Financial risk control method and device based on responsibility chain mode

[0049] Example 1

[0050] see figure 1 , this embodiment provides a financial risk control method based on the chain of responsibility model, including:

[0051] Build a mandatory database and an optional database, the mandatory database includes multiple mandatory steps for risk event detection, and the optional database includes multiple optional steps for risk event detection; select all mandatory steps from the mandatory database, Or at the same time, select optional steps from the optional database to configure the detection responsibility chain of risk events; based on the detection responsibility chain, risk control detection is performed on risk events.

[0052] In the financial risk control method based on the chain of responsibility model provided by this embodiment, by constructing a mandatory database and an optional database, the user can select mandatory steps and optional steps according to the detection needs of the risk event and configure the chain of responsibility for detecting the risk event. , and use the chain of detection responsibility to conduct risk control detection for this type of risk event.

[0053] It can be seen that this embodiment can dynamically configure the detection responsibility chain. On the premise of ensuring the effectiveness of risk control detection of each risk event, the detection steps in the risk event detection process are reduced, and the risk control detection efficiency and detection flexibility are improved.

[0054] In the above embodiment, the method for constructing a mandatory database and an optional database includes: analyzing a plurality of detection steps of a plurality of risk events, classifying and counting each detection step, and calculating the usage ratio of each detection step in the risk event; When the usage ratio of the detection step exceeds the threshold, it is defined as a mandatory step and placed in the mandatory database; when the usage ratio of the detection step does not exceed the threshold, it is defined as an optional step and placed in the optional database.

[0055] like image 3 As shown, the mandatory database includes the step of verifying mandatory fields, the step of event mapping, the step of checking event status, the step of checking interface parameters, the step of checking event rules, the step of calculating interface parameters, and the step of throwing investigation records; the optional database includes Review the merchant's personal scene steps, flow control steps, execute asynchronous engine steps, execute immunization steps, calculate derived parameters steps, throw machine fingerprint steps, execute rule engine steps, and respond to special processing steps.

[0056] During specific implementation, the existing technology needs to conduct risk control detection according to the detection responsibility chain of fixed detection steps in the risk control detection process of risk events. figure 2 The shown chain of responsibility executes each detection step in turn, that is, the risk control detection of a risk event requires a total of 15 detection steps. Each of the detection steps must be executed, and each detection step corresponds to a different function. If there is an error in one of the detection steps in the middle, the detection process of the risk event will be terminated immediately. This sequential execution scenario is very suitable for the chain of responsibility. In the mode, each responsible person is responsible for his own steps. If the execution is successful, the event and the execution result will be handed over to the next responsible person. Otherwise, it will exit directly and the process will be terminated.

[0057] However, in practice, most risk events do not need to perform these 15 detection steps completely, and only need to perform part of them to complete the risk control detection. Therefore, we divide these 15 detection steps into mandatory steps and optional steps, as shown in the appendix figure 2 Shown: More than half of these are optional steps.

[0058] Mandatory steps are the steps that must be performed for each risk event, and are also indispensable. Optional steps are the steps that are not required to be performed. Each risk event can configure the optional steps that need to be performed according to its own situation. The chain of responsibility for such a risk event may be as follows: Figure 4 As shown, the simplest chain of responsibility model only includes mandatory steps, which is equivalent to only a simple detection process; such as Figure 5 As shown, a slightly more complex Chain of Responsibility model may include optional steps such as flow control; such as Image 6 As shown, a more detailed chain of responsibility model will perform more fine-grained risk control detection based on the type of derived parameters, such as wool party, blacklist, account freezing, etc. In general, compared with the full-step responsibility chain model in the prior art, the flexible configuration of the responsibility chain model in this embodiment can reduce unnecessary detection steps, greatly improve the detection efficiency of risk events, and the TPS of transactions per second. Indicators will also be greatly improved.

[0059] In the above embodiment, all mandatory steps are selected from the mandatory database, or optional steps are selected from the optional database at the same time, and the method for configuring the detection responsibility chain of the risk event includes:

[0060] Based on the detection requirements of risk events, select mandatory steps from the mandatory database by means of visual drag and drop, and configure a risk event detection responsibility chain that only includes mandatory steps; and/or, use visual drag and drop from the mandatory database Select mandatory steps from the Option database, and select optional steps from the optional database to configure a chain of responsibility for risk event detection that includes mandatory and optional steps.

[0061] In the above-mentioned embodiment, before configuring the chain of responsibility for detecting the risk event, it also includes: presetting the chain building priority of each mandatory step and each optional step, so as to realize the data from the mandatory database and/or the optional database. After the detection step is selected, the detection responsibility chain is automatically configured based on the priority order.

[0062] In specific implementation, by setting the priority of each detection step, after the user selects a group of detection steps, these detection steps can be automatically chained according to the preset priority, thereby improving the configuration efficiency of the detection responsibility chain.

[0063] Optionally, in the above embodiment, when there are multiple derived parameters in the risk event, a parallel computing mode is adopted when the step of calculating the derived parameters is performed.

[0064] During specific implementation, the most time-consuming detection step in the risk control detection process is the step of calculating derived parameters, which happens to be an optional step, and not every risk event needs to be executed. Since the step of calculating derived parameters needs to call an external interface to obtain relevant results and prepare data for subsequent execution of the asynchronous engine step, this embodiment proposes a mode of executing the external interface in parallel for the step of calculating derived parameters. Assuming that a risk event has three derived parameters that need to call external results, the calculation of derived parameters in the prior art belongs to the serial calculation mode, such as Figure 7 As shown, the total waiting time is T A +T B +T C. The derivative parameter calculation in this embodiment adopts parallel calculation, and the total waiting time is the one that takes the longest to adjust the external interface among the three parameters. Figure 8 shown, that is, T B , it is obvious that the parallel computing mode saves more time, especially when the external interface is adjusted a lot of times, the performance improvement is more obvious.

[0065] The above embodiment also includes: preset multiple event templates for use when configuring the detection responsibility chain; the event template includes common detection steps corresponding to risk events, and each detection step supports parameter adjustment and deletion and addition functions.

[0066] During specific implementation, several types of templates are provided for risk events with relatively large traffic in advance according to their scenario types, such as instant discount templates, interest-free activity templates, payment rebate activity templates, etc. These templates include general-purpose templates for such events. derived parameters, regular expressions, and optional steps to reduce the workload of manual configuration. This embodiment mainly distinguishes existing risk events and newly added risk events.

[0067] For a new risk event, the general template of this type of event can be automatically matched according to the event code. If you need special configuration for the event, you only need to do a small amount of personalized configuration on the basis of the template, such as Figure 9 shown. Optional steps are attached Figure 10 As shown, since this event is an active discount event, it will automatically match the instant discount template. The template is configured with derived parameters and rule nodes. Therefore, in addition to the mandatory steps in the actual detection process, you only need to select the calculation derived parameters, It is enough to execute the rule engine and respond to special processing.

[0068] For the stock risk event, its interface parameters, derivative parameters, rule nodes and other information have been configured. If you do not need to configure optional steps, the optional steps of the stock risk event are all executed by default, such as Figure 11 shown. Of course, we can also rearrange the corresponding optional steps according to the characteristics of the existing risk events to form a new responsibility chain, the specific method is the same as the new event.

[0069] It can be seen that this embodiment can verify the validity of the configuration information to ensure the validity of the responsibility chain. The dynamic modification of the chain of responsibility takes effect in near real time for running events, that is, it takes effect when the event is detected next time, and takes effect in real time for non-running events.

[0070] To sum up, the technical effects of the solution in this embodiment are mainly reflected in the aspects of flexibility, scalability, robustness, and performance optimization:

[0071] 1. Increase the flexibility and scalability of the risk control system. Whether it is adding or reducing detection steps, or changing the execution order between detection steps, the chain of responsibility model is fully compatible with the expansion of subsequent risk control models. At the same time, the responsibility chain of each risk event is dynamically configurable, which increases the flexibility of risk event detection.

[0072] 2. Enhance the robustness of the risk control system. In this embodiment, the bloated and complex model of the risk control system is divided into several relatively independent steps. Each step can be maintained independently, reducing the coupling between steps and enhancing the risk control system. robustness.

[0073] 3. The performance of the risk control system has been improved. It can be seen from experiments that the average TPS pressure measurement result of the system before the solution of this embodiment is 5863, and the average TPS pressure measurement result of the system after the solution of this embodiment is adopted is 9182, and the performance is improved by 50%. %above.

[0074] Embodiment 2

[0075] This embodiment provides a financial risk control device based on a chain of responsibility model, including:

[0076] A database building unit is used to construct a mandatory database and an optional database, the mandatory database includes multiple mandatory steps for risk event detection, and the optional database includes multiple optional steps for risk event detection;

[0077] a configuration unit, configured to select all mandatory steps from the mandatory database, or select optional steps from the optional database at the same time, to configure the chain of responsibility for detecting risk events;

[0078] A detection unit, configured to perform risk control detection on the risk event based on the detection responsibility chain.

[0079] Preferably, the mandatory database includes a step of checking mandatory fields, an event mapping step, a step of checking event status, a step of checking interface parameters, a step of checking event rules, a step of calculating interface parameters, and a step of throwing investigation records;

[0080] The optional database includes the steps of reviewing the merchant's personal scene, flow control step, executing asynchronous engine step, executing immune step, calculating derivative parameter step, throwing machine fingerprint step, executing rule engine step and responding special processing step.

[0081] Compared with the prior art, the beneficial effect of the financial risk control device based on the chain of responsibility mode provided by the embodiment of the present invention is the same as the beneficial effect of the financial risk control method based on the chain of responsibility mode provided by the above-mentioned first embodiment. Repeat.

[0082] Embodiment 3

[0083] This embodiment provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is run by a processor, the steps of the above-mentioned financial risk control method based on the chain of responsibility model are executed.

[0084] Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this embodiment are the same as the beneficial effects of the financial risk control method based on the chain of responsibility model provided by the above technical solution, which will not be repeated here.

[0085] Those of ordinary skill in the art can understand that all or part of the steps in the above-mentioned inventive method can be completed by instructing relevant hardware through a program, and the above-mentioned program can be stored in a computer-readable storage medium, and when the program is executed, it includes For each step of the method in the above embodiment, the storage medium may be: ROM/RAM, magnetic disk, optical disk, memory card, and the like.