CN-121980157-A - Industrial chain quality and efficiency evaluation method and system based on collaborative filtering

CN121980157ACN 121980157 ACN121980157 ACN 121980157ACN-121980157-A

Abstract

The invention relates to the technical field of industry chain collaborative management, and in particular discloses an industry chain quality and efficiency evaluation method and system based on collaborative filtering, wherein the method comprises the steps of obtaining collaborative behavior data of a plurality of entities in an industry chain; extracting a plurality of entity association features according to the collaborative behavior data, acquiring the performance similarity among the entities based on the association features, generating a collaborative performance evaluation matrix according to the performance similarity corresponding to all the entity association features, acquiring historical performance data and real-time performance indexes, and acquiring a real-time dynamic performance evaluation value based on the collaborative performance evaluation matrix. According to the invention, the dynamic evaluation matrix is constructed by integrating the cooperative behavior data among the entities, and the trend analysis is carried out by combining the history and the real-time data, so that the problems of isolation, static state and homogenization of the traditional method are solved.

Inventors

LIU XIAOMING
JIANG JIANPING
ZHENG PEI
JIANG LILIANG
SU WEN
WEI CHUNMEI
LIANG SUMEI

Assignees

浙江省质量科学研究院

Dates

Publication Date: 20260505
Application Date: 20251128

Claims (10)

1. The industrial chain quality and efficiency evaluation method based on collaborative filtering is applied to an industrial chain collaborative management platform, and the platform integrates interaction data of a plurality of industrial chain entities, and is characterized by comprising the following steps: acquiring collaborative behavior data of a plurality of entities in an industrial chain; extracting a plurality of entity association features according to the collaborative behavior data, and acquiring performance similarity among the entities based on the association features; generating a collaborative performance evaluation matrix according to the performance similarity corresponding to all the entity association features; Acquiring historical efficiency data and real-time efficiency indexes, and acquiring a real-time dynamic efficiency evaluation value based on the collaborative efficiency evaluation matrix; Acquiring a historical dynamic efficiency evaluation set, and generating an industrial chain quality efficiency risk evaluation result according to the historical dynamic efficiency evaluation set and the real-time dynamic efficiency evaluation value.
2. The collaborative filtering-based industrial chain quality performance evaluation method according to claim 1, wherein the step of obtaining collaborative behavior data of a plurality of entities in an industrial chain comprises: Acquiring a database of an industrial chain collaborative management platform, acquiring entity interaction records within a preset time range according to the database, and acquiring original interaction data according to the entity interaction records, wherein the original interaction data comprises transaction frequency, number of collaborative projects, resource transfer quantity and quality data sharing times; carrying out data cleaning and normalization processing on the original interaction data, unifying transaction frequency, the number of cooperative projects, the resource transfer amount and the quality data sharing times into a standardized format through a data conversion rule, and generating cleaning data; identifying direct association and indirect association among entities according to the cleaning data, wherein the direct association is calculated by counting co-occurrence frequencies of transactions and cooperative events, and the indirect association is deduced by analyzing a shared resource path and a quality index dependency relationship; Constructing an entity relationship network diagram based on direct association and indirect association, wherein the entity relationship network diagram is generated by using nodes to represent entities and edges to represent collaborative behavior intensity according to association strength and data topological structure of the entity relationship network diagram; and acquiring collaborative behavior data according to the entity relationship network diagram.
3. The method for evaluating quality and performance of an industrial chain based on collaborative filtering according to claim 1, wherein the step of extracting a plurality of entity association features according to the collaborative behavior data and obtaining performance similarities among entities based on the association features comprises: Obtaining a transaction record in the collaborative behavior data, wherein the transaction record comprises transaction frequency and transaction amount fluctuation rate, and obtaining transaction stability characteristics according to the transaction frequency and the transaction amount fluctuation rate; Acquiring cooperative project information in the cooperative behavior data, wherein the cooperative project information comprises project duration and participation depth, and acquiring cooperative depth characteristics according to the project duration and the participation depth; Acquiring a resource exchange log in the collaborative behavior data, wherein the resource exchange log comprises a resource turnover rate and a resource utilization rate, and acquiring a resource utilization efficiency characteristic according to the resource turnover rate and the resource utilization rate; acquiring quality index sharing content in the collaborative behavior data, wherein the quality index sharing content comprises index integrity and timeliness, and acquiring quality index completion characteristics according to the index integrity and timeliness; And carrying out standardization processing on the transaction stability characteristics, the cooperation depth characteristics, the resource utilization efficiency characteristics and the quality index completion degree characteristics which are obtained through calculation, constructing feature vectors based on the standardized feature values, and generating the performance similarity among the entities by adopting a cosine similarity algorithm.
4. The method for evaluating quality and performance of an industrial chain based on collaborative filtering according to claim 1, wherein the step of generating a collaborative performance evaluation matrix according to performance similarities corresponding to all entity association features comprises: generating a performance similarity list among the entities according to the reserved similarity value; constructing an initial collaborative performance evaluation matrix according to a performance similarity list among entities; smoothing the initial cooperative efficiency evaluation matrix to obtain a smoothed matrix; the symmetry check is carried out on the matrix after the smoothing treatment, If the difference value between the check element in the check matrix and the preset element is smaller than the preset symmetry tolerance threshold value, judging that the matrix is symmetrical; If the difference value between the check element in the check matrix and the preset element is larger than the preset symmetry tolerance threshold, the check matrix is determined to be asymmetric; checking the matrix in a non-negative way, checking whether all matrix elements are larger than or equal to zero, and if negative elements exist, judging that the matrix does not meet the non-negative requirement; when the matrix fails the symmetrical check and the non-negative check, acquiring the characteristic weight parameters again according to the adjustment efficiency similarity, recalculating the efficiency similarity among the entities and generating a new evaluation matrix until all the checks are passed; and taking the checked matrix as a final cooperative efficiency evaluation matrix.
5. The collaborative filtering-based industrial chain quality performance evaluation method according to claim 1, wherein the step of obtaining historical performance data and real-time performance metrics and obtaining real-time dynamic performance evaluation values based on the collaborative performance evaluation matrix comprises: Acquiring historical efficiency data, and acquiring efficiency records in a preset period according to the historical efficiency data, wherein the efficiency records comprise entity efficiency scores and timestamp information; acquiring a real-time efficiency index, wherein the real-time efficiency index comprises a resource turnover rate, a cooperation response time and a quality standard reaching rate; based on the collaborative performance evaluation matrix, acquiring weight factors of the entity in historical data and real-time data; According to the weight factors and the real-time efficiency indexes, carrying out trend fitting by adopting a time sequence analysis method to obtain a trend fitting result; and generating a real-time dynamic performance evaluation value based on the trend fitting result.
6. The collaborative filtering-based industrial chain quality performance evaluation method according to claim 1, wherein the step of obtaining a historical dynamic performance evaluation set and generating an industrial chain quality performance risk evaluation result according to the historical dynamic performance evaluation set and a real-time dynamic performance evaluation value comprises: acquiring a historical dynamic performance evaluation set, and acquiring performance evaluation values and related metadata at a plurality of time points according to the historical dynamic performance evaluation set; performing data screening and sorting on the efficacy evaluation values and related metadata at a plurality of time points, and reserving the latest preset number of records to generate a screening historical dynamic efficacy evaluation set; comparing the real-time dynamic performance evaluation value with the screening historical dynamic performance evaluation set to obtain a change trend and a deviation index; generating an industrial chain quality efficacy risk grade based on the variation trend and the deviation index, wherein the risk grade is classified into a low grade, a medium grade and a high grade; If the risk level is medium or high, generating a detailed risk assessment report including risk reasons, influence ranges and improvement suggestions; and if the risk level is low, updating the real-time dynamic performance evaluation value into a historical dynamic performance evaluation set.
7. An industrial chain quality and efficacy evaluation system based on collaborative filtering, comprising: The collaborative behavior data acquisition module is used for acquiring collaborative behavior data of a plurality of entities in the industry chain; the performance similarity acquisition module is used for extracting a plurality of entity association features according to the collaborative behavior data and acquiring performance similarity among the entities based on the association features; the evaluation matrix generation module is used for generating a collaborative performance evaluation matrix according to the performance similarity corresponding to the correlation characteristics of all the entities; the real-time dynamic performance evaluation value acquisition module is used for acquiring historical performance data and real-time performance indexes and acquiring a real-time dynamic performance evaluation value based on the collaborative performance evaluation matrix; the evaluation module is used for acquiring a historical dynamic efficiency evaluation set and generating an industrial chain quality efficiency risk evaluation result according to the historical dynamic efficiency evaluation set and the real-time dynamic efficiency evaluation value.
8. The collaborative filtering-based industrial chain quality performance evaluation system of claim 7, wherein the collaborative behavior data acquisition module comprises: The system comprises an original interaction data acquisition unit, a data processing unit and a data processing unit, wherein the original interaction data acquisition unit is used for acquiring a database of an industrial chain collaborative management platform, acquiring entity interaction records within a preset time range according to the database, and acquiring original interaction data according to the entity interaction records, wherein the original interaction data comprises transaction frequency, number of collaborative projects, resource transfer quantity and quality data sharing times; The cleaning data generation unit is used for carrying out data cleaning and normalization processing on the original interaction data, unifying transaction frequency, the number of cooperative projects, the resource transfer quantity and the quality data sharing times into a standardized format through a data conversion rule, and generating cleaning data; The cleaning data recognition entity-entity recognition unit is used for recognizing direct association and indirect association among the entities according to the cleaning data, wherein the direct association is calculated by counting the co-occurrence frequency of transactions and cooperative events, and the indirect association is deduced by analyzing the dependence relationship between the shared resource path and the quality index; The relationship network diagram generating unit is used for constructing an entity relationship network diagram based on direct association and indirect association, wherein the relationship network diagram with nodes representing the entity and edges representing the collaborative behavior intensity is generated according to the association strength and the data topological structure of the entity relationship network diagram; And the collaborative behavior data acquisition unit is used for acquiring collaborative behavior data according to the entity relationship network diagram.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 6 when the computer program is executed.
10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 6.

Description

Industrial chain quality and efficiency evaluation method and system based on collaborative filtering Technical Field The invention relates to the technical field of industrial chain collaborative management, in particular to an industrial chain quality and efficiency evaluation method and system based on collaborative filtering. Background Along with the continuous and deep collaborative development of the industrial chain, the accurate evaluation of the quality and efficiency of the industrial chain has become a key link for improving the toughness and safety level of the industrial chain. The existing evaluation method often adopts an isolated index system, and cannot effectively capture the synergistic relationship and interaction effect among all entities in the industrial chain. The method is usually based on independent financial indexes or operation data, ignores complex network relations formed by actions such as transaction, cooperation, resource exchange and the like among entities, causes the fragmentation defect of an evaluation result, is difficult to comprehensively reflect the overall efficiency of an industrial chain, and meanwhile, the conventional method is usually based on data snapshot at a fixed time point for evaluation, and cannot effectively track the dynamic change process of the efficiency of the industrial chain. The static evaluation is difficult to identify efficacy fluctuation trend in time, so that early warning is delayed, and prospective guidance cannot be provided for industrial chain risk management. Disclosure of Invention The invention aims to provide an industrial chain quality and efficiency evaluation method and system based on collaborative filtering, which are used for solving the technical problems in the background technology. In order to achieve the above purpose, the present invention provides the following technical solutions: the industrial chain quality and efficiency evaluation method based on collaborative filtering is applied to an industrial chain collaborative management platform, and the platform integrates interaction data of a plurality of industrial chain entities, and comprises the following steps: acquiring collaborative behavior data of a plurality of entities in an industrial chain; extracting a plurality of entity association features according to the collaborative behavior data, and acquiring performance similarity among the entities based on the association features; generating a collaborative performance evaluation matrix according to the performance similarity corresponding to all the entity association features; Acquiring historical efficiency data and real-time efficiency indexes, and acquiring a real-time dynamic efficiency evaluation value based on the collaborative efficiency evaluation matrix; Acquiring a historical dynamic efficiency evaluation set, and generating an industrial chain quality efficiency risk evaluation result according to the historical dynamic efficiency evaluation set and the real-time dynamic efficiency evaluation value. Preferably, the step of obtaining collaborative behavior data of a plurality of entities in an industry chain includes: Acquiring a database of an industrial chain collaborative management platform, acquiring entity interaction records within a preset time range according to the database, and acquiring original interaction data according to the entity interaction records, wherein the original interaction data comprises transaction frequency, number of collaborative projects, resource transfer quantity and quality data sharing times; carrying out data cleaning and normalization processing on the original interaction data, unifying transaction frequency, the number of cooperative projects, the resource transfer amount and the quality data sharing times into a standardized format through a data conversion rule, and generating cleaning data; identifying direct association and indirect association among entities according to the cleaning data, wherein the direct association is calculated by counting co-occurrence frequencies of transactions and cooperative events, and the indirect association is deduced by analyzing a shared resource path and a quality index dependency relationship; Constructing an entity relationship network diagram based on direct association and indirect association, wherein the entity relationship network diagram is generated by using nodes to represent entities and edges to represent collaborative behavior intensity according to association strength and data topological structure of the entity relationship network diagram; and acquiring collaborative behavior data according to the entity relationship network diagram. Preferably, the step of extracting a plurality of entity association features according to the collaborative behavior data and obtaining the performance similarity between the entities based on the association features includes: Obtaining a transaction record in the co