / geography / behavior tags), version number; for abnormal inventory turnover, the key elements may include: warehouse area, commodity category, supplier, logistics channel, order type (pre-sale / spot); for fraud transaction detection, the key elements may include: transaction time (early morning / non-working hours), transaction location (abroad / non-resident city), transaction amount, card type, merchant category; for credit risk detection, the key elements may include: customer type (enterprise / individual), loan term, repayment method, industry (retail / manufacturing), regional economic indicators; for abnormal energy consumption detection, the key elements may include: time period (weekday / holiday), area (production workshop / floor), equipment category (air conditioner / lighting / production line), weather data (temperature / humidity); for abnormal medical indicator monitoring, the key elements may include: department (respiratory department / operating room), medical staff, patient type (inpatient / outpatient), equipment model, operation process node; for analysis of adverse drug reactions, the key elements may include: patient age, gender, underlying diseases, drug dosage, medication duration.

[0024] Step S202: Perform a full permutation and combination of multiple key elements to obtain a multi-dimensional combination list.

[0025] In the exemplary embodiment, the multi-dimensional combination list includes multiple dimensional combinations; each dimensional combination contains multiple key elements; the multi-dimensional combination list is associated with a pre-constructed data model. For example, for abnormal public transportation flow detection, the key elements may be date (weekday / weekend), time period (morning and evening rush hours), line type (urban line / suburban line), event (concert / holiday), and transportation type (subway / vehicle). Performing a full permutation and combination of date (weekday / weekend), time period (morning and evening rush hours), line type (urban line / suburban line), event (concert / holiday), and transportation type (subway / vehicle), the obtained multi-dimensional combination list is a set of 5*4*3*2*1 = 120 dimensional combinations.

[0026] The pre - constructed data model includes a star data model or a constellation data model. Among them, the construction process of the pre - constructed data model includes: obtaining all the data related to the root cause of the metric anomaly in the target scenario and generating a fact table; using multiple key elements in the target scenario as dimensions to generate dimension tables; and associating the fact table with the dimension tables based on a preset algorithm to generate a star data model or a constellation data model.

[0027] Specifically, the OLAP algorithm is used to associate the fact table with the dimension tables to generate a star data model or a constellation data model. For example, a star model or a constellation model is used to construct a data structure, associating the fault records (fact table) with dimension tables such as date, line, and equipment, providing structured data for multi - dimensional analysis. Through the drill - down operation of OLAP, from high - level dimensions (such as "year") to low - level dimensions (such as "month → line → equipment") layer by layer, aggregation results (such as the mean and sum of metrics for each dimension combination) are generated.

[0028] Step S203, during the process of traversing the multi - dimensional combination list, filter the pre - constructed data model associated with the multi - dimensional combination list based on the anomaly rule to obtain the target multi - dimensional combination that meets the anomaly rule.

[0029] In the exemplary embodiment, the anomaly rule refers to a rule formulated through a preset threshold, comparison relationship, or change trend under specific dimension conditions. For example, to analyze whether the "monthly fault count" of a certain subway line is abnormal, dimension: date (month); metric: fault count; the anomaly rule is: the fault count is greater than a preset absolute value threshold. Example: If the anomaly rule is: the fault count is greater than 30, and the fault count in May 2025 is 35 times (fault count 35 times > 30 times), which meets the anomaly rule, then it is determined to be abnormal. Or, the anomaly rule is: the fault count is greater than the historical mean. Example: If the anomaly rule is: the fault count in a certain month > 1.5 times the mean of the past 12 months, then it is determined to be abnormal.

[0030] Step S204, obtain the dimension values corresponding to the target multi - dimensional combination from the pre - constructed data model, and generate a dimension value combination sequence and the corresponding metric anomaly values based on the dimension values.

[0031] It is understandable that a dimension value combination sequence refers to a combination formed by arranging dimension values of multiple dimensions in a specific order, which is used to uniquely identify a specific analysis scenario. For example, for subway traffic fault detection, the analysis dimensions are date, line, and equipment type, and the dimension values of each dimension are as follows: Date: May 2025, June 2025; Line: Subway Line 1, Subway Line 2; Equipment type: elevator, signal machine. Then the possible dimension value combination sequences include: Single dimension: May 2025, Subway Line 1, elevator; Two dimensions: May 2025 → Subway Line 1, Subway Line 1 → elevator, June 2025 → signal machine; Three dimensions: May 2025 → Subway Line 1 → elevator.

[0032] An index outlier refers to a specific value where the actual value of an index deviates from the normal range under a specific dimension value combination sequence, which is used to quantify the degree of abnormality. For example, if the number of faults under the combination of "May 2025 → Subway Line 1" is 50 times, and the average value in the same period of the past 12 months is 20 times (the abnormality rule is: the number of faults is greater than 1.5 times the average value or the number of faults > 30 times), then the number of faults 50 is an outlier. Another example, if the failure rate under the combination of "Subway Line 1 → elevator" is 5%, and the average failure rate of elevators on all lines is 1% (the abnormality rule is: the number of faults is greater than 2 times the average failure rate or the failure rate > 2%), then the failure rate 5% is an outlier.

[0033] Implement the above steps S201 to S204 to determine multiple key elements associated with the root cause of index abnormality through the root cause determination of index abnormality in the target scenario; obtain a multi-dimensional combination list by performing a full permutation and combination of multiple key elements; during the process of traversing the multi-dimensional combination list, screen the pre-constructed data model associated with the multi-dimensional combination list through the abnormality rule to obtain the target multi-dimensional combination that meets the abnormality rule; obtain the dimension values corresponding to the target multi-dimensional combination through the pre-constructed data model, and generate a dimension value combination sequence and the corresponding index outlier based on the dimension values; on the one hand, by constructing a multi-dimensional combination list, it can more comprehensively cover the dimensions associated with the root cause of index abnormality, and then can comprehensively screen the pre-constructed data model to obtain the dimension value combination sequence and index outlier, achieving the purpose of analyzing the root cause of index abnormality from multiple dimensions, and effectively improving the accuracy of the analysis result; on the other hand, by traversing the multi-dimensional combination list through the pre-constructed analysis model and abnormality rule, the analysis efficiency and the accuracy of the analysis result can be effectively improved.

[0034] In another embodiment of the present application, in order to improve the analysis efficiency and the accuracy of the analysis result, before step S203, the method further includes: screening the multi-dimensional combination list based on a preset screening rule to obtain a multi-dimensional combination list to be analyzed.

[0035] In an exemplary embodiment, the preset screening rule is a conventional arrangement corresponding to a dimension in a target scenario. For example, in a target scenario, the root cause of an abnormal indicator is the number of failures that occurred in the subway in 2025. Taking the date as the dimension, the preset screening rule is year / month / day. Or, taking the location where the failure occurred as the dimension, the preset screening rule is line - station.

[0036] Then, step S203 may include: during the process of traversing the list of multi - dimensional combinations to be analyzed, screening the pre - constructed data model associated with the list of multi - dimensional combinations based on the abnormal rule to obtain the target multi - dimensional combination that meets the abnormal rule.

[0037] In a specific embodiment, during the process of traversing the list of multi - dimensional combinations to be analyzed, screening the pre - constructed data model associated with the list of multi - dimensional combinations based on the abnormal rule to obtain the target multi - dimensional combination that meets the abnormal rule may include the following steps S301 to S305. Specifically: Step S301: Traverse the list of multi - dimensional combinations to be analyzed to obtain the number of dimension combinations included in the list of multi - dimensional combinations to be analyzed. Step S302: Analyze and count the list of multi - dimensional combinations to be analyzed one by one in the order of the list of multi - dimensional combinations to be analyzed. Step S303: During the process of analyzing the list of multi - dimensional combinations to be analyzed, determine whether the current multi - dimensional combination in the list of multi - dimensional combinations to be analyzed meets the abnormal rule. Step S304: If it meets the abnormal rule, determine the current multi - dimensional combination as the target multi - dimensional combination, and save the target multi - dimensional combination and the dimension values corresponding to the target multi - dimensional combination. Step S305: If it does not meet the abnormal rule, skip the current multi - dimensional combination and analyze the next multi - dimensional combination in the list of multi - dimensional combinations to be analyzed until the count value is equal to the number of dimension combinations, and end the traversal.

[0038] Further, step S305 may include the following steps S3051 to S3053. Specifically: Step S3051: If it does not meet the abnormal rule, skip the current multi - dimensional combination and determine whether the traversal of the list of multi - dimensional combinations to be analyzed is completed. Step S3052: If the traversal is not completed, analyze the next multi - dimensional combination in the list of multi - dimensional combinations to be analyzed. Step S3053: If the traversal is completed, end the analysis.

[0039] Combined with Figure 3As can be understood from the above embodiments, after the server obtains the root cause of the abnormal index in the target scenario, it executes step S1 to perform a full permutation and combination of multiple key elements to obtain a multi-dimensional combination list; then it executes step S2 to determine whether each multi-dimensional combination in the multi-dimensional combination list meets the preset screening rules; if not, it executes step S3 to screen the multi-dimensional combination list, eliminate the multi-dimensional combinations that do not meet the preset screening rules, and arrange the conforming multi-dimensional combinations in the original order to obtain the multi-dimensional combination list to be analyzed; if it meets, it executes step S4 to arrange the conforming multi-dimensional combinations in the original order to obtain the multi-dimensional combination list to be analyzed; then it executes step S5 to analyze and count one by one according to the order of the multi-dimensional combination list to be analyzed; during the analysis process, it executes step S6 to determine whether the current multi-dimensional combination in the multi-dimensional combination list to be analyzed meets the abnormal rule; if it meets, it executes step S7 to determine that the current multi-dimensional combination is the target multi-dimensional combination, and save the target multi-dimensional combination and the dimension values corresponding to the target multi-dimensional combination, and then execute step S8; if it does not meet, it directly skips the current multi-dimensional combination and executes step S8 to determine whether the count value is equal to the number of dimension combinations; if they are equal, the traversal is completed, and step S9 is executed to obtain the dimension values corresponding to all the target multi-dimensional combinations, generate a dimension value combination sequence and the corresponding index abnormal value; if they are not equal, the traversal is not completed, and step S5 is executed to analyze the next multi-dimensional combination in the multi-dimensional combination list to be analyzed, and then steps S6, S7, S8 and / or step S9 are executed in sequence until the count value is equal to the number of dimension combinations, the multi-dimensional combination list to be analyzed is traversed, and the analysis is ended and the dimension value combination sequence and the corresponding index abnormal value are output.

[0040] It should be noted that if all the multi-dimensional combinations in the multi-dimensional combination list to be analyzed do not meet the abnormal rule, an empty set or the previous historical dimension value combination sequence and the corresponding index abnormal value are returned.

[0041] By returning an empty set or the previous historical dimension value combination sequence and the corresponding index abnormal value, it is a supplement to a special case of the analysis result, which can further improve the accuracy of the analysis result.

[0042] Based on the same inventive concept, an embodiment of the present application also provides a multi-dimensional index abnormal detection device for implementing the above-mentioned multi-dimensional index abnormal detection method. The implementation solution provided by this device to solve the problem is similar to the implementation solution described in the above method. Therefore, the specific limitations in one or more embodiments of the multi-dimensional index abnormal detection device provided below can refer to the limitations on the multi-dimensional index abnormal detection method in the above text, and will not be repeated here.

[0043] In an exemplary embodiment, such as Figure 4As shown in the figure, a multi-dimensional index anomaly detection device is provided. The multi-dimensional index anomaly detection device 400 includes: a determination module 401, a combination module 402, an analysis module 403, and a generation module 404, where: The determination module 401 is configured to determine a plurality of key elements associated with the root cause of the index anomaly based on the root cause of the index anomaly in the target scenario; The combination module 402 is configured to perform a full permutation and combination on the plurality of key elements to obtain a multi-dimensional combination list; the multi-dimensional combination list includes a plurality of dimension combinations; each dimension combination includes a plurality of key elements; the multi-dimensional combination list is associated with a pre-constructed data model; The analysis module 403 is configured to, during the process of traversing the multi-dimensional combination list, screen the pre-constructed data model associated with the multi-dimensional combination list based on the anomaly rule to obtain a target multi-dimensional combination that meets the anomaly rule; The generation module 404 is configured to obtain the dimension values corresponding to the target multi-dimensional combination from the pre-constructed data model, and generate a dimension value combination sequence and the corresponding index anomaly value based on the dimension values.

[0044] As an optional implementation manner, the above multi-dimensional index anomaly detection device 400 further includes a screening module. The screening module is configured to screen the multi-dimensional combination list based on a preset screening rule to obtain a to-be-analyzed multi-dimensional combination list; specifically, the analysis module 403 is configured to, during the process of traversing the to-be-analyzed multi-dimensional combination list, screen the pre-constructed data model associated with the multi-dimensional combination list based on the anomaly rule to obtain a target multi-dimensional combination that meets the anomaly rule.

[0045] As an optional implementation manner, the analysis module 403 is further specifically configured to traverse the to-be-analyzed multi-dimensional combination list to obtain the number of dimension combinations included in the to-be-analyzed multi-dimensional combination list; analyze and count the to-be-analyzed multi-dimensional combination list one by one in the order of the to-be-analyzed multi-dimensional combination list; during the process of analyzing the to-be-analyzed multi-dimensional combination list, determine whether the current multi-dimensional combination in the to-be-analyzed multi-dimensional combination list meets the anomaly rule; if it meets the anomaly rule, determine the current multi-dimensional combination as the target multi-dimensional combination, and save the target multi-dimensional combination and the dimension values corresponding to the target multi-dimensional combination; if it does not meet the anomaly rule, skip the current multi-dimensional combination and analyze the next multi-dimensional combination in the to-be-analyzed multi-dimensional combination list until the to-be-analyzed multi-dimensional combination list is traversed.

[0046] As an optional implementation manner, the analysis module 403 is further configured to, if it does not meet the anomaly rule, skip the current multi-dimensional combination and determine whether the traversal of the to-be-analyzed multi-dimensional combination list is completed; if the traversal is not completed, analyze the next multi-dimensional combination in the to-be-analyzed multi-dimensional combination list; if the traversal is completed, the analysis ends.

[0047] As an alternative implementation, the above-mentioned multi-dimensional index anomaly detection device 400 further includes a second generation module. The second generation module is configured to, if no multi-dimensional combination that meets the anomaly rule is selected after traversing the multi-dimensional combination list, return an empty set or the previous historical dimension value combination sequence and the corresponding index anomaly value.

[0048] As an alternative implementation, the above-mentioned multi-dimensional index anomaly detection device 400 further includes a construction module. The construction module is configured to: obtain all data associated with the root cause of index anomaly in the target scenario and generate a fact table; use multiple key elements in the target scenario as dimensions to generate a dimension table; and associate the fact table with the dimension table based on a preset algorithm to generate a star-shaped or constellation data model.

[0049] Among them, when implementing this implementation, multiple key elements associated with the root cause of index anomaly are determined through the root cause of index anomaly in the target scenario; a multi-dimensional combination list is obtained by performing a full permutation and combination of the multiple key elements; during the process of traversing the multi-dimensional combination list, the pre-constructed data model associated with the multi-dimensional combination list is screened through the anomaly rule to obtain the target multi-dimensional combination that meets the anomaly rule; the dimension values corresponding to the target multi-dimensional combination are obtained through the pre-constructed data model, and a dimension value combination sequence and the corresponding index anomaly value are generated based on the dimension values; on the one hand, by constructing the multi-dimensional combination list, the dimensions associated with the root cause of index anomaly can be more comprehensively covered, and then the pre-constructed data model can be comprehensively screened to obtain the dimension value combination sequence and the index anomaly value, achieving the purpose of analyzing the root cause of index anomaly from multiple dimensions and effectively improving the accuracy of the analysis result; on the other hand, by traversing the multi-dimensional combination list through the pre-constructed analysis model and the anomaly rule, the analysis efficiency and the accuracy of the analysis result can be effectively improved.

[0050] In an exemplary embodiment, a computer device is provided. The computer device can be a server or a terminal, and its internal structure diagram can be as Figure 5As shown in the figure. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store multi-dimensional metric anomaly detection data. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals through a network connection. When the computer program is executed by the processor, it implements a multi-dimensional metric anomaly detection method.

[0051] Those skilled in the art can understand that Figure 5 the structure shown in the figure is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different component layout.

[0052] In an exemplary embodiment, a computer device is further provided, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, the steps in the above method embodiments are implemented.

[0053] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program. When the computer program is executed by the processor, the steps in the above method embodiments are implemented.

[0054] In an exemplary embodiment, a computer program product is provided, including a computer program. When the computer program is executed by the processor, the steps in the above method embodiments are implemented.

[0055] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data that have been consented to by the user or fully consented to by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0056] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memories can include read-only memory (ROM), magnetic tapes, floppy disks, flash memories, optical memories, high-density embedded non-volatile memories, resistive random access memories (ReRAM), magnetoresistive random access memories (MRAM), ferroelectric random access memories (FRAM), phase change memories (PCM), graphene memories, etc. Volatile memories can include random access memory (RAM) or external cache memories, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0057] The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, etc., without limitation.

[0058] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0059] Specific examples are used in this article to elaborate on the principles and implementation manners of the present application. The descriptions of the above embodiments are only used to help understand the methods and core ideas of the present application; at the same time, for those of ordinary skill in the art, according to the ideas of the present application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. A multi-dimensional index anomaly detection method, characterized in that The multi-dimensional index anomaly detection method includes: Determine multiple key elements associated with the root cause of the index anomaly based on the root cause of the index anomaly in the target scenario; Perform a full permutation and combination of the multiple key elements to obtain a multi-dimensional combination list; the multi-dimensional combination list includes multiple dimension combinations; each dimension combination contains the multiple key elements; the multi-dimensional combination list is associated with a pre-constructed data model; During the process of traversing the multi-dimensional combination list, filter the pre-constructed data model associated with the multi-dimensional combination list based on the anomaly rule to obtain a target multi-dimensional combination that meets the anomaly rule; Obtain the dimension values corresponding to the target multi-dimensional combination from the pre-constructed data model, and generate a dimension value combination sequence and the corresponding index anomaly value based on the dimension values.

2. The multi-dimensional index anomaly detection method according to claim 1, wherein, After performing the full permutation and combination of the multiple key elements to obtain the multi-dimensional combination list, the method further includes: Filter the multi-dimensional combination list based on a preset filtering rule to obtain a multi-dimensional combination list to be analyzed; The step of, during the process of traversing the multi-dimensional combination list, filtering the pre-constructed data model associated with the multi-dimensional combination list based on the anomaly rule to obtain a target multi-dimensional combination that meets the anomaly rule includes: During the process of traversing the multi-dimensional combination list to be analyzed, filter the pre-constructed data model associated with the multi-dimensional combination list based on the anomaly rule to obtain the target multi-dimensional combination that meets the anomaly rule.

3. The multi-dimensional index anomaly detection method according to claim 2, wherein The step of, during the process of traversing the multi-dimensional combination list to be analyzed, filtering the pre-constructed data model associated with the multi-dimensional combination list based on the anomaly rule to obtain the target multi-dimensional combination that meets the anomaly rule includes: Traverse the multi-dimensional combination list to be analyzed to obtain the number of dimension combinations included in the multi-dimensional combination list to be analyzed; Analyze and count the multi-dimensional combination list to be analyzed one by one in the order of the multi-dimensional combination list to be analyzed; During the process of analyzing the multi-dimensional combination list to be analyzed, determine whether the current multi-dimensional combination of the multi-dimensional combination list to be analyzed meets the anomaly rule; If it meets the anomaly rule, determine the current multi-dimensional combination as the target multi-dimensional combination, and save the target multi-dimensional combination and the dimension values corresponding to the target multi-dimensional combination; If it does not meet the anomaly rule, skip the current multi-dimensional combination and analyze the next multi-dimensional combination of the multi-dimensional combination list to be analyzed until the count value is equal to the number of dimension combinations, and end the traversal.

4. The multi-dimensional index anomaly detection method according to claim 3, characterized in that, The step of, if it does not meet the anomaly rule, skipping the current multi-dimensional combination and analyzing the next multi-dimensional combination of the multi-dimensional combination list to be analyzed includes: If it does not meet the anomaly rule, skip the current multi-dimensional combination and determine whether the traversal of the multi-dimensional combination list to be analyzed is completed; If the traversal is not completed, analyze the next multi-dimensional combination of the multi-dimensional combination list to be analyzed; If the traversal is completed, the analysis ends.

5. The multi-dimensional index anomaly detection method according to claim 1, wherein The method further includes: If no multi-dimensional combination that meets the anomaly rule is selected after traversing the multi-dimensional combination list, an empty set or the previous historical dimension value combination sequence and the corresponding indicator anomaly value are returned.

6. The multi-dimensional index anomaly detection method according to claim 1, wherein The construction process of the pre-constructed data model includes: Obtain all data associated with the root cause of indicator anomalies in the target scenario and generate a fact table; Use multiple key elements in the target scenario as dimensions to generate a dimension table; Based on a preset algorithm, associate the fact table with the dimension table to generate a star-shaped or constellation data model.

7. A multi-dimensional index anomaly detection device, characterized in that, The multi-dimensional indicator anomaly detection device includes: A determination module for determining multiple key elements associated with the root cause of indicator anomalies based on the root cause of indicator anomalies in the target scenario; A combination module for performing a full permutation and combination of the multiple key elements to obtain a multi-dimensional combination list; the multi-dimensional combination list includes multiple dimension combinations; each dimension combination contains the multiple key elements; the multi-dimensional combination list is associated with a pre-constructed data model; An analysis module for screening the pre-constructed data model associated with the multi-dimensional combination list based on the anomaly rule during the traversal of the multi-dimensional combination list to obtain a target multi-dimensional combination that meets the anomaly rule; A generation module for obtaining the dimension values corresponding to the target multi-dimensional combination from the pre-constructed data model and generating a dimension value combination sequence and the corresponding indicator anomaly value based on the dimension values.

8. A computer device, comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the multi-dimensional indicator anomaly detection method according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the multi-dimensional indicator anomaly detection method according to any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the multi-dimensional indicator anomaly detection method according to any one of claims 1-6.

Images