Insurance premium calculation rule matching method and system

Abstract

The invention provides an insurance premium calculation rule matching method and system. According to the matching method, an existing insurance premium calculation rule is deconstructed into matching dimensions, duplicate removal is carried out, and a feature factor of each matching dimension and an associated insurance premium calculation rule set are recorded; when a corresponding insurance premium calculation rule is retrieved or inquired, analyzing a corresponding inquiry dimension (belonging to the matching dimension) according to an inquiry condition, and determining a condition value corresponding to each inquiry dimension; on the basis of each query dimension and the corresponding condition value thereof, determining a feature factor of each query matching dimension as a query feature factor; and calculating an intersection of the associated premium calculation rule sets of all query feature factors as a final matched premium calculation rule. According to the technical scheme provided by the invention, the one-by-one matching process of each premium calculation rule during query of the premium calculation rules is converted into the intersection solving process of the related rule set, so that the calculation amount of premium calculation rule matching is greatly reduced, and the premium calculation rule matching efficiency is improved.

Description

Technical Field

[0001] This application relates to the field of algorithmic processing in rule matching engines, specifically to a method and system for matching premium calculation rules. Background Technology

[0002] Currently, insurance products often set different premiums for different customer groups. Insurance agents need to input different coefficients into the corresponding algorithm formulas based on the different customer's choices to calculate the corresponding premium. As the number of calculation coefficients and corresponding rules involved in premium calculation increases, when calculating the premium for a specific customer, the corresponding coefficient values ​​are usually obtained by insurance agents using a query engine to find the applicable rules in the premium calculation rule library through business attribute input parameters, thus retrieving the coefficient values ​​configured for the corresponding premium calculation rule. For example... Figure 1 As shown, the premium calculation rules mainly consist of two parts: conditions and results. The conditions part mainly consists of factors plus operational logic, while the results part is often the corresponding configured coefficient values.

[0003] As shown in Table 1, the determination of a coefficient used in the premium calculation process for vehicle insurance involves ten rules, from rule 1001 to rule 1010 in Table 1. Each rule outputs a coefficient value. Each factor consists of a matching logic + value, which can be empty, indicating no distinction. When determining the applicable coefficient value for a specific user, it is usually necessary to fill in the user's actual situation. The computer program then matches each matching dimension of each premium calculation rule to obtain the applicable coefficient value. In the example in Table 1, each rule has four matching dimensions: "Usage Type," "Customer Age," "Dealer Code," and "Whether Vehicle Insurance is Insured." Each matching dimension has several feature factors (for example, for the matching dimension "Customer Age," it has "... "and" (Two feature factors). When a user's actual situation matches all the feature factors corresponding to a premium calculation rule across all matching dimensions, the coefficient value corresponding to that premium calculation rule is used for premium calculation. From the above matching process, it can be seen that the matching efficiency of the existing matching scheme for premium calculation rules is O(N), where N is the total number of relevant premium calculation rules.

[0004] Table 1: Premium Calculation Rules Corresponding to a Vehicle Insurance Premium Calculation Factor Value

[0005] The existing matching scheme for premium calculation rules is facing increasing business volume and improvements in premium calculation rules (business departments frequently configure and adjust the rules). As a result, the matching time is getting longer and the matching efficiency is getting worse, leading to increasingly uncontrollable or unacceptable system responsiveness. Summary of the Invention

[0006] To address the aforementioned shortcomings of existing premium calculation rule matching schemes, and to reduce the complexity of matching algorithms and improve matching efficiency, this application provides a premium calculation rule matching scheme based on the characteristics that the total number of premium calculation rules is uncontrollable, while the matching dimensions and the number of corresponding feature factors are relatively controllable. This matching scheme reduces the impact of the total number of premium calculation rules on matching efficiency through matching algorithm optimization.

[0007] The first aspect of this application provides a method for matching premium calculation rules. The matching method includes: After dividing all premium calculation rules into matching dimensions and removing duplicates, a list of matching dimensions is obtained. The feature factors of each matching dimension in the list of matching dimensions and its associated premium calculation rule set are recorded. The associated premium calculation rule set refers to the part of all premium calculation rules that matches a single feature factor. When each premium calculation rule is matched, the conditions corresponding to each matching dimension must satisfy a logical AND relationship.

[0008] The matching dimensions in the matching dimension list are prioritized according to the number of their respective feature factors, and the priority ranking results are saved. Matching dimensions with more feature factors have higher priority, and matching dimensions with the same number of feature factors have equal priority. Generally, matching dimensions with more feature factors can better exclude premium calculation rules that do not meet the query conditions, reducing the number of premium calculation rules that need to be calculated and processed subsequently.

[0009] The user-input query conditions are parsed to obtain relevant matching dimensions as query dimensions, and the corresponding condition values ​​for each query dimension are determined. Based on each query dimension and its corresponding condition value, the feature factors of each query matching dimension that match the query conditions are identified as query feature factors. The intersection of the associated premium calculation rule sets of all query feature factors is calculated as the final matched premium calculation rule.

[0010] Further, the process of finding the intersection of the associated premium calculation rule sets of all query feature factors includes: sorting the query feature factors according to the priority of their respective query dimensions, with query feature factors corresponding to higher priority query dimensions appearing first in the sorting. The intersection of the associated premium calculation rule set of the first-ranked query feature factor and the associated premium calculation rule set of the second-ranked query feature factor is performed to obtain the initial hit rule set. This initial hit rule set is taken as the current hit rule set corresponding to the third-ranked query feature factor. Following the sorting of the query feature factors, other query feature factors are sequentially used as the current query feature factors and processed as follows: the intersection of the associated premium calculation rule set of the current query feature factor and the current hit rule set is calculated. If the current query feature factor is the last query feature factor, the premium calculation rule in the intersection is taken as the final matched premium calculation rule; otherwise, the intersection is set as the current hit rule set corresponding to the next-ranked query feature factor.

[0011] In the above process of finding the intersection, since the intersection operation is first performed on the feature factors of the query dimensions that are ranked first, irrelevant premium calculation rules can be eliminated as soon as possible, further saving the amount of computation in subsequent intersection operations.

[0012] Furthermore, in order to adapt to the ever-expanding premium calculation rules and matching dimensions, the matching method also includes: when the premium calculation rules are adjusted, reconstructing the premium calculation rules.

[0013] Furthermore, the matching method also includes: if the intersection of the associated premium calculation rule set of the currently queried feature factor and the currently hit rule set is empty, then directly return a prompt message indicating that the corresponding premium calculation rule does not exist.

[0014] Corresponding to the matching method described above, a second aspect of this application also provides a matching system for premium calculation rules. The matching system includes: a rule deconstruction module and a rule filtering module.

[0015] The rule deconstruction module is used to deconstruct the premium calculation rules: it divides all premium calculation rules into matching dimensions, removes duplicates to obtain a list of matching dimensions, records the feature factors of each matching dimension in the list and its associated premium calculation rule set, and ensures that the conditions corresponding to each matching dimension satisfy a logical AND relationship when each premium calculation rule is matched; it sorts the matching dimensions in the list according to their respective number of feature factors and saves the priority sorting results, with matching dimensions having more feature factors having higher priority, and matching dimensions with the same number of feature factors having equal priority; the associated premium calculation rule set refers to the part of all premium calculation rules that matches a single feature factor.

[0016] The rule filtering module includes a query condition parsing submodule and a rule matching submodule. The query condition parsing submodule parses the query conditions input by the user to obtain relevant matching dimensions as query dimensions, and determines the condition value corresponding to each query dimension.

[0017] The rule matching submodule identifies the feature factors of each query matching dimension that are matched by the query condition based on each query dimension and its corresponding condition value, and uses them as query feature factors; it calculates the intersection of the associated premium calculation rule sets of all query feature factors as the final matched premium calculation rule.

[0018] Furthermore, the matching system also includes a rule destructuring trigger module. When an adjustment to the premium calculation rule occurs, the rule destructuring trigger module triggers the rule destructuring module to re-destruct the premium calculation rule. The implementation details of this matching system are the same as those of the aforementioned premium calculation rule matching method, and will not be repeated here.

[0019] A third aspect of this application also provides a computer-readable storage medium. The computer-readable storage medium contains a computer program, which, when processed by a computer, describes a matching method for the aforementioned premium calculation rules.

[0020] The technical solution provided in this application takes advantage of the fact that the total number of premium calculation rules is uncontrollable, while the number of matching dimensions and feature factors is relatively controllable. It transforms the process of matching each premium calculation rule one by one into a process of finding the intersection of related rule sets, which greatly reduces the amount of computation for premium calculation rule matching, improves the efficiency of premium calculation rule matching, and reduces the system overhead during matching operations. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the specific embodiments of this application or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram illustrating the matching process for existing premium calculation rules.

[0023] Figure 2 A flowchart of one embodiment of the matching method for providing premium calculation rules in this application.

[0024] Figure 3 A schematic diagram of a premium calculation rule matching system provided for this application in one embodiment.

[0025] Explanation of reference numerals in the attached figures: 100 - Premium calculation rule matching system, 111 - Rule deconstruction module, 112 - Rule filtering module, 1121 - Query condition parsing submodule, 1122 - Rule matching submodule, 113 - Storage unit, 114 - Rule deconstruction trigger module. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0027] like Figure 2 In the illustrated embodiment, the premium calculation rule matching method provided in this application includes the following steps: S1. Deconstructing Premium Calculation Rules: This step includes: First, dividing all premium calculation rules into matching dimensions, removing duplicates to obtain a list of matching dimensions, and recording the feature factors of each matching dimension in the list and its associated premium calculation rule set. The associated premium calculation rule set refers to the portion of all premium calculation rules that matches a single feature factor. When each premium calculation rule is matched, the conditions corresponding to each matching dimension must satisfy a logical AND relationship. For example, Figure 1 The premium calculation rules shown can be deconstructed into two matching dimensions: "gender" and "age".

[0028] Then, the matching dimensions in the matching dimension list are sorted by priority according to the number of their respective feature factors, and the priority sorting results are saved. Matching dimensions with more feature factors have higher priority, and matching dimensions with the same number of feature factors have equal priority. For Figure 3 In the example shown, we will get a sorting order like "Dealer Code" --> "Usage Type" --> "Customer Age" --> "Whether Vehicle Damage Insurance is Purchased". Generally, matching dimensions with more feature factors can better exclude premium calculation rules that do not meet the query conditions, thus reducing the number of premium rules that need to be calculated and processed subsequently, and improving the matching efficiency of premium calculation rules.

[0029] S2. Parse query conditions: Parse the user-input query conditions to obtain relevant matching dimensions as query dimensions, and determine the condition values ​​corresponding to each query dimension. For example, Figure 1 The query conditions are parsed into "gender-male" and "age-25" in the form of "query dimension-condition value".

[0030] S3. Match premium calculation rules based on the parsing results of query conditions and the deconstruction results of premium calculation rules: Based on each query dimension and its corresponding condition value, identify the feature factors of each query matching dimension that the query condition matches as query feature factors. Calculate the intersection of the associated premium calculation rule sets of all query feature factors as the final matched premium calculation rule.

[0031] Further, the process of finding the intersection of the associated premium calculation rule sets of all query feature factors includes: sorting the query feature factors according to the priority of their respective query dimensions, with query feature factors corresponding to higher priority query dimensions appearing first in the sorting. The intersection of the associated premium calculation rule set of the first-ranked query feature factor and the associated premium calculation rule set of the second-ranked query feature factor is performed to obtain the initial hit rule set. This initial hit rule set is taken as the current hit rule set corresponding to the third-ranked query feature factor. Following the sorting of the query feature factors, other query feature factors are sequentially used as the current query feature factors and processed as follows: the intersection of the associated premium calculation rule set of the current query feature factor and the current hit rule set is calculated. If the current query feature factor is the last query feature factor, the premium calculation rule in the intersection is taken as the final matched premium calculation rule; otherwise, the intersection is set as the current hit rule set corresponding to the next-ranked query feature factor.

[0032] Based on the above intersection process, it's clear that the later the intersection of the set of associated premium calculation rules for the current query feature factor and the current hit rule set involves, the fewer premium calculation rules are involved, resulting in less computational resource consumption. Furthermore, this process minimizes the number of premium calculation rules used for multiple intersection calculations, further reducing the computational load required for premium calculation rule matching.

[0033] Furthermore, the matching method also includes: if the intersection of the associated premium calculation rule set of the currently queried feature factor and the currently hit rule set is empty, then directly return a prompt message indicating that the corresponding premium calculation rule does not exist.

[0034] Preferably, when the premium calculation rules are adjusted, the premium calculation rules are re-deconstructed. Compared to matching the premium calculation rules, modifications and additions to the premium calculation rules themselves are low-frequency events. Thus, by sacrificing a small amount of overhead, the above matching method can adapt to the ever-expanding premium calculation rules and matching dimensions.

[0035] Corresponding to the matching method described above, this application also provides a matching system for premium calculation rules. For example... Figure 2In the embodiment shown, the matching system 100 provided in this application includes: a rule deconstruction module 111, a rule filtering module 112, and a storage unit 113.

[0036] The rule deconstruction module 111 is used to perform the following deconstruction processing on the premium calculation rules: dividing all premium calculation rules into matching dimensions, obtaining a list of matching dimensions after deduplication, recording the feature factors of each matching dimension in the list of matching dimensions and its associated premium calculation rule set (stored in storage unit 113), and when each premium calculation rule is matched, the conditions corresponding to each matching dimension must satisfy the logical AND relationship; sorting the matching dimensions in the list of matching dimensions according to their respective number of feature factors, and saving the priority sorting results (stored in storage unit 113), the matching dimensions with more feature factors have higher priority, and the matching dimensions with the same number of feature factors have the same priority; the associated premium calculation rule set refers to the part of all premium calculation rules that matches a single feature factor.

[0037] The rule filtering module 112 includes a query condition parsing submodule 1121 and a rule matching submodule 1122. The query condition parsing submodule 1121 parses the query conditions input by the user to obtain relevant matching dimensions as query dimensions, and determines the condition value corresponding to each query dimension.

[0038] The rule matching submodule 1122, based on each query dimension and its corresponding condition value, confirms the feature factors of each query matching dimension that the query condition matches, and uses them as query feature factors; it calculates the intersection of the associated premium calculation rule sets of all query feature factors as the final matched premium calculation rule.

[0039] like Figure 3 As shown, the matching system also has a rule destructuring trigger module 114. When a premium calculation rule is adjusted, the rule destructuring trigger module 114 triggers the rule destructuring module 111 to re-destruct the premium calculation rule. The implementation details of this matching system are the same as the implementation details of the aforementioned premium calculation rule matching method, and will not be repeated here.

[0040] In addition, this application also provides a computer-readable storage medium. The computer-readable storage medium contains a computer program, which, when processed by a computer, describes a matching method for the aforementioned premium calculation rules. The computer-readable storage medium includes, but is not limited to, hard disks, USB flash drives, RAM, flash memory, and network storage.

[0041] The above description is merely an embodiment of this application and is not intended to limit this application. For those skilled in the art, the technical solutions provided in this application can be modified and varied in various ways. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for matching premium calculation rules, characterized in that, The matching method includes: Deconstructing the premium calculation rules: All premium calculation rules are divided into matching dimensions, and after deduplication, a list of matching dimensions is obtained. The feature factors of each matching dimension in the list and its associated premium calculation rule set are recorded. The matching dimensions in the list are prioritized according to the number of their respective feature factors, and the priority ranking results are saved. Matching dimensions with more feature factors have higher priority, and matching dimensions with the same number of feature factors have equal priority. The associated premium calculation rule set refers to the portion of all premium calculation rules that matches a single feature factor. When each premium calculation rule is matched, the conditions corresponding to each matching dimension must satisfy a logical AND relationship. Parse query conditions: Parse the query conditions entered by the user to obtain the relevant matching dimensions as query dimensions, and determine the condition value corresponding to each query dimension; The premium calculation rule is matched based on the parsing results of the query conditions: Based on each query dimension and its corresponding condition value, the feature factors of each query matching dimension that the query condition matches are identified as query feature factors; the intersection of the associated premium calculation rule sets of all query feature factors is calculated as the final matched premium calculation rule.

2. The matching method as described in claim 1, characterized in that, The intersection of the associated premium calculation rule sets of all query feature factors is obtained as follows: Query feature factors are sorted according to the priority of their respective query dimensions, with those having higher priority corresponding to the query dimension appearing first in the sorted order; the intersection of the associated premium calculation rule set of the first-ranked query feature factor and the associated premium calculation rule set of the second-ranked query feature factor is performed to obtain the initial hit rule set; the initial hit rule set is taken as the current hit rule set corresponding to the third-ranked query feature factor, and other query feature factors are processed sequentially as the current query feature factors according to their sorting order: the intersection of the associated premium calculation rule set of the current query feature factor and the current hit rule set is calculated. If the current query feature factor is the last query feature factor, the premium calculation rule in the intersection is taken as the final matched premium calculation rule; otherwise, the intersection is set as the current hit rule set corresponding to the next-ranked query feature factor.

3. The matching method as described in claim 2, characterized in that, The matching method further includes: if the intersection of the associated premium calculation rule set of the currently queried feature factor and the currently hit rule set is empty, then directly return a prompt message that the corresponding premium calculation rule does not exist.

4. The matching method as described in any one of claims 1-3, characterized in that, The matching method further includes: when the premium calculation rules are adjusted, re-deconstructing the premium calculation rules.

5. A matching system for premium calculation rules, characterized in that, The matching system includes: a rule deconstruction module and a rule filtering module; The rule deconstruction module is used to deconstruct the premium calculation rules: it divides all premium calculation rules into matching dimensions, removes duplicates to obtain a list of matching dimensions, records the feature factors of each matching dimension in the list and its associated premium calculation rule set; it sorts the matching dimensions in the list according to the number of their respective feature factors and saves the priority sorting results, with matching dimensions having more feature factors having higher priority, and matching dimensions with the same number of feature factors having equal priority; the associated premium calculation rule set refers to the part of all premium calculation rules that matches a single feature factor; The rule filtering module includes: a query condition parsing submodule and a rule matching submodule; The query condition parsing submodule parses the query conditions input by the user to obtain relevant matching dimensions as query dimensions, and determines the condition value corresponding to each query dimension. The rule matching submodule identifies the feature factors of each query matching dimension that are matched by the query condition based on each query dimension and its corresponding condition value, and uses them as query feature factors; it calculates the intersection of the associated premium calculation rule sets of all query feature factors as the final matched premium calculation rule.

6. The matching system as described in claim 5, characterized in that, The rule matching submodule sorts the query feature factors according to the priority of their respective query dimensions, with query feature factors of higher priority corresponding to the query dimension being sorted first. It calculates the intersection of the associated premium calculation rule set of the first-ranked query feature factor and the associated premium calculation rule set of the second-ranked query feature factor as the initial hit rule set. The initial hit rule set is then taken as the current hit rule set corresponding to the third-ranked query feature factor. Following the sorting of the query feature factors, other query feature factors are sequentially treated as current query feature factors and processed as follows: the intersection of the associated premium calculation rule set of the current query feature factor and the current hit rule set is calculated. If the current query feature factor is the last query feature factor, the premium calculation rule in the intersection is taken as the final matched premium calculation rule; otherwise, the intersection is set as the current hit rule set corresponding to the next-ranked query feature factor.

7. The matching system as described in claim 6, characterized in that, If the intersection of the associated premium calculation rule set of the currently queried feature factor and the currently hit rule set is empty, the rule matching submodule will directly return a prompt message indicating that the corresponding premium calculation rule does not exist.

8. The matching system as described in any one of claims 5-7, characterized in that, The matching system also includes a rule destructuring trigger module; when the premium calculation rule is adjusted, the rule destructuring trigger module triggers the rule destructuring module to re-destruct the premium calculation rule.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium contains a computer program that, when processed by a computer, implements the matching method for the premium calculation rules according to any one of claims 1-4.

Images