Vehicle insurance claim settlement operation auxiliary method based on multi-factor model

Abstract

The invention discloses a vehicle insurance claim settlement operation auxiliary method based on a multi-factor model, and the method comprises the following steps: S1), setting a plurality of rule factors, and binding each rule factor with a business model; s2) establishing a survey rule factor library, a loss assessment rule factor library and an auditing rule factor library, setting links of each factor library application, and binding each factor to the corresponding factor library; s3) setting an affiliation branch company, a rule type and a trigger link selection rule factor, configuring a rule factor trigger condition, and serially connecting the factor trigger condition with AND, OR and NOT logic to form an auxiliary operation rule; s4) setting a rule triggering reminding type and reminding content; and S5) when the auxiliary operation rule is stored, generating a drools rule in real time and thermally loading the drools rule to a rule engine, thereby realizing timely effectiveness of the rule. According to the method, the problem that intensive and large-scale claim settlement popularization is hindered due to the fact that the regional operation requirement difference is too large can be solved.

Description

Technical Field

[0001] The present invention relates to an assisting method for automobile insurance claims, and in particular to an assisting method for automobile insurance claims based on a multi-factor model. Background Art

[0002] Auto insurance claims are underwritten in different regions, and there are differences in claims policies for accident cases. With the steady advancement of online and intensive auto insurance claims, cross-regional case handling has become the norm, and the rules and processing to assist cross-regional operations have become increasingly complex. Hard-coded rule reminders are now commonly used to implement this. When the operating rules change, a complete release process is required to implement it. In terms of complexity, flexibility, and data processing timeliness, it is difficult to meet business development. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a method for assisting automobile insurance claims settlement operations based on a multi-factor model, which can solve the problem that the promotion of intensive and large-scale claims settlement is hindered by the large differences in regional operation requirements.

[0004] The technical solution adopted by the present invention to solve the above technical problems is to provide a method for assisting auto insurance claims processing based on a multi-factor model, comprising the following steps: S1) setting multiple rule factors and binding each rule factor to a business model; S2) establishing an inspection rule factor library, a damage assessment rule factor library and an audit rule factor library, setting the application links of each factor library, and binding each factor to the corresponding factor library; S3) setting the affiliated branch, rule type, and trigger link selection rule factor, configuring the rule factor triggering conditions, and forming an auxiliary operation rule by logically connecting the factor triggering conditions with AND, OR, and NOT; S4) setting the type and content of the rule trigger reminder; S5) when saving the auxiliary operation rule, generating drools rules in real time and hot loading them to the rule engine to realize the timely effectiveness of the rules.

[0005] Furthermore, the step S1 sets the name, type and constraint condition of the rule factor, and the type of the rule factor includes time, value, string and enumeration.

[0006] Furthermore, the rule factors in step S1 include a liability ratio factor, a damage amount factor, a target vehicle brand factor and an inspection area factor. The liability ratio factor in step S1 is bound to the liability ratio attribute of the inspection report business model, the damage amount factor is bound to the damage amount attribute of the damage estimate form, the target vehicle brand factor is bound to the vehicle brand attribute of the insured vehicle of the order copying information business model, and the inspection area factor is bound to the inspection area attribute of the inspection report business model.

[0007] Furthermore, when binding the business model in step S1, the business model is compared with the business model stored in the database. If there is an abnormal situation in which the data business model is not matched, rebinding is prompted through an abnormal prompt.

[0008] Furthermore, the application links of each factor library set in step S2 include page opening, page entry and data submission; when the front-end page is initialized, the business model in the factor library of the corresponding link is loaded, and the page elements and trigger events are cyclically compared and bound; when the page loading is completed or the bound page element value changes or triggers business submission, the auxiliary operation rules are called.

[0009] Furthermore, for the survey page, the responsibility ratio business model is bound to the survey page responsibility ratio input box. When the survey page is initialized and the survey factor library is loaded, the responsibility ratio factor trigger link is traversed to the page entry. The responsibility ratio input box monitors the leaving of the input focus and triggers the rule execution.

[0010] Furthermore, the type of the reminder triggered by the rule in step S4 is blocked operation, highlighted reminder or forced reading.

[0011] Furthermore, the step S5 includes parsing the rule into a drl rule statement of the drools rule engine through the freemarker template engine, and updating the kieContainer of the rule engine.

[0012] Furthermore, the step S5 configures the survey submission auxiliary operation rules as follows: when saving the rule, first find the survey rule template, then solidify the template content into drl regular statements, use the template engine to traverse and replace rule factors, operators and rule execution results, produce executable drl statements, and finally execute the KieContainerImpl.updateToKieModule method to hot update the rule library.

[0013] Furthermore, the step S5 uses the RETE algorithm to construct a network composed of nodes, wherein the nodes represent different conditions, and the nodes in the network are divided into working memory nodes and conditional nodes; the working memory nodes store current facts, and the conditional nodes are used to compare these facts; when new facts are input, they are matched with the conditions in the network; when a fact enters the working memory, it is checked whether the fact meets the conditions of the conditional nodes; if so, the corresponding rules are activated.

[0014] Compared with the prior art, the present invention has the following beneficial effects: the motor insurance claims processing assistance method based on a multi-factor model provided by the present invention enables the operation center to create differentiated claims policy assistance rules for different regions. Operators do not need to understand the special and complex claims policies of various regions regarding rescue fees, special agreements, vehicle properties, etc. During the operation, the system quickly identifies risk points and gives operation instructions, solving the problem of hindering the promotion of intensive and large-scale claims processing due to large differences in regional operation requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 The present invention is an auxiliary flow chart of the auto insurance claims settlement operation based on the multi-factor model. DETAILED DESCRIPTION

[0016] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0017] Figure 1 The present invention is an auxiliary flow chart of the auto insurance claims settlement operation based on the multi-factor model.

[0018] See also Figure 1 The invention provides a method for assisting automobile insurance claims based on a multi-factor model, which factors and models the automobile insurance claims business model, and uses a drools rule engine with hot deployment characteristics to configure and execute auxiliary operation rules for the automobile insurance claims business model.

[0019] Different from the existing hard-coded rule method, the auxiliary operation is now implemented by using independent rule management. The specific algorithm is as follows:

[0020] (1) Define factors: define the name, type (such as time, value, string, enumeration, etc.), and constraints (such as value range, etc.) of the rule factor

[0021] (2) Factor binding business models: for example, the liability ratio factor is bound to the liability ratio attribute of the survey report business model, the damage amount factor is bound to the damage amount attribute of the loss estimate, the target vehicle brand factor is bound to the vehicle brand attribute of the insured vehicle in the order information business model, and the survey area factor is bound to the survey area attribute of the survey report business model. When binding a business model, it will be compared with the business model stored in the database. If there is an abnormal situation where the data business model is not matched, it will be re-bound through the abnormal prompt.

[0022] (3) Define the factor library, establish rule factor libraries for investigation, damage assessment, and audit, define the application links of the factor library, including page opening, page entry, and data submission, and bind the factors to the factor library.

[0023] (4) Define rules: define the affiliated branch, rule type, trigger link (page loading, button trigger, etc.), select the factors that the rule needs to reference, and configure the rule factor trigger conditions (such as the responsibility ratio is no responsibility, etc.), and connect the factor trigger conditions in series with AND, OR, or NOT logic to form auxiliary operation rules.

[0024] (5) Define the type and content of reminders triggered by rules; reminder types include blocked jobs, highlighted reminders, and forced reading, etc.

[0025] (6) Save the auxiliary operation rules, parse the rules into drl rule statements of the drools rule engine through the freemarker template engine, and update the kieContainer of the rule engine; for example, configure an auxiliary operation rule for inspection submission, and the rule content is to remind you to verify the vehicle type when the target vehicle type is not less than 6 seats. When saving the rule, find the inspection rule template, solidify the drl regular statement in the template content, use the template engine to traverse and replace the rule factors, operators and rule execution results, produce executable drl statements, execute the KieContainerImpl.updateToKieModule method to hot update the rule library, and realize hot deployment that the original technology cannot achieve. The rules do not need to be published and take effect in real time.

[0026] (7) Page element binding: When the front-end page is initialized, the business model in the factor library of the corresponding link is loaded, and the page elements and trigger events are cyclically compared and bound. For example, on the survey page, the responsibility ratio business model is bound to the responsibility ratio input box of the survey page. When the survey page is initialized and the survey factor library is loaded, the responsibility ratio factor trigger link is traversed to the page input. After the responsibility ratio input box detects that the input focus has left, the rule execution is triggered.

[0027] (8) When the page is loaded or the bound page element value changes or the business submission is triggered, the auxiliary job rule is called.

[0028] (9) When executing rules, the page business model data is used to generate rule execution parameters according to the rule factor binding relationship, and the drools rule engine is used to execute auxiliary job rules.

[0029] The front end blocks submission or displays auxiliary homework reminders in a personalized manner based on the rule type.

[0030] The present invention builds a rule base by itself, visually configures auxiliary operation rules, and generates drools rules in real time and hot loads them to the rule engine when saving auxiliary operation rules, so as to realize the timely effectiveness of the rules, thereby achieving flexible regulation of auxiliary operation reminders. Using the built-in RETE algorithm of drools (designed and invented by DrCharles L.Forgy of Carnegie Mellon University), the RETE network is first generated according to the facts in the rules, the new facts are inserted into the working memory, and the RETE network is updated; the nodes of the network are matched to find the rules that meet the conditions, and when all conditions are met, the corresponding rules are triggered to execute; the RETE algorithm first builds a network composed of nodes, which represent different conditions, and the nodes in the network are divided into two categories: working memory nodes and conditional nodes. The working memory nodes store the current facts, and the conditional nodes are used to compare these facts; when new facts are input into the system, the RETE network will start to process these facts. Each fact will be matched with the conditions in the network. When a fact enters the working memory, the algorithm will check whether the fact meets the conditions of the conditional node. If it is met, the corresponding rule is activated. Unlike the traditional full rematch method, RETE uses an incremental update method to detect only the changed parts, thereby improving efficiency. Once the rule is activated, the system will execute the relevant actions or rules to produce corresponding results or further reasoning. The application of the RETE algorithm enables the auxiliary operation rules to be efficiently executed within 0.5 seconds.

[0031] Although the present invention has been disclosed as above with preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art may make some modifications and improvements without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be based on the definition of the claims.

Claims

1. A method for assisting automobile insurance claims settlement based on a multi-factor model, characterized in that: The steps include: S1) setting multiple rule factors and binding each rule factor to a business model; S2) Establishing the factor library of survey rules, the factor library of damage assessment rules and the factor library of audit rules, setting the application links of each factor library and binding each factor to the corresponding factor library; S3) Setting the affiliated branch, rule type, triggering link selection rule factor, configuring the rule factor triggering condition, and connecting the factor triggering condition with AND, OR, or NOT logic to form an auxiliary operation rule; S4) Setting the type and content of the reminder triggered by the rule; S5) When saving auxiliary operation rules, drools rules are generated in real time and hot loaded into the rule engine to realize the timely effectiveness of the rules.

2. The method for assisting automobile insurance claims settlement based on a multi-factor model as claimed in claim 1, characterized in that: The step S1 sets the name, type and constraint of the rule factor, and the type of the rule factor includes time, value, string and enumeration.

3. The method for assisting automobile insurance claims settlement based on a multi-factor model as claimed in claim 1, characterized in that: The rule factors in step S1 include a liability ratio factor, a damage amount factor, a target vehicle brand factor and an inspection area factor. The liability ratio factor in step S1 is bound to the liability ratio attribute of the inspection report business model, the damage amount factor is bound to the damage amount attribute of the damage estimate form, the target vehicle brand factor is bound to the vehicle brand attribute of the insured vehicle of the order copying information business model, and the inspection area factor is bound to the inspection area attribute of the inspection report business model.

4. The method for assisting automobile insurance claims settlement based on a multi-factor model as claimed in claim 1, characterized in that: When binding the business model in step S1, the business model is compared with the business model stored in the database. If the data business model is not matched, rebinding is performed through an exception prompt.

5. The method for assisting automobile insurance claims settlement based on a multi-factor model as claimed in claim 1, characterized in that: The application links of each factor library set in step S2 include page opening, page entry and data submission; when the front-end page is initialized, the business model in the factor library of the corresponding link is loaded, and the page elements and trigger events are cyclically compared and bound; when the page loading is completed or the bound page element value changes or triggers business submission, the auxiliary operation rules are called.

6. The method for assisting automobile insurance claims settlement based on a multi-factor model as claimed in claim 5, characterized in that: For the survey page, the responsibility ratio business model is bound to the survey page responsibility ratio input box. When the survey page is initialized and the survey factor library is loaded, the responsibility ratio factor trigger link is traversed to the page entry. The responsibility ratio input box monitors the leaving of the input focus and triggers the rule execution.

7. The method for assisting automobile insurance claims settlement based on a multi-factor model as claimed in claim 1, characterized in that: The type of the rule-triggered reminder in step S4 is a blocked job, a highlighted reminder, or a mandatory reading.

8. The method for assisting automobile insurance claims settlement based on a multi-factor model as claimed in claim 1, characterized in that: The step S5 includes parsing the rule into a drl rule statement of the drools rule engine through the freemarker template engine, and updating the kieContainer of the rule engine.

9. The method for assisting automobile insurance claims settlement based on a multi-factor model as claimed in claim 8, characterized in that: The step S5 configures the survey submission auxiliary operation rules as follows: when saving the rule, first find the survey rule template, then solidify the template content into drl regular statements, use the template engine to traverse and replace rule factors, operators and rule execution results, produce executable drl statements, and finally execute the KieContainerImpl.updateToKieModule method to hot update the rule library.

10. The method for assisting automobile insurance claims settlement based on a multi-factor model as claimed in claim 8, characterized in that: The step S5 uses the RETE algorithm to construct a network composed of nodes, wherein the nodes represent different conditions, and the nodes in the network are divided into working memory nodes and conditional nodes; the working memory nodes store current facts, and the conditional nodes are used to compare these facts; when new facts are input, they are matched with the conditions in the network; when a fact enters the working memory, it is checked whether the fact meets the conditions of the conditional nodes; if so, the corresponding rules are activated.

Images