A method, device and apparatus for determining abnormal resource transfer situation
By obtaining the resource transfer information set of the target system and using the decision tree model and growth function model, the importance of resource transfer and monitoring threshold are determined, and the problem of inability to accurately judge abnormal transactions in the existing technology is solved, achieving efficient and accurate Exception resource transfer monitoring.
Patent Information
- Application Number
- CN202110515482.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-05-12
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2041-05-12
AI Technical Summary
The existing technology cannot accurately determine whether there are abnormal transactions, resulting in false positives or missed reports.
By obtaining the target resource transfer information set of the target system in the target time period, the pre-trained decision tree model is used to determine the importance of resource transfer failure, and the monitoring threshold is determined based on the growth function model to determine whether there is an abnormal resource transfer situation.
It realizes efficient and accurate judgment of resource transfer information sets within different time periods, avoids false alarms or missed reports, and improves the accuracy of monitoring.
Smart Images

Figure CN113112043B_ABST
Abstract
Description
Technical Field
[0001] The embodiments of this specification relate to the field of artificial intelligence technology, and in particular, to a method, device and apparatus for determining abnormal resource transfer conditions. Background Art
[0002] In the daily production and maintenance work of production and maintenance organizations, transaction monitoring occupies an important position in operation and maintenance. In the big data scenario, operation and maintenance personnel can respond quickly through transaction monitoring alarms to reduce emergency time and thus reduce business impact. The current alarm mechanism in production is to alarm when the ratio of the number of failed transactions within 5 minutes to the total number of transactions within 5 minutes is greater than or equal to the set fixed threshold, where the above-mentioned fixed threshold is usually set by operation and maintenance personnel based on experience. However, the importance of different applications and different transactions of the same application may vary. Therefore, only using the same fixed threshold set by human experience to make judgments may result in false alarms or omissions. It can be seen that the use of existing technologies cannot accurately determine whether there are abnormal transactions.
[0003] To address the above problems, no effective solution has been proposed yet. Summary of the invention
[0004] The embodiments of this specification provide a method, device and apparatus for determining abnormal resource transfer conditions to solve the problem in the prior art that it is impossible to accurately determine whether there is an abnormal transaction situation.
[0005] An embodiment of the present specification provides a method for determining an abnormal resource transfer situation, including: obtaining a target resource transfer information set of a target system within a target time period; when it is determined that there is a resource transfer failure in the target resource transfer information set, determining a characterization value of the target resource transfer information set based on the target resource transfer information set using a target decision tree model; wherein the target decision tree model is a pre-trained model for determining the importance of resource transfer failures, and the characterization value is used to represent the importance of failed resource transfers in the target resource transfer information set; based on the characterization value, determining a monitoring threshold through a growth function model; wherein the monitoring threshold is used to determine whether there is an abnormal resource transfer situation within the target time period.
[0006] An embodiment of the present specification also provides a device for determining an abnormal resource transfer situation, including: an acquisition module, used to obtain a target resource transfer information set of a target system within a target time period; a processing module, used to determine a characterization value of the target resource transfer information set using a target decision tree model based on the target resource transfer information set when it is determined that there is a resource transfer failure in the target resource transfer information set; wherein the target decision tree model is a pre-trained model for determining the importance of resource transfer failures, and the characterization value is used to represent the importance of failed resource transfers in the target resource transfer information set; a determination module, used to determine a monitoring threshold through a growth function model based on the characterization value; wherein the monitoring threshold is used to determine whether there is an abnormal resource transfer situation within the target time period.
[0007] The embodiment of the present specification also provides a device for determining an abnormal resource transfer situation, including a processor and a memory for storing processor executable instructions, and the processor implements the steps of the method for determining the abnormal resource transfer situation when executing the instructions.
[0008] The embodiments of the present specification also provide a computer-readable storage medium on which computer instructions are stored, and when the instructions are executed, the steps of the method for determining abnormal resource transfer conditions are implemented.
[0009] The embodiment of this specification provides a method for determining abnormal resource transfer situations. The method can obtain a target resource transfer information set of a target system within a target time period. When it is determined that there is a resource transfer failure in the target resource transfer information set, the target decision tree model is used to determine the characterization value of the target resource transfer information set according to the target resource transfer information set, so that the importance of the failed resource transfer in the target resource transfer information set can be efficiently and accurately determined. Among them, the above-mentioned target decision tree model is a model obtained by pre-training for determining the importance of resource transfer failure. Further, the monitoring threshold of the target resource transfer information set within the target time period can be determined based on the characterization value, and then it can be determined whether there is an abnormal resource transfer situation within the target time period based on the monitoring threshold. Therefore, the monitoring threshold can be determined for the resource transfer information sets in different time periods based on the importance of their resource transfer failures, so that the monitoring threshold finally determined can meet the characteristics of resource transfer, effectively avoiding the occurrence of false alarms or omissions. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] The drawings described herein are used to provide a further understanding of the embodiments of this specification, constitute a part of the embodiments of this specification, and do not constitute a limitation of the embodiments of this specification. In the drawings:
[0011] Figure 1It is a schematic diagram of the steps of a method for determining abnormal resource transfer conditions provided in an embodiment of this specification;
[0012] Figure 2 is a schematic diagram of a curve showing changes in the monitoring threshold with the importance of resource transfer failure according to an embodiment of this specification;
[0013] Figure 3 It is a structural schematic diagram of a device for determining abnormal resource transfer conditions provided in an embodiment of this specification;
[0014] Figure 4 It is a structural diagram of a device for determining abnormal resource transfer conditions provided in an embodiment of this specification. DETAILED DESCRIPTION
[0015] The principles and spirit of the embodiments of this specification will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided only to enable those skilled in the art to better understand and implement the embodiments of this specification, and are not intended to limit the scope of the embodiments of this specification in any way. On the contrary, these embodiments are provided to make the disclosure of the embodiments of this specification more thorough and complete, and to fully convey the scope of the disclosure to those skilled in the art.
[0016] Those skilled in the art know that the implementation of the embodiments of this specification can be implemented as a system, device, method or computer program product. Therefore, the embodiments disclosed in this specification can be specifically implemented in the following forms, namely: complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.
[0017] Although the processes described below include multiple operations that occur in a particular order, it should be clear that these processes may include more or fewer operations, which may be performed sequentially or in parallel (eg, using parallel processors or a multi-threaded environment).
[0018] See also Figure 1 , this embodiment can provide a method for determining abnormal resource transfer situations. The method for determining abnormal resource transfer situations can be used to determine the importance of resource transfer failures for resource transfer information within a period of time when determining that there are resource transfer failures within the period of time, and determine the corresponding monitoring threshold based on the determined importance of resource transfer failures, so as to accurately monitor whether there are abnormal transactions within the period of time. The above-mentioned method for determining abnormal resource transfer situations can include the following steps.
[0019] S101: Acquire a target resource transfer information set of a target system within a target time period.
[0020] In this embodiment, the resource transfer information generated in each time interval can be recorded at a preset time interval to obtain a resource transfer information set. For example, a resource transfer information set can be recorded every 5 minutes. Based on this, when it is necessary to analyze whether there is an abnormal resource transfer situation at present, the target resource transfer information set of the target system in the target time period can be obtained, wherein the target time period can be the time interval for obtaining the resource transfer information at present.
[0021] In this embodiment, the target resource transfer information set may include resource transfer information generated within the target time period. Since different types of resource transfer information may have different importance, the types of resource transfers in the target resource transfer information set may be the same, so that it is possible to determine whether there is an abnormality in the current resource transfer of the same type for the same type of resource transfer information, thereby improving the accuracy of monitoring. It can be seen that different resource transfer types within the target time period can be recorded in a resource transfer information set, and corresponding monitoring thresholds can be determined for resource transfer information sets of different resource transfer types.
[0022] In this embodiment, the resource transfer information can be used to characterize the resource transfer situation, and specifically, can include the resource transfer initiator, the resource transfer recipient, the resource transfer amount, the application to which it belongs, the resource transfer level, the result of the resource transfer (whether it is successful), the scope of the resource transfer impact, the time of the resource transfer, etc. In some embodiments, other possible resource transfer information can also be included, such as the level of the application to which it belongs, etc., which can be determined according to the actual situation, and the embodiments of this specification do not limit this.
[0023] In this embodiment, the above resources may include: income and profit, real estate fixed assets, human resources, expense amount, etc. Resource transfer can be used to indicate the transfer of resources between users, between users and third-party users, or between third-party users. In one embodiment, the types of resource transfer may include: transfer, third-party payment, collection of water and electricity, purchase of financial products, etc. The specific types can be determined according to actual conditions, and this specification embodiment does not limit this.
[0024] S102: When it is determined that there is a resource transfer failure in the target resource transfer information set, determine the characterization value of the target resource transfer information set based on the target resource transfer information set using a target decision tree model; wherein the target decision tree model is a pre-trained model for determining the importance of the resource transfer failure, and the characterization value is used to represent the importance of the failed resource transfer in the target resource transfer information set.
[0025] In this embodiment, if there is no request for failed resource transfer in the target resource transfer information set, it means that there is no abnormality in the resource transfer type corresponding to the target resource transfer information set within the target time period. If there is a resource transfer failure in the target resource transfer information set, it means that there may be a resource transfer abnormality, and the characterization value of the target resource transfer information set can be determined based on the target resource transfer information set using the target decision tree model.
[0026] In this embodiment, the above-mentioned characterization value is used to indicate the importance of the failed resource transfer in the target resource transfer information set, and the characterization value is a value greater than 0, which can be determined according to the actual situation, and the embodiments of this specification do not limit this. The larger the above-mentioned characterization value, the more important the failed resource transfer is, for example: 0.000 means "the failure is not important and is sporadic", and 4.000 means "the failure is very important".
[0027] In this embodiment, since the target resource transfer information set records the resource transfer situation within the target time period, the importance of the failed resource transfer in the target resource transfer information set can be determined using a pre-trained target decision tree model based on the target resource transfer information set. The target decision tree model can be a pre-trained machine learning model for determining the importance of failed resource transfers.
[0028] In this embodiment, the decision tree model is essentially a tree composed of multiple judgment nodes, and feature parameter judgment is made at each judgment node of the tree, and then the last branch (leaf node) of the tree can make the best judgment on the value of the variable of interest. Usually, a decision tree contains a root node, several internal nodes and several leaf nodes, and the leaf nodes correspond to the decision classification results.
[0029] In this embodiment, since the target resource transfer information set records the original data generated when the resource is transferred, it may not contain all the characteristic parameters required for the target decision tree model to judge. Therefore, in some embodiments, when it flows to each judgment node of the target decision tree, the resource transfer information in the target resource transfer information set can be statistically analyzed to obtain the corresponding required characteristic parameters. For example: the success rate of resource transfer within the target time period can be obtained based on the results of each resource transfer in the target resource transfer information set. Of course, the statistical analysis method is not limited to the above examples. Technical personnel in the relevant field may make other changes under the inspiration of the technical essence of the embodiments of this specification, but as long as the functions and effects achieved are the same or similar to the embodiments of this specification, they should be covered within the protection scope of the embodiments of this specification.
[0030] S103: Based on the characterization value, determine a monitoring threshold through a growth function model; wherein the monitoring threshold is used to determine whether there is an abnormal resource transfer situation within a target time period.
[0031] In this embodiment, since the above-mentioned characterization value can be used to indicate the importance of the failed resource transfer in the target resource transfer information set, the monitoring threshold can be determined by the growth function model based on the above-mentioned characterization value. The above-mentioned monitoring threshold can be used to determine whether there is an abnormal resource transfer situation within the target time period.
[0032] In this embodiment, the growth function model can be expressed as follows.
[0033]
[0034] Among them, y is the indicator threshold; x is the importance of resource transfer failure (characterization value); k, a, b, c, threshold are constant terms, and the above constant terms can be obtained by fitting historical records.
[0035] In this embodiment, if the ratio of the number of failed resource transfers to the total number of resource transfers within the target time period is greater than or equal to the monitoring threshold, it can be determined that there is an abnormal resource transfer situation, and an alarm can be issued at this time, and the abnormal resource transfer situation can be fed back to the relevant processing personnel. If the ratio of the number of failed resource transfers to the total number of resource transfers within the target time period is less than the above monitoring threshold, it means that there is no abnormal resource transfer situation within the target time period.
[0036] In this embodiment, the monitoring threshold may be a value greater than 0 and less than 1, such as 0.5, 0.6, 0.81, 0.99, etc. The specific value may be determined based on actual conditions, and the embodiments of this specification do not limit this.
[0037] From the above description, it can be seen that the embodiment of this specification realizes the following technical effects: by obtaining the target resource transfer information set of the target system within the target time period, when it is determined that there is a resource transfer failure in the target resource transfer information set, the target decision tree model is used to determine the characterization value of the target resource transfer information set according to the target resource transfer information set, so that the importance of the failed resource transfer in the target resource transfer information set can be efficiently and accurately determined. Among them, the above-mentioned target decision tree model is a model obtained by pre-training for determining the importance of resource transfer failure. Further, the monitoring threshold of the target resource transfer information set within the target time period can be determined based on the characterization value, and then it can be determined whether there is an abnormal resource transfer situation within the target time period based on the monitoring threshold. Therefore, the monitoring threshold can be determined for the resource transfer information sets in different time periods based on the importance of their resource transfer failures, so that the monitoring threshold finally determined can meet the characteristics of resource transfer, effectively avoiding the occurrence of false alarms or omissions.
[0038] In one embodiment, according to the target resource transfer information set, a target decision tree model is used to determine the characterization value of the target resource transfer information set, which can include: obtaining corresponding target feature parameters according to the target resource transfer information set when flowing to each judgment node of the target decision tree model, until the characterization value is obtained when flowing to the leaf node of the target decision tree model; wherein the target decision tree model includes multiple judgment nodes, and each judgment node corresponds to a target feature parameter.
[0039] In this embodiment, the target decision tree model may include multiple judgment nodes and multiple leaf nodes, each judgment node may correspond to a target feature parameter, wherein the leaf node corresponds to the decision classification result. The corresponding target feature parameter judgment is performed at each judgment node of the target decision tree model, and then the last branch (leaf node) of the target decision tree model can make the best judgment on the value of the concerned variable.
[0040] In this embodiment, each leaf node of the target decision tree model can correspond to a specific numerical value of importance or a range of importance values, so that when the flow reaches the leaf node of the target decision tree model, the specific numerical value of importance or the range of importance values corresponding to the leaf node can be used as the obtained representation value.
[0041] In one embodiment, the target characteristic parameters may include: application level, resource transfer level, resource transfer failure rate within the target time period, resource transfer time period, scope affected by resource transfer, number of monitoring false alarms before resource transfer, etc. Of course, it can be understood that in some embodiments, other characteristic parameters may also be included, such as resource transfer frequency, etc., which can be determined according to actual conditions, and the embodiments of this specification do not limit this.
[0042] In one embodiment, before determining the characterization value of the target resource transfer information set using the target decision tree model according to the target resource transfer information set, it can also include: obtaining an initial historical monitoring data set; wherein the historical monitoring data set includes multiple alarm information, and each alarm information includes multiple candidate feature parameters; the initial historical monitoring data set can be preprocessed to obtain the target monitoring data set. Further, based on the target monitoring data set, a first training data set and a first test data set can be generated, and the target decision tree model can be constructed using the first training data set and the first test data set.
[0043] In this embodiment, an initial historical monitoring data set may be obtained as data for training, and the initial historical monitoring data set may include multiple historical alarm information, and the historical alarm information is a historical record of abnormal resource transfer within a preset time interval and determining that an alarm is required. The above initial historical monitoring data set may include 10,000 or more or less alarm information, which may be determined according to actual conditions, and the embodiments of this specification are not limited to this.
[0044] In this embodiment, each alarm message may include multiple candidate feature parameters, and the candidate feature parameters may include: the application level, resource transfer level, resource transfer failure rate within a preset time interval, the scope affected by the resource transfer, resource transfer time period, resource transfer frequency, monitoring frequency, number of monitoring false alarms before resource transfer, estimated time to resolve the alarm, manual steps to resolve the alarm, etc. Among them, the preset time interval is the same as the time interval corresponding to the target time period. Of course, the candidate feature parameters are not limited to the above examples, and technicians in the relevant field may make other changes under the inspiration of the technical essence of the embodiments of this specification, but as long as the functions and effects achieved are the same or similar to the embodiments of this specification, they should be covered within the protection scope of the embodiments of this specification.
[0045] In this embodiment, the description of the above candidate feature parameters can be as shown in Table 1.
[0046] Table 1
[0047]
[0048]
[0049] In this embodiment, since the data in the initial historical monitoring data set may contain abnormal values, missing values, etc., the initial historical monitoring data set can be preprocessed to obtain the target monitoring data set. Among them, in the preprocessing process, the abnormal value of the resource transfer failure rate within the preset time interval and the monitoring false alarm rate before the resource transfer can be processed. The minimum value of the two variables of the resource transfer failure rate within the preset time interval and the monitoring false alarm rate before the resource transfer is 0, which is an abnormal value. Therefore, the 0 value of the two variables of the resource transfer failure rate within the preset time interval and the monitoring false alarm rate before the resource transfer can be processed as a NaN value.
[0050] In this embodiment, the target monitoring data set can be imported and divided into a first training data set and a first test data set using the train_test_split function. Further, the first training data set and the first test data set can be used to establish a decision tree model using sklearn.DecisionTreeClassifier in Python. The above-mentioned Python is a computer programming language that provides efficient high-level data structures and can also perform simple and effective object-oriented programming.
[0051] In one embodiment, using the first training data set and the first test data set to construct a target decision tree model may include: using the first training data set and the first test data set to construct an initial decision tree model. Importance labels may be added to each alarm information in the first training data set to obtain a second training data set. And according to the second training data set and the initial decision tree model, the importance ranking results of each candidate feature parameter are determined. Further, a preset number of target feature parameters may be screened according to the importance ranking results of each candidate feature parameter. And based on the preset number of target feature parameters, a target decision tree model is constructed.
[0052] In this embodiment, the initial decision tree model can be established using the first training data set and the first test data set and the sklearn.DecisionTreeClassifier in Python. When establishing the initial decision tree model, the tree is fully expanded by default (the tree continues to branch until all leaf nodes are pure). When establishing the initial decision tree model, the random_state of the tree can be fixed to solve the tie problem internally. random_state is used to set the parameters of the random pattern in the branch, and the default is None. The random_state of the tree is fixed to ensure that the same training set and test set are split each time the program runs, so that the model stabilizes. Otherwise, the same algorithm model will have different effects on different training sets and test sets.
[0053] In this embodiment, the data can be recursively divided repeatedly until each divided area (each leaf node of the decision tree) contains only a single target value (a single category or a single regression value). If the target values of the data points contained in a leaf node in the tree are the same, then the leaf node is pure. To predict a new data point, first check which area of the feature space the point is located in, and then use the majority of the target values in the area (if it is a pure leaf node, it is a single target value) as the prediction result. Starting from the root node, the tree can be traversed to find this area. Each step to the left or right depends on whether the corresponding judgment conditions are met.
[0054] In this embodiment, the existence of pure leaf nodes indicates that the accuracy of this tree on the training set is 100%. Each data point in the training set is located in a leaf node that is correctly classified. The initial decision tree model is constructed until all leaf nodes are pure leaf nodes, which will cause the initial decision tree model to be very complex and highly overfit to the training data. Therefore, in order to control the complexity of the decision tree and avoid overfitting, the initial decision tree model can be optimized.
[0055] In this embodiment, the growth of the tree can be stopped early by pre-pruning to prevent overfitting; or by post-pruning or pruning, the tree is constructed first, but nodes with little information are deleted or collapsed later to prevent overfitting. The specific method can be determined according to the actual situation, and the embodiments of this specification are not limited to this.
[0056] In this embodiment, the initial decision tree model can be optimized by pre-pruning to improve the generalization degree of new data. The constraints of pre-pruning may include limiting the maximum depth of the tree, limiting the maximum number of leaf nodes, or stipulating the minimum number of data points in a node to prevent further division.
[0057] In this embodiment, since the decision tree model is essentially a binary tree, each judgment node corresponds to a feature hyperparameter. In order to limit the maximum depth of the tree, the candidate feature parameters can be screened first, and unimportant features can be filtered out before model fitting, for example, the features that are least relevant to the results can be filtered out. In one embodiment, importance labels can be added to each alarm information in the first training data set to obtain a second training data set, and the importance ranking results of each candidate feature parameter can be determined using the feature_importances_ method of sklearn.DecisionTreeClassifier based on the second training data set and the initial decision tree model.
[0058] In this embodiment, the feature_importances_method can be used to determine the importance of each candidate feature parameter by learning the second training data set. The feature_importances_method can be measured by the number of times the feature is split or the gain after the feature is split.
[0059] In this embodiment, the candidate feature parameters can be screened by setting max_depth (maximum depth). For example, if max_depth=6 is set, the preset number can be 6, and the feature parameters ranked first 6 in the order of importance of the candidate feature parameters can be used as target feature parameters. Of course, it can be understood that in some embodiments, feature parameters whose importance is greater than a certain value can also be used as target feature parameters. The specific value can be determined according to actual conditions, and this specification embodiment does not limit this. In a scenario example, when max_depth=6 is set, the screening results for the candidate feature parameters in Table 1 can be as shown in Table 2.
[0060] Table 2
[0061]
[0062]
[0063] In this embodiment, a target decision tree model can be constructed based on the target feature parameters obtained by screening, and the depth value of the target decision tree model is the total number of the above target feature parameters. Limiting the depth of the tree can reduce overfitting, which will reduce the accuracy of the training set, but can improve the accuracy of the test set.
[0064] In one embodiment, constructing a target decision tree model based on a preset number of target feature parameters may include: determining multiple classification dimensions of the importance of resource transfer failure based on the preset number, and adding a resource transfer failure importance label to each alarm information in the first training data set and the first test data set according to the multiple classification dimensions of the importance of resource transfer failure, to obtain a third training data set and a second test data set. Further, a target decision tree model may be constructed based on the third training data set, the second test data set, and the preset number of target feature parameters.
[0065] In this embodiment, when determining the max_depth of the tree, the number of leaf nodes of the tree can be determined accordingly. For example, the number of leaf nodes of a binary tree with a max_depth of 6 is 2. 5=32, therefore, the importance of resource transfer failure can be divided into 32 categories. When the importance value is [0,4], it can be divided into 32 classification dimensions of 0.000, 0.125, 0.250, 0.500, ..., 3.875, 4.000. The larger the data, the higher the importance of the failed resource transfer. In the leaf node of the target decision tree model, the importance can be a custom number or a compliance, as long as it can show the importance. The specific importance can be determined according to the actual situation, and this specification embodiment does not limit this.
[0066] In this embodiment, in order to train the decision tree model, the importance of resource transfer failure can be added to each alarm information in the first training data set and the first test data set according to multiple classification dimensions of the importance of resource transfer failure, so as to obtain a third training data set and a second test data set. Thus, the importance of resource transfer failure corresponding to each alarm information belongs to a certain classification dimension.
[0067] In this embodiment, an initial decision tree model can be established based on the third training data set, the second test data set, a preset number of target feature parameters, and sklearn.DecisionTreeClassifier in Python.
[0068] In one embodiment, before determining the monitoring threshold through the growth function model based on the characterization value, the method may further include: establishing an initial monitoring threshold calculation formula through the growth function model; wherein the initial monitoring threshold calculation formula includes multiple constant terms. Further, the third training data set and the monitoring threshold calculation formula may be used to fit the values of the multiple constant terms to obtain the target monitoring threshold calculation formula.
[0069] In this embodiment, the growth function model can be expressed as follows.
[0070]
[0071] Among them, y is the indicator threshold; x is the importance of resource transfer failure; k, a, b, c, and threshold are constant terms.
[0072] In this embodiment, the monitoring threshold corresponding to each data in the third training data set can be determined according to the historical records, so as to fit the values of k, a, b, c, and threshold in the above formula to obtain the final target monitoring threshold calculation formula. In some embodiments, 0.5≤y≤0.99 and 0<x≤4 can be set to ensure a higher monitoring effect. Of course, the value ranges of y and x are not limited to the above examples. Technicians in the relevant field may make other changes under the inspiration of the technical essence of the embodiments of this specification, but as long as the functions and effects achieved are the same or similar to the embodiments of this specification, they should be covered within the protection scope of the embodiments of this specification.
[0073] In this embodiment, the curve of the monitoring threshold changing with the importance of resource transfer failure can be as follows: Figure 2 As shown in , the horizontal axis is the importance of resource transfer failure, and the vertical axis is the monitoring threshold.
[0074] In one embodiment, determining the monitoring threshold based on the characterization value through a growth function model may include: calculating the monitoring threshold corresponding to the characterization value based on the characterization value and the target monitoring threshold calculation formula; wherein the characterization value is a classification dimension of the importance of resource transfer failure.
[0075] In this embodiment, the characterization value can be substituted into the target monitoring threshold calculation formula to calculate the monitoring threshold corresponding to the characterization value, wherein the characterization value belongs to a classification dimension of the importance of resource transfer failure. Then, it can be determined whether an alarm is required based on the calculated monitoring threshold, and the judgment formula can be as follows.
[0076]
[0077] Where n is the number of failed resource transfers within the target time period; m is the total number of completed resource transfers within the target time period; and ρ is the calculated monitoring threshold. If the corresponding values are substituted into the above equation and it holds, an alarm is required; if not, no alarm is required. By setting dynamic thresholds, false alarms can be effectively reduced, thereby improving the maturity of the application.
[0078] Based on the same inventive concept, an apparatus for determining an abnormal resource transfer situation is also provided in the embodiments of this specification, such as the following embodiments. Since the principle of solving the problem by the apparatus for determining an abnormal resource transfer situation is similar to the method for determining an abnormal resource transfer situation, the implementation of the apparatus for determining an abnormal resource transfer situation can refer to the implementation of the method for determining an abnormal resource transfer situation, and the repeated parts will not be repeated. As used below, the term "unit" or "module" can be a combination of software and / or hardware that implements a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived. Figure 3 is a structural block diagram of a device for determining abnormal resource transfer conditions in an embodiment of this specification, such as Figure 3 As shown, it may include: an acquisition module 301, a processing module 302, and a determination module 303. The structure is described below.
[0079] The acquisition module 301 may be used to acquire a target resource transfer information set of a target system within a target time period;
[0080] The processing module 302 may be used to determine the representation value of the target resource transfer information set by using the target decision tree model according to the target resource transfer information set when it is determined that there is a resource transfer failure in the target resource transfer information set; wherein the target decision tree model is a pre-trained model for determining the importance of the resource transfer failure, and the representation value is used to indicate the importance of the failed resource transfer in the target resource transfer information set;
[0081] The determination module 303 may be used to determine a monitoring threshold through a growth function model based on the characterization value; wherein the monitoring threshold is used to determine whether there is an abnormal resource transfer situation within a target time period.
[0082] The embodiment of this specification also provides an electronic device, which can be specifically referred to in Figure 4The electronic device composition structure diagram of the method for determining abnormal resource transfer situations provided by the embodiments of this specification is shown, and the electronic device may specifically include an input device 41, a processor 42, and a memory 43. Among them, the input device 41 can be specifically used to input the target resource transfer information set of the target system within the target time period. The processor 42 can be specifically used to determine the characterization value of the target resource transfer information set using the target decision tree model according to the target resource transfer information set when it is determined that there is a resource transfer failure in the target resource transfer information set; wherein the target decision tree model is a pre-trained model for determining the importance of resource transfer failures, and the characterization value is used to indicate the importance of failed resource transfers in the target resource transfer information set; based on the characterization value, the monitoring threshold is determined by the growth function model; wherein the monitoring threshold is used to determine whether there is an abnormal resource transfer situation within the target time period. The memory 43 can be specifically used to store parameters such as the characterization value and the monitoring threshold.
[0083] In this embodiment, the input device can specifically be one of the main devices for information exchange between the user and the computer system. The input device can include a keyboard, a mouse, a camera, a scanner, a light pen, a handwriting input board, a voice input device, etc.; the input device is used to input the original data and the program for processing these numbers into the computer. The input device can also obtain and receive data transmitted from other modules, units, and devices. The processor can be implemented in any appropriate manner. For example, the processor can take the form of a computer-readable medium, a logic gate, a switch, an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a programmable logic controller, and an embedded microcontroller, etc., such as a microprocessor or a processor and a computer-readable program code (such as software or firmware) that can be executed by the (micro) processor. The memory can specifically be a memory device used to store information in modern information technology. The memory can include multiple levels. In a digital system, anything that can store binary data can be a memory; in an integrated circuit, a circuit with a storage function without a physical form is also called a memory, such as a RAM, a FIFO, etc.; in a system, a storage device with a physical form is also called a memory, such as a memory stick, a TF card, etc.
[0084] In this embodiment, the functions and effects specifically realized by the electronic device can be explained in comparison with other embodiments and will not be described in detail here.
[0085] In the implementation manner of the embodiments of the present specification, a computer storage medium for a determination method based on abnormal resource transfer situations is also provided, and the computer storage medium stores computer program instructions, which can achieve the following when the computer program instructions are executed: obtaining a target resource transfer information set of a target system within a target time period; when it is determined that there is a resource transfer failure in the target resource transfer information set, determining a characterization value of the target resource transfer information set based on the target resource transfer information set using a target decision tree model; wherein the target decision tree model is a pre-trained model for determining the importance of resource transfer failures, and the characterization value is used to represent the importance of failed resource transfers in the target resource transfer information set; based on the characterization value, determining a monitoring threshold through a growth function model; wherein the monitoring threshold is used to determine whether there is an abnormal resource transfer situation within the target time period.
[0086] In this embodiment, the storage medium includes, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a cache, a hard disk (HDD), or a memory card. The memory may be used to store computer program instructions. The network communication unit may be an interface for network connection communication set in accordance with the standard specified by the communication protocol.
[0087] In this embodiment, the functions and effects specifically implemented by the program instructions stored in the computer storage medium can be explained in comparison with other embodiments and will not be repeated here.
[0088] Obviously, those skilled in the art should understand that the modules or steps of the above-mentioned embodiments of this specification can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order from that here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. In this way, the embodiments of this specification are not limited to any specific combination of hardware and software.
[0089] Although the embodiments of this specification provide method operation steps as described in the above embodiments or flowcharts, more or fewer operation steps may be included in the method based on routine or no creative labor. In the steps where there is no necessary causal relationship logically, the execution order of these steps is not limited to the execution order provided by the embodiments of this specification. When the device or terminal product of the method described in practice is executed, it can be executed in sequence or in parallel according to the method shown in the embodiments or drawings (for example, a parallel processor or a multi-threaded processing environment).
[0090] It should be understood that the above description is for illustration and not for limitation. By reading the above description, many embodiments and many applications beyond the examples provided will be apparent to those skilled in the art. Therefore, the scope of the embodiments of this specification should not be determined with reference to the above description, but should be determined with reference to the full scope of the aforementioned claims and the equivalents possessed by these claims.
[0091] The above is only a preferred embodiment of the embodiment of this specification, and is not intended to limit the embodiment of this specification. For those skilled in the art, the embodiment of this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the embodiment of this specification shall be included in the protection scope of the embodiment of this specification.
Claims
1. A method for determining abnormal resource transfer conditions, It is characterized in that include: Obtaining a target resource transfer information set of a target system within a target time period; When it is determined that there is a resource transfer failure in the target resource transfer information set, a characterization value of the target resource transfer information set is determined using a target decision tree model according to the target resource transfer information set; wherein the target decision tree model is a pre-trained model for determining the importance of resource transfer failure, and the characterization value is used to represent the importance of failed resource transfers in the target resource transfer information set; Based on the characterization value, determining a monitoring threshold through a growth function model; wherein the monitoring threshold is used to determine whether there is an abnormal resource transfer situation within the target time period; Wherein, determining the monitoring threshold value through a growth function model based on the characterization value includes: The monitoring threshold is determined according to the following formula: Among them, y is the monitoring threshold; x is the characterization value; k, a, b, c, and threshold are constant terms.
2. The method according to claim 1, It is characterized in that Determining a characterization value of the target resource transfer information set using a target decision tree model according to the target resource transfer information set includes: When flowing to each judgment node of the target decision tree model, the corresponding target feature parameter is obtained according to the target resource transfer information set, until it flows to the leaf node of the target decision tree model to obtain the characterization value; wherein, the target decision tree model includes multiple judgment nodes, and each judgment node corresponds to a target feature parameter.
3. The method according to claim 2, It is characterized in that The target characteristic parameters include: application level, resource transfer level, resource transfer failure rate within the target time period, time period of resource transfer, scope affected by resource transfer, and number of monitoring false alarms before resource transfer.
4. The method according to claim 1, It is characterized in that Before determining the characterization value of the target resource transfer information set by using a target decision tree model according to the target resource transfer information set, the method further includes: Acquire an initial historical monitoring data set; wherein the historical monitoring data set includes multiple alarm information, and each alarm information includes multiple candidate feature parameters; Preprocessing the initial historical monitoring data set to obtain a target monitoring data set; Based on the target monitoring data set, generating a first training data set and a first test data set; The target decision tree model is constructed using the first training data set and the first test data set.
5. The method according to claim 4, It is characterized in that Using the first training data set and the first test data set to construct the target decision tree model includes: Using the first training data set and the first test data set, construct an initial decision tree model; Adding importance labels to each piece of alarm information in the first training data set to obtain a second training data set; Determine the importance ranking result of each candidate feature parameter according to the second training data set and the initial decision tree model; A preset number of target feature parameters are obtained by screening according to the importance ranking results of the candidate feature parameters; Based on the preset number of target feature parameters, the target decision tree model is constructed.
6. The method according to claim 5, It is characterized in that Based on the preset number of target feature parameters, constructing the target decision tree model includes: Determine multiple classification dimensions of the importance of the resource transfer failure based on the preset number; According to the multiple classification dimensions of the importance of the resource transfer failure, adding a resource transfer failure importance label to each alarm information in the first training data set and the first test data set to obtain a third training data set and a second test data set; The target decision tree model is constructed based on the third training data set, the second test data set and the preset number of target feature parameters.
7. The method according to claim 6, It is characterized in that Before determining the monitoring threshold through the growth function model based on the characterization value, the method further includes: Establishing an initial monitoring threshold calculation formula through a growth function model; wherein the initial monitoring threshold calculation formula contains multiple constant terms; The third training data set and the monitoring threshold calculation formula are used to fit the values of the multiple constant terms to obtain a target monitoring threshold calculation formula.
8. The method according to claim 7, It is characterized in that Based on the characterization value, determining a monitoring threshold through a growth function model also includes: Based on the characterization value and the target monitoring threshold calculation formula, the monitoring threshold corresponding to the characterization value is calculated; wherein the characterization value belongs to a classification dimension of the importance of the resource transfer failure.
9. A device for determining abnormal resource transfer conditions, It is characterized in that include: An acquisition module, used to acquire a target resource transfer information set of a target system within a target time period; A processing module, configured to determine, when determining that a resource transfer failure exists in the target resource transfer information set, a characterization value of the target resource transfer information set using a target decision tree model according to the target resource transfer information set; wherein the target decision tree model is a pre-trained model for determining the importance of resource transfer failures, and the characterization value is used to indicate the importance of failed resource transfers in the target resource transfer information set; A determination module, configured to determine a monitoring threshold through a growth function model based on the characterization value; wherein the monitoring threshold is used to determine whether there is an abnormal resource transfer situation within the target time period; Wherein, determining the monitoring threshold value through a growth function model based on the characterization value includes: The monitoring threshold is determined according to the following formula: Among them, y is the monitoring threshold; x is the characterization value; k, a, b, c, and threshold are constant terms.
10. A device for determining abnormal resource transfer conditions, It is characterized in that The method comprises a processor and a memory for storing processor-executable instructions, wherein the processor implements the steps of the method according to any one of claims 1 to 8 when executing the instructions.
11. A computer-readable storage medium, It is characterized in that Computer instructions are stored thereon, and when the instructions are executed, the steps of the method according to any one of claims 1 to 8 are implemented.
Citation Information
Patent Citations
Abnormal trading behavior monitoring method, device, computer device and storage medium
CN109509097A
Risk transaction identification method and device, server and storage medium
CN110309840A
