CN114595252B - Self-iteration method and system of policy matching mechanism - Google Patents
Self-iteration method and system of policy matching mechanism Download PDFInfo
- Publication number
- CN114595252B CN114595252B CN202210107193.5A CN202210107193A CN114595252B CN 114595252 B CN114595252 B CN 114595252B CN 202210107193 A CN202210107193 A CN 202210107193A CN 114595252 B CN114595252 B CN 114595252B
- Authority
- CN
- China
- Prior art keywords
- enterprise
- policy
- feature
- library
- matching
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2455—Query execution
- G06F16/24564—Applying rules; Deductive queries
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Computational Linguistics (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a dynamic self-iteration method and a system of a policy matching mechanism, wherein the system comprises: an event trigger, a policy router, a feature vector generator, a feature matcher, a rule engine, and a recommendation engine. The invention utilizes part of enterprise information for carrying out policy reporting in advance, introduces feature vectors, a rule engine technology and establishes a feature matching space, the state conversion of the enterprise reporting policy process drives the updating of a policy matching mechanism in an event form, the feature matching space is automatically and continuously updated along with the increase of the number of the enterprise reporting and the complementation of state data of the reporting process, the matching rule is optimized, a more accurate policy matching set is generated, the workload of manual participation is greatly reduced, and the requirements of evolution and optimization of a matching algorithm are solved.
Description
Technical Field
The invention belongs to the technical field of computer software, relates to a data processing technology, and particularly relates to a self-iteration method and a self-iteration system of a policy matching mechanism.
Background
In recent years, the policy strength of the general financial field in China is gradually enhanced, the effect is obvious, but with the change of the internal and external economic situation, a plurality of enterprises still face great challenges in management. How to make a plurality of enterprises, especially a large number of small and medium-sized enterprises occupying the main body of the enterprise in China enjoy the policy bonus of affording finance, excite the market vitality, improve the commercial environment, and continuously serve the enterprise with full life cycle is an important problem.
There have been some organizations exploring and practicing how policies are recommended to enterprises. Most of the existing technologies automatically analyze, correlate and assist in extracting labels based on NLP, knowledge graph and other technologies, and correct labels by manual intervention, so that the aim of policy recommendation is achieved through bidirectional matching of enterprises and policy labels. In actual production practice, since the content of the policy file is unstructured, there may be many brand new proper noun terms in the content description, and some word descriptions have implicit meanings, so the accuracy of policy matching depends on the "expertise" of policy data analysis, meaning that if policy recommendation is desired to be more accurate, a large number of specialized policy analysts are required to participate in content analysis and label extraction, and the requirements of rapid increase of policies and rapid and accurate policy matching cannot be met.
Disclosure of Invention
In order to solve the problems, the invention discloses a dynamic self-iteration method and a system of a policy matching mechanism.
In order to achieve the above purpose, the technical scheme of the invention is as follows:
a dynamic self-iterating system of a policy matching mechanism, comprising: the system comprises an event trigger, a strategy router, a feature vector generator, a feature matcher, a rule engine and a recommendation engine;
the event trigger is used for triggering subsequent actions according to a trigger strategy;
the strategy router is used for receiving the message notification of the event trigger and determining the executed specific strategy;
the feature vector generator is used for generating or updating enterprise feature vectors for specific enterprises or enterprises meeting the conditions according to the execution strategy determined by the strategy router, and writing the enterprise feature vectors into the enterprise feature library;
the feature matcher is used for traversing and searching in the enterprise feature library according to the sequence of the first sequence and the sequence of the second sequence, arranging and combining feature sets according to requirements to form feature set vectors, and writing the feature set vectors into the enterprise feature library;
the rule engine is used for reading the enterprise feature library, generating a rule logic expression and an expression group according to the feature set generated by the feature matcher, and writing the rule logic expression and the expression group into the rule library;
the recommendation engine is used for screening out a qualified enterprise list based on the current policy execution rule expression, adding the current policy into the policy matching result set of each enterprise, and establishing a result set for the enterprise.
Further, the event trigger subscribes to various enterprise reporting process events, when the event occurs, the event trigger firstly writes in an event library, reads out a strategy library, and decides whether to trigger subsequent actions according to a preset event trigger strategy.
Further, the enterprise reporting process event refers to the enterprise reporting policy by itself, filling in related forms and submitting enterprise information, and dynamically updating the process state of the policy reporting until reporting success or reporting failure.
Further, the preset event triggering strategy comprises at least one of the following strategies:
a. triggering the next action when the number of submitted declared enterprises reaches a threshold value N1; n1=a preset value;
b. triggering the next action when the accepted enterprise proportion reaches a threshold value N2; n2=the number of businesses in the current policy acceptance/the total number of businesses declared by the current policy;
c. triggering the next action when the successfully declared enterprise proportion reaches a threshold value N3; n3=current policy claims number of successful enterprises/current policy claims total enterprises.
Further, the policy router is configured to receive a message notification from the event trigger, read the policy repository, and determine, according to the number of enterprises in each state stage of reporting, a specific policy to be executed, where the policy router includes the following branches:
policy branch a: recommending a default policy result set to enterprise matching, screening and sorting according to a plurality of conditions such as release time, viewing quantity, collection quantity, text issuing units and the like when selecting policies, wherein the policy result set takes effect globally;
policy branch B: recommending a manually specified policy result set to the enterprise matching, and manually specifying the ordering of the result set when the policy is selected;
policy branch C: and recommending a policy result set generated after a series of feature calculation, extraction, rule generation and data screening to the enterprise matching.
Further, the policies are decided by the policy router to be executed singly or in combination; and configuring the sequence of the recommended result set in the strategy library when the combination is executed, and displaying the result set according to the sequence of the sequence.
Further, the feature vector generator reads the enterprise information base, acquires all attribute information of the enterprise, and generates feature vectors according to preset attribute ordering; and when the attribute information of the enterprise is illegal or invalid, carrying out emptying processing, and if the quantity of the illegal or invalid attribute information reaches a preset threshold value of the quantity of all information of the current enterprise, directly discarding the feature vector of the current enterprise.
Further, for the current column, the feature matcher stores technical metadata and business metadata information of the current column in an enterprise feature library, and generates a feature set of the current column according to metadata in the feature library, wherein the feature set is a single value, a set of a plurality of values or a certain range; and finally, arranging and combining the feature sets as required to form feature set vectors and writing the feature set vectors into an enterprise feature library.
Further, in a single expression group generated by the rule engine, all expressions evaluate to true, and then the matching rule is considered to hit; the generation logic of the single expression includes at least one of the following:
single value matching: enterprise attribute Value = feature set Value; multi-value matching: the enterprise attribute Value belongs to the feature set; time range matching: the enterprise attribute Value is within the time range specified by the feature set; prefix matching; the numerical ranges match.
The invention provides a dynamic self-iteration method of a policy matching mechanism, which comprises the following steps:
step 1, an enterprise self-declares a policy, fills in a related form and submits enterprise information, and the process state of the policy declaration is dynamically updated to generate an enterprise declaration process event;
step 2, subscribing various upstream enterprise reporting process events by the event trigger, internally maintaining an event library, writing the event library when the event occurs, reading a strategy library, and determining whether to trigger subsequent actions according to a preset event triggering strategy;
step 3, reading a strategy library according to the information notification of the event trigger, and determining the executed specific strategy according to the number of enterprises in each state stage;
step 4, generating or updating enterprise feature vectors for a specific enterprise or an enterprise meeting the conditions according to the determined executing strategy, and writing the enterprise feature vectors into an enterprise feature library;
step 5, traversing and searching in the enterprise feature library according to the sequence of the first step and the sequence of the second step, arranging and combining feature sets according to requirements to form feature set vectors, and writing the feature set vectors into the enterprise feature library;
step 6, reading an enterprise feature library, generating a rule logic expression and an expression group according to a feature set generated by a feature matcher, and writing the rule logic expression and the expression group into the rule library;
and 7, screening out a qualified enterprise list based on the current policy execution rule expression, adding the current policy into a policy matching result set of each enterprise, and establishing a result set for the enterprise.
The beneficial effects of the invention are as follows:
the invention focuses on one end of an enterprise, utilizes partial enterprise information which is used for conducting policy reporting in advance, introduces feature vectors, adopts a rule engine technology, establishes a feature matching space, drives the updating of a policy matching mechanism in an event form by state conversion of an enterprise reporting policy process, automatically and continuously updates the feature matching space along with the increase of the number of the enterprise reporting and the complement of state data of the reporting process, optimizes a matching rule, generates a more accurate policy matching set, greatly reduces the workload of manual participation, and solves the requirements of evolution and optimization of a matching algorithm.
Drawings
FIG. 1 is a diagram of a dynamic self-iterating system architecture of the policy matching mechanism provided by the present invention.
Fig. 2 is a schematic diagram of feature vector generation.
FIG. 3 is a schematic diagram of feature set vectors written into an enterprise feature library.
Fig. 4 is a schematic diagram of a matching rule.
Detailed Description
The technical scheme provided by the present invention will be described in detail with reference to the following specific examples, and it should be understood that the following specific examples are only for illustrating the present invention and are not intended to limit the scope of the present invention. Additionally, the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer executable instructions, and although a logical order is illustrated in the flowcharts, in some cases the steps illustrated or described may be performed in an order other than that herein.
The invention provides a dynamic self-iterative system of policy matching mechanism, the architecture of which is shown in figure 1, comprising: an event trigger, a policy router, a feature vector generator, a feature matcher, a rule engine, and a recommendation engine. The event trigger is used for triggering subsequent actions according to the trigger strategy; the strategy router is used for receiving the message notification of the event trigger and determining the executed specific strategy; the feature vector generator is used for generating or updating enterprise feature vectors for specific enterprises or enterprises meeting the conditions according to the execution strategy determined by the strategy router, and writing the enterprise feature vectors into the enterprise feature library; the feature matcher is used for traversing and searching in the enterprise feature library according to the sequence of the first sequence, arranging and combining feature sets according to requirements to form feature set vectors, and writing the feature set vectors into the enterprise feature library; the rule engine is used for reading the enterprise feature library, generating a rule logic expression and an expression group according to the feature set generated by the feature matcher, and writing the rule logic expression and the expression group into the rule library; the recommendation engine is used for screening out a qualified enterprise list based on the current policy execution rule expression, adding the current policy into the policy matching result set of each enterprise, and establishing a result set for the enterprise.
Specifically, the event trigger subscribes to various upstream enterprise reporting process events, an event library is maintained in the event trigger, the event library is written when the event occurs, the strategy library is read, and whether the follow-up action is triggered is determined according to a preset event triggering strategy. The enterprise reporting process event refers to that an enterprise reports a policy by itself, fills in related forms and submits enterprise information, and the process state of the policy reporting is dynamically updated until reporting success or reporting failure. The declared process states include "to audit", "under process", "success", "failure", etc. The state conversion of the enterprise reporting policy process drives the updating of the policy matching mechanism in an event form, and along with the increase of the number of the enterprise reporting, the system uses the enterprise number reaching each state stage as a reference, and the 'gradual' and 'stepwise' thinning and narrowing matching algorithm achieves the aim of more accurate matching.
The preset event trigger strategy is illustrated as follows.
For a certain policy (it is sufficient to meet any of the following conditions),
a. triggering the next action when the number of enterprises submitting the declaration (i.e. reaching the state to be audited) reaches a threshold value N1; n1=a preset value;
b. triggering the next action when the proportion of the enterprises which are declared to be accepted (namely, the in-process state is reached) reaches a threshold value N2; n2=the number of businesses in the current policy acceptance/the total number of businesses declared by the current policy;
c. triggering the next action when the enterprise proportion of the success declaration (i.e. the success state is reached) reaches a threshold value N3; n3=current policy reporting success number of enterprises/current policy reporting total number of enterprises;
the strategy router is used for receiving the information notice of the event trigger, reading the strategy library and determining the executed specific strategy according to the number of enterprises or the proportion of enterprises in the states of being declared to be audited, being declared under the process, being declared to be successful and the like. Policy routing includes the following branches:
policy branch a: and recommending a default policy result set to the enterprise by matching, wherein the policy can be selected and ordered according to a plurality of conditions such as release time, viewing quantity, collection quantity, and a texting unit, and the policy result set is globally effective.
Policy branch B: and recommending a manually specified policy result set to the enterprise matching, and manually specifying the ordering of the result set when the policy is selected.
Policy branch C: and recommending a policy result set generated after a series of feature calculation, extraction, rule generation and data screening to the enterprise matching. The specific calculation process will be described below.
The various policies may be enforced by the policy router decisions alone or in combination. For example, for a complex scenario, the core logic for multiple policy combinations is described as follows: when the proportion of enterprises that declare successful (i.e., reach the "successful" state) reaches a threshold of 50%, the policy router receives the event message sent upstream, reads the policy library, and decides to execute policy B and policy C for the event combination. The result set will be generated by the steps of: according to the policy B, a policy result set R1 which is designated and ordered by the operator according to the attributes of the enterprise industry, the region and the like is selected in advance, then according to the policy C, a result set R2 is generated, R1 deduplication and ordering are performed on R2, that is, R2-r1=r2 'operation is performed, and the final result set is generated as { R1, R2'.
The feature vector generator is configured to generate or update an enterprise feature vector for a particular enterprise or an enterprise that meets a condition (e.g., is in a state of reporting "to be audited") based on an execution policy determined by the policy router. The feature vector generator reads the enterprise information base, acquires each attribute information (including inherent basic attribute information, additional supplementary information automatically filled in and the like) of the enterprise, and generates feature vectors according to preset attribute sequences, as shown in fig. 2. And when the enterprise attribute information is illegal or invalid, carrying out emptying processing, and if the quantity of the illegal or invalid attribute information reaches a preset threshold value, such as 30%, of all information quantity of the current enterprise, directly discarding the feature vector of the current enterprise. The feature vectors of a plurality of enterprises finally form a large two-dimensional feature space, and the large two-dimensional feature space is written into an enterprise feature library. The enterprise feature library establishes a preliminary association of policies with enterprise features.
The feature matcher is used for traversing and searching in the enterprise feature library according to the sequence of the first step and the sequence of the second step. For the current column Col-1, technical metadata (such as field type information of character string type, integer type, decimal type, time type and the like) and business metadata information (such as business description information such as enumeration, prefix matching and the like) of the column Col-1 are stored in an enterprise feature library, a feature set of the current column is generated according to the metadata in the feature library, the feature set can be a single value (such as numerical value and text), a set of a plurality of values (such as a plurality of numerical values and a plurality of texts) or a certain range (such as a time range, that is, earlier, later or in a certain time period), and finally the feature set is arranged and combined according to requirements to form a feature set vector and written into the enterprise feature library. For example, column Col-1 is an integer, a numeric value is used for enumeration, all instance values of the current column are traversed according to the metadata information, and if all instance values are 20 (type: single), a feature set { type: single, value:20} is generated; column Col-2 metadata is consistent with column Col-1, all instance values of column Col-2 are traversed, 3 cases (type: multiple) are found, 20,30,40 respectively, and a feature set { type: multiple, value: {20,30,40 }; column Col-3 is a time type (type: range), traverses all instance values, takes minimum and maximum values, and generates a feature set { type: range, value: {2021-01-01,2099-01-01 }. All feature set ranks form feature set vectors. As shown in fig. 3.
The rule engine is used for reading the enterprise feature library, generating a rule logic expression and an expression group according to the feature set generated by the feature matcher, and writing the rule logic expression and the expression group into the rule library. Within a single expression set, all expressions evaluate to true, and a matching rule hit is considered. The generation logic of a single expression is illustrated as follows:
single value matching: enterprise attribute Value = feature set Value;
multi-value matching: the enterprise attribute Value belongs to a feature set, e.g. "30" belongs to the set {10,25,30,50,100};
time range matching: the enterprise attribute Value is within a time range specified by the feature set, such as a date 2021-01-01 between "2020-01-01 and 2099-01-01".
Other matching modes are just like prefix matching, numerical range matching and the like. As shown in fig. 4.
The recommendation engine is used for screening out a qualified enterprise list based on the rule expression of the current policy execution after the rule expression is generated, adding the current policy into a policy matching result set of each enterprise, adjusting the ordering of the policy result set based on the operation policy, setting information recommendation channels and the like, establishing a result set for the enterprise, and writing a recommendation library and a cache for subsequent quick query.
Based on the system, the invention also provides a dynamic self-iteration method of the policy matching mechanism, which comprises the following steps:
step 1, the enterprise self-declares the policy, fills in the related forms and submits enterprise information, and the process state of the policy declaration is dynamically updated to generate enterprise declaration process events.
Step 2, subscribing various upstream enterprise reporting process events by the event trigger, internally maintaining an event library, writing the event library when the event occurs, reading a strategy library, and determining whether to trigger subsequent actions according to a preset event triggering strategy; the predetermined event triggering strategy has been described above.
And step 3, reading a strategy library according to the information notification of the event trigger, and determining the executed specific strategy according to the number of enterprises in each state stage of reporting. The specific possible policies that may be implemented have been described in the paragraph regarding policy routers, which policies may be implemented alone or in combination.
And step 4, generating or updating enterprise feature vectors for a specific enterprise or an enterprise meeting the conditions according to the determined executed strategy, and writing the enterprise feature vectors into an enterprise feature library. Reference may be made in particular to the paragraph relating to the feature vector generator.
And 5, traversing and searching the enterprise feature library according to the sequence of the first step and the sequence of the second step, arranging and combining feature sets according to requirements to form feature set vectors, and writing the feature set vectors into the enterprise feature library. Reference may be made in particular to the paragraph relating to feature matchers.
And 6, reading the enterprise feature library, generating a rule logic expression and an expression group according to the feature set generated by the feature matcher, and writing the rule logic expression and the expression group into the rule library. Reference may be made in particular to paragraphs relating to the rules engine.
And 7, screening out a qualified enterprise list based on the current policy execution rule expression, adding the current policy into a policy matching result set of each enterprise, and establishing a result set for the enterprise. Reference may be made in particular to paragraphs relating to recommendation engines.
It should be noted that the foregoing merely illustrates the technical idea of the present invention and is not intended to limit the scope of the present invention, and that a person skilled in the art may make several improvements and modifications without departing from the principles of the present invention, which fall within the scope of the claims of the present invention.
Claims (7)
1. A dynamic self-iterating system for a policy matching mechanism, comprising: the system comprises an event trigger, a strategy router, a feature vector generator, a feature matcher, a rule engine and a recommendation engine;
the event trigger is used for triggering subsequent actions according to a trigger strategy;
the event trigger subscribes to various enterprise reporting process events, writes into an event library when the events occur, reads a strategy library, and decides whether to trigger subsequent actions according to a preset event triggering strategy;
the enterprise reporting process event refers to the enterprise reporting policy by itself, filling the related forms and submitting enterprise information, the process state of the policy reporting is dynamically updated, and reporting success or reporting failure state is used;
the preset event triggering strategy comprises at least one of the following strategies:
a. triggering the next action when the number of submitted declared enterprises reaches a threshold value N1; n1=a preset value;
b. triggering the next action when the accepted enterprise proportion reaches a threshold value N2; n2=the number of businesses in the current policy acceptance/the total number of businesses declared by the current policy;
c. triggering the next action when the successfully declared enterprise proportion reaches a threshold value N3; n3=current policy reporting success number of enterprises/current policy reporting total number of enterprises;
the strategy router is used for receiving the message notification of the event trigger and determining the executed specific strategy;
the feature vector generator is used for generating or updating enterprise feature vectors for specific enterprises or enterprises meeting the conditions according to the execution strategy determined by the strategy router, and writing the enterprise feature vectors into the enterprise feature library;
the feature matcher is used for traversing and searching in the enterprise feature library according to the sequence of the first sequence and the sequence of the second sequence, arranging and combining feature sets according to requirements to form feature set vectors, and writing the feature set vectors into the enterprise feature library;
the rule engine is used for reading the enterprise feature library, generating a rule logic expression and an expression group according to the feature set generated by the feature matcher, and writing the rule logic expression and the expression group into the rule library;
the recommendation engine is used for screening out a qualified enterprise list based on the current policy execution rule expression, adding the current policy into the policy matching result set of each enterprise, and establishing a result set for the enterprise.
2. The dynamic self-iterating system of a policy matching mechanism of claim 1, wherein the policy router is configured to receive a message notification from an event trigger, read a policy repository, determine a specific policy to execute based on the number of businesses in each state stage of reporting, and wherein the policy router includes the following branches:
policy branch a: recommending a default policy result set to enterprise matching, screening and sorting according to release time, viewing quantity, collection quantity and text unit condition when selecting policies, wherein the policy result set takes effect globally;
policy branch B: recommending a manually specified policy result set to the enterprise matching, and manually specifying the ordering of the result set when the policy is selected;
policy branch C: and recommending a policy result set generated after a series of feature calculation, extraction, rule generation and data screening to the enterprise matching.
3. The dynamic self-iterating system of a policy matching mechanism according to claim 1, wherein a plurality of policies are decided by a policy router to be executed alone or in combination; and configuring the sequence of the recommended result set in the strategy library when the combination is executed, and displaying the result set according to the sequence of the sequence.
4. The dynamic self-iterative system of policy matching mechanism according to claim 1, wherein said feature vector generator reads an enterprise information base, obtains each attribute information of an enterprise, and generates feature vectors according to a preset attribute ordering; and when the attribute information of the enterprise is illegal or invalid, carrying out emptying processing, and if the quantity of the illegal or invalid attribute information reaches a preset threshold value of the quantity of all information of the current enterprise, directly discarding the feature vector of the current enterprise.
5. The dynamic self-iterative system of policy matching mechanism according to claim 1, wherein said feature matcher stores technical metadata and business metadata information of the current column in an enterprise feature library for the current column, and generates a feature set of the current column according to metadata in the feature library, the feature set being a single value, a set of a plurality of values, or a certain range; and finally, arranging and combining the feature sets as required to form feature set vectors and writing the feature set vectors into an enterprise feature library.
6. The dynamic self-iterating system of a policy matching mechanism of claim 1, wherein within a single set of expressions generated by the rules engine, all expressions evaluate to "true" and a matching rule hit is considered; the generation logic of the single expression includes at least one of the following:
single value matching: enterprise attribute Value = feature set Value; multi-value matching: the enterprise attribute Value belongs to the feature set; time range matching: the enterprise attribute Value is within the time range specified by the feature set; prefix matching; the numerical ranges match.
7. A dynamic self-iteration method of a policy matching mechanism, comprising the steps of:
step 1, an enterprise self-declares a policy, fills in a related form and submits enterprise information, and the process state of the policy declaration is dynamically updated to generate an enterprise declaration process event;
step 2, subscribing various upstream enterprise reporting process events by the event trigger, internally maintaining an event library, writing the event library when the event occurs, reading a strategy library, and determining whether to trigger subsequent actions according to a preset event triggering strategy;
the enterprise reporting process event refers to the enterprise reporting policy by itself, filling the related forms and submitting enterprise information, the process state of the policy reporting is dynamically updated, and reporting success or reporting failure state is used;
the preset event triggering strategy comprises at least one of the following strategies:
a. triggering the next action when the number of submitted declared enterprises reaches a threshold value N1; n1=a preset value;
b. triggering the next action when the accepted enterprise proportion reaches a threshold value N2; n2=the number of businesses in the current policy acceptance/the total number of businesses declared by the current policy;
c. triggering the next action when the successfully declared enterprise proportion reaches a threshold value N3; n3=number of successful enterprises for reporting current policy/total number of enterprises for reporting current policy
Step 3, reading a strategy library according to the information notification of the event trigger, and determining the executed specific strategy according to the number of enterprises in each state stage;
step 4, generating or updating enterprise feature vectors for a specific enterprise or an enterprise meeting the conditions according to the determined executing strategy, and writing the enterprise feature vectors into an enterprise feature library;
step 5, traversing and searching in the enterprise feature library according to the sequence of the first step and the sequence of the second step, arranging and combining feature sets according to requirements to form feature set vectors, and writing the feature set vectors into the enterprise feature library;
step 6, reading an enterprise feature library, generating a rule logic expression and an expression group according to a feature set generated by a feature matcher, and writing the rule logic expression and the expression group into the rule library;
and 7, screening out a qualified enterprise list based on the current policy execution rule expression, adding the current policy into a policy matching result set of each enterprise, and establishing a result set for the enterprise.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210107193.5A CN114595252B (en) | 2022-01-28 | 2022-01-28 | Self-iteration method and system of policy matching mechanism |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210107193.5A CN114595252B (en) | 2022-01-28 | 2022-01-28 | Self-iteration method and system of policy matching mechanism |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114595252A CN114595252A (en) | 2022-06-07 |
| CN114595252B true CN114595252B (en) | 2023-11-10 |
Family
ID=81804758
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210107193.5A Active CN114595252B (en) | 2022-01-28 | 2022-01-28 | Self-iteration method and system of policy matching mechanism |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114595252B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109408683A (en) * | 2018-10-31 | 2019-03-01 | 广州高企云信息科技有限公司 | A kind of policy intelligent Matching system and method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7797289B2 (en) * | 2007-09-05 | 2010-09-14 | Oracle International Corporation | Method and apparatus for automatically executing rules in enterprise systems |
-
2022
- 2022-01-28 CN CN202210107193.5A patent/CN114595252B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109408683A (en) * | 2018-10-31 | 2019-03-01 | 广州高企云信息科技有限公司 | A kind of policy intelligent Matching system and method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114595252A (en) | 2022-06-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7676453B2 (en) | Partial query caching | |
| US10002122B2 (en) | Forming knowledge information based on a predetermined threshold of a concept and a predetermined threshold of a target word extracted from a document | |
| US20120047125A1 (en) | Executing a query plan with display of intermediate results | |
| CN112015741A (en) | Method and device for storing massive data in different databases and tables | |
| US8352470B2 (en) | Adaptive aggregation: improving the performance of grouping and duplicate elimination by avoiding unnecessary disk access | |
| CN115422372A (en) | A method and system for building a knowledge map based on software testing | |
| CN111400449A (en) | Regular expression extraction method and device | |
| CN118822705A (en) | Financial artificial intelligence management service platform and financial business processing method | |
| CN113656440A (en) | A kind of database statement optimization method, device and equipment | |
| Vu et al. | FTKHUIM: a fast and efficient method for mining top-K high-utility itemsets | |
| Si et al. | Establishment and improvement of financial decision support system using artificial intelligence and big data | |
| CN114595252B (en) | Self-iteration method and system of policy matching mechanism | |
| Nin et al. | Speed up gradual rule mining from stream data! A b-tree and owa-based approach | |
| US12045215B2 (en) | Machine learning based duplicate invoice detection | |
| CN117194687A (en) | Multimedia recommendation method, device, equipment and storage medium | |
| Ni et al. | ARTAR: Temporal association rule mining algorithm based on attribute reduction | |
| De Waal | Solving the error localization problem by means of vertex generation | |
| CN116185684A (en) | Construction method and application of server fault root cause analysis model | |
| Alsarkhi et al. | Optimizing inverted index blocking for the matrix comparator in linking unstandardized references | |
| Mior | JSONoid: Monoid-based Enrichment for Configurable and Scalable Data-Driven Schema Discovery | |
| CN116910337B (en) | Entity object selection method, query method, device, server and medium | |
| Nunez et al. | Expansion span: Combining fading memory and retrieval in hybrid state space models | |
| CN113987372B (en) | Hot data acquisition method, device and equipment of domain business object model | |
| CN120429414B (en) | A technology development situation awareness system and method | |
| US20250173733A1 (en) | Attention-based deep neural architectures for multi-point response generation in virtual business assistant ai engines |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |