CN-122022339-A - Inspection mechanism-oriented review dispatcher scheme generation method, device and medium

CN122022339ACN 122022339 ACN122022339 ACN 122022339ACN-122022339-A

Abstract

The embodiment of the disclosure provides a method, a device and a medium for generating a review dispatcher solution for a checking organization. The method comprises the steps of reading a plurality of key judgment elements and attribute data of all candidate reviewers, screening the attribute data based on preset fairness avoidance conditions to obtain a set of candidate reviewers, constructing a hierarchical data structure for optimizing a dispatcher scheme based on a hierarchical analysis method, calculating weights of all quantized feature parameters in the plurality of quantized feature parameters based on a preset judgment matrix and carrying out consistency check to obtain an effective weight set, generating a plurality of candidate review group schemes according to the set of candidate reviewers, calculating comprehensive evaluation values according to the effective weight set and the corresponding attribute data for each candidate review group scheme, and determining optimal dispatcher decisions in the plurality of candidate review group schemes based on the comprehensive evaluation values. By using the method of the embodiment of the disclosure, the efficiency and the accuracy of the decision of the dispatcher can be obviously improved.

Inventors

LIU JIA
Niu Xingrong
WANG ZHIYUAN
LI ANG
DOU WEIWEI
LI BIN

Assignees

中国合格评定国家认可中心

Dates

Publication Date: 20260512
Application Date: 20260130

Claims (10)

1. A method for generating a review dispatcher scenario for a checking institution, the method comprising: Reading a plurality of key judgment elements and attribute data of all candidate reviewers, wherein the key judgment elements are from a prestored rule base, and the attribute data are from a prestored database; screening the attribute data based on preset fairness avoidance conditions to obtain a set of to-be-selected reviewers; Constructing a hierarchical data structure for optimizing a dispatcher solution based on a hierarchical analysis method, wherein the hierarchical data structure comprises an optimization target configuration, a parameter operation layer and a solution layer, the parameter operation layer comprises a plurality of quantitative characteristic parameters determined based on the plurality of key judgment elements, and the solution layer is used for accommodating candidate review group solutions; Based on a preset discrimination matrix, the optimization target is used as an optimization target, the weight of each quantized feature parameter in the plurality of quantized feature parameters is calculated, and consistency verification is carried out, so that an effective weight set is obtained; responding to the generation of a plurality of candidate review group schemes in the scheme layer according to the set of to-be-selected reviewers, and calculating the comprehensive evaluation value of each candidate review group scheme according to the effective weight set and the corresponding attribute data aiming at each candidate review group scheme; and determining an optimal dispatcher solution from a plurality of candidate review group solutions contained in the solution layer based on the comprehensive evaluation value.
2. The method of claim 1, wherein the screening the attribute data based on the preset fairness avoidance condition to obtain a set of candidate reviewers comprises: reading the preset fairness avoidance conditions, wherein the fairness avoidance conditions comprise benefit association detection conditions, service time window conditions and specific relation identification conditions; matching the attribute data of the candidate reviewers with the fairness avoidance conditions, and identifying candidate reviewer identifiers conforming to any one of the fairness avoidance conditions; And removing all data records corresponding to the identified candidate reviewer identifications from the initial candidate reviewer set, and generating the candidate reviewer set.
3. The method of claim 1, wherein the hierarchically constructing a hierarchical data structure for optimizing a dispatcher solution based on the hierarchical analysis method comprises: establishing a dominant relationship among the optimization target configuration, the parameter operation layer and the scheme layer, wherein the optimization target configuration dominates the parameter operation layer, and the parameter operation layer dominates the scheme layer; and grouping the quantization characteristic parameters according to the attributes of the quantization characteristic parameters to form a hierarchical structure of the parameter operation layer.
4. The method according to claim 1, wherein the calculating weights of each of the plurality of quantization characteristic parameters and performing consistency check based on the preset discrimination matrix with the optimization objective configured as an optimization objective, to obtain an effective weight set, includes: solving the maximum eigenvalue and the corresponding eigenvector of the preset discrimination matrix based on the optimization target configuration, wherein the preset discrimination matrix is pre-constructed based on a Satty scale method; normalizing the feature vector to obtain initial weights of all quantized feature parameters; calculating a consistency index CI according to a consistency index formula CI= (lambda_max-n)/(n-1), wherein lambda_max is the maximum eigenvalue, and n is the order of the preset judging matrix; Querying an average random consistency index RI value corresponding to the order n; calculating a consistency ratio CR according to a consistency ratio formula cr=ci/RI; And in response to the consistency ratio CR being smaller than a preset consistency ratio threshold, determining that the initial weight passes the consistency check, and determining the initial weight as a valid weight set.
5. The method of claim 1, wherein the generating a plurality of candidate panel group scenarios in the scenario layer from the set of candidate panel players comprises: Generating an initial review group combination meeting preset role constraints from the to-be-selected reviewer set based on the resource demand configuration of the current review task, wherein the resource demand configuration comprises the number of reviewer groups, role qualification requirements and professional field coverage requirements; Filtering and screening the initial panel group combination by applying space-time availability constraint and cost constraint, wherein the space-time availability constraint comprises no task conflict in a panel period and the travel distance is within a preset threshold, the cost constraint is set based on estimated travel cost and estimated panel workload, the estimated panel workload is measured by the number of panel days, and the number of panel days is the product of the number of panel members and estimated panel days; and carrying out optimizing calculation on the filtered review group combination through a combination optimizing algorithm to generate a plurality of candidate review group schemes, and loading the candidate review group schemes into the scheme generation layer.
6. The method of claim 1, wherein for each candidate review group plan, calculating a comprehensive evaluation value of the candidate review group plan according to the valid weight set and the attribute data, comprises: determining a quantization score of each reviewer in the candidate review group scheme on each quantization characteristic parameter according to the attribute data; based on the weight of each quantization characteristic parameter in the effective weight set, weighting calculation is carried out on the quantization scores to obtain individual evaluation values corresponding to each reviewer; and integrating the individual evaluation values of all the panelists in the candidate panel scheme, and calculating to obtain the integrated evaluation value of the candidate panel scheme.
7. The method of claim 1, wherein the determining an optimal dispatcher solution from a plurality of candidate review group solutions contained in the solution layer based on the composite evaluation value comprises: screening candidate schemes with comprehensive evaluation values meeting a preset evaluation threshold from the candidate review group schemes; in the case that the number of the screened candidate schemes is one, determining the candidate scheme as an optimal dispatcher decision; And under the condition that the number of the screened candidate schemes is a plurality of, selecting a candidate review group scheme with relatively lowest review cost from the plurality of candidate schemes as an optimal dispatcher decision based on a preset review cost calculation rule.
8. A review dispatcher scenario generation device for a verification institution, the device comprising: The data acquisition unit is configured to read a plurality of key judgment elements and attribute data of all candidate reviewers, wherein the key judgment elements are from a prestored rule base, and the attribute data are from a prestored database; the data screening unit is configured to screen the attribute data based on a preset fairness avoidance condition to obtain a set of to-be-selected reviewers; The modeling unit is configured to construct a hierarchical data structure for optimizing a dispatcher solution based on a hierarchical analysis method, wherein the hierarchical data structure comprises an optimization target configuration, a parameter operation layer and a solution layer, the parameter operation layer comprises a plurality of quantitative characteristic parameters determined based on the plurality of key judgment elements, and the solution layer is used for accommodating candidate review group solutions; the weight calculation unit is configured to calculate the weight of each quantized feature parameter in the plurality of quantized feature parameters and perform consistency verification based on a preset discrimination matrix by taking the optimization target as an optimization target, so as to obtain an effective weight set; An evaluation value determining unit configured to respond to the generation of a plurality of candidate evaluation group schemes in the scheme layer according to the set of candidate evaluation members, and calculate, for each candidate evaluation group scheme, a comprehensive evaluation value of the candidate evaluation group scheme according to the valid weight set and the corresponding attribute data; And a decision determining unit configured to determine an optimal dispatcher solution from a plurality of candidate review group solutions contained in the solution layer based on the comprehensive evaluation value.
9. A computer-readable storage medium storing a computer program for executing the inspection agency-oriented review dispatcher solution generation method of claims 1 to 7.
10. An electronic device, characterized in that the electronic device comprises a processor, a memory for storing executable instructions of the processor, and the processor is used for reading the executable instructions from the memory and executing the instructions to realize the review dispatcher solution generation method facing a checking mechanism according to claims 1-7.

Description

Inspection mechanism-oriented review dispatcher scheme generation method, device and medium Technical Field The disclosure relates to the technical field of management application of information technology, in particular to a method, a device and a medium for generating a review dispatcher proposal for a checking mechanism. Background With the rapid development of the inspection industry, the acceptance of review requirements continues to grow. At present, the industry generally adopts a traditional manual dispatcher mode, and the mode relies on the experience of staff and manually screens reviewers matching basic conditions (such as professional fields and regions). However, the method has the obvious limitations that firstly, the consistency of the dispatcher results is poor due to excessive dependence on subjective experience, secondly, hidden criteria such as historical performance, fairness avoidance, culture requirements and the like of the reviewer lack quantitative means, multi-dimensional accurate matching is difficult to achieve, thirdly, a dispatcher effect feedback mechanism is lacked, and a decision closed loop cannot be formed to achieve continuous optimization. Although decision methods such as neural networks and fuzzy comprehensive evaluation exist, the scene is difficult to adapt. The neural network needs a large number of training samples, and the samples are insufficient due to the special approval review projects, and the weight determination of the fuzzy comprehensive evaluation is strong in subjectivity and lacks a deviation correcting mechanism. Therefore, the prior art cannot meet the requirements of approval review dispatchers on quantitative decisions and small sample adaptability, and an innovative solution capable of systematically integrating multidimensional criteria and realizing scientific quantitative decisions is needed. Disclosure of Invention In order to solve the technical problems, a scheme of the present disclosure is provided. The embodiment of the disclosure provides a method, a device and a medium for generating a review dispatcher solution for a checking mechanism. According to a first aspect of an embodiment of the present disclosure, there is provided a review dispatcher scenario generation method for a verification institution, wherein the method includes: Reading a plurality of key judgment elements and attribute data of all candidate reviewers, wherein the key judgment elements are from a prestored rule base, and the attribute data are from a prestored database; screening the attribute data based on preset fairness avoidance conditions to obtain a set of to-be-selected reviewers; Constructing a hierarchical data structure for optimizing a dispatcher solution based on a hierarchical analysis method, wherein the hierarchical data structure comprises an optimization target configuration, a parameter operation layer and a solution layer, the parameter operation layer comprises a plurality of quantitative characteristic parameters determined based on the plurality of key judgment elements, and the solution layer is used for accommodating candidate review group solutions; Based on a preset discrimination matrix, the optimization target is used as an optimization target, the weight of each quantized feature parameter in the plurality of quantized feature parameters is calculated, and consistency verification is carried out, so that an effective weight set is obtained; responding to the generation of a plurality of candidate review group schemes in the scheme layer according to the set of to-be-selected reviewers, and calculating the comprehensive evaluation value of each candidate review group scheme according to the effective weight set and the corresponding attribute data aiming at each candidate review group scheme; and determining an optimal dispatcher solution from a plurality of candidate review group solutions contained in the solution layer based on the comprehensive evaluation value. According to a second aspect of the embodiments of the present disclosure, there is provided a review dispatcher scenario generation apparatus for a verification institution, wherein the apparatus includes: The data acquisition unit is configured to read a plurality of key judgment elements and attribute data of all candidate reviewers, wherein the key judgment elements are from a prestored rule base, and the attribute data are from a prestored database; the data screening unit is configured to screen the attribute data based on a preset fairness avoidance condition to obtain a set of to-be-selected reviewers; The modeling unit is configured to construct a hierarchical data structure for optimizing a dispatcher solution based on a hierarchical analysis method, wherein the hierarchical data structure comprises an optimization target configuration, a parameter operation layer and a solution layer, the parameter operation layer comprises a plurality of quantitative characteristic