Search

CN-121998262-A - Supplier comprehensive management method and system based on real-time AI analysis

CN121998262ACN 121998262 ACN121998262 ACN 121998262ACN-121998262-A

Abstract

The invention provides a provider comprehensive management method and system based on real-time AI analysis, which relate to the technical field of provider comprehensive management and comprise the steps of extracting provider core capability characteristics from standardized provider historical cooperation data based on a depth factor decomposition model; the method comprises the steps of extracting dynamic performance characteristics from provider real-time operation data through a time sequence attention mechanism, carrying out AI comprehensive evaluation on the dynamic performance characteristics based on provider core capability characteristics to obtain provider real-time rating results, carrying out weight adjustment by combining purchasing demand priority to generate a provider hierarchical directory, constructing a provider digital portrait system based on the provider hierarchical directory, carrying out cooperation suitability clustering through a density clustering algorithm to obtain a provider classification cooperation model, carrying out intelligent matching and cooperation strategy optimization on purchasing demands to generate a customized compound operation scheme, and carrying out dynamic iteration update by combining real-time market dynamics to realize the comprehensive management of the full life cycle of the provider.

Inventors

  • LI XIAOCHUN

Assignees

  • 杭州友成科技有限公司

Dates

Publication Date
20260508
Application Date
20260409

Claims (10)

  1. 1. A method for comprehensive supplier management based on real-time AI analysis, the method comprising: performing data standardization processing on the pre-collected provider history cooperative data, and extracting provider core capability features from the standardized provider history cooperative data based on a depth factor decomposition model; Carrying out multidimensional data acquisition and real-time verification on the whole supplier cooperation process based on the multisource real-time data acquisition terminal to obtain supplier real-time operation data, and extracting dynamic performance characteristics from the supplier real-time operation data through a time sequence attention mechanism; Performing AI comprehensive evaluation on the dynamic performance characteristics based on the core capacity characteristics of the suppliers to obtain real-time grading results of the suppliers, and performing weight adjustment on the real-time grading results of the suppliers by combining purchasing demand priorities to generate a grading directory of the suppliers; Constructing a provider digital portrait system based on the provider hierarchical directory, and carrying out cooperation suitability clustering on providers in the portrait system through a density clustering algorithm to obtain a provider classification cooperation model; and carrying out intelligent matching and cooperation strategy optimization on purchasing requirements based on the supplier classification cooperation model to generate a customized compound operation scheme, and carrying out dynamic iterative updating on the customized compound operation scheme by combining with real-time market dynamics to realize comprehensive management of the whole life cycle of the supplier.
  2. 2. The method of claim 1, wherein the process of extracting the vendor core capability features from the normalized vendor history collaboration data based on the depth factor decomposition model comprises: Extracting multi-dimensional historical feature data from standardized provider historical cooperation data, and performing feature discretization and feature normalization on the multi-dimensional historical feature data based on a preset feature dimension system to obtain a provider basic feature set; Extracting cross feature data from the multi-dimensional historical feature data, and performing feature embedding on the cross feature data to obtain a provider cross feature set; Inputting the effective basic feature set and the provider cross feature set into a depth factor decomposition model to obtain a first-order feature interaction vector and a high-order feature interaction vector; calculating the attention weight distribution of the first-order feature interaction vector and the high-order feature interaction vector based on a feature interaction attention layer to obtain feature interaction weights; the first-order feature interaction vector and the high-order feature interaction vector are subjected to weighted fusion based on the feature interaction weight, and a fusion feature vector is obtained; and performing feature dimension reduction on the fusion feature vector to obtain the core capability feature of the provider.
  3. 3. The method for comprehensively managing suppliers based on real-time AI analysis according to claim 2, wherein the process of acquiring real-time operation data of the suppliers by multi-dimensional data acquisition and real-time verification of the cooperation whole flow of the suppliers based on the multi-source real-time data acquisition terminal comprises the following steps: Acquiring order execution data of a supplier based on an order management system interface, adding a time stamp mark to obtain real-time order data, acquiring process quality data and batch sampling inspection data in the production process of a product based on a quality inspection equipment data acquisition interface, adding a time stamp mark to obtain real-time quality data, acquiring cargo transportation data based on a logistics GPS tracking interface, adding a time stamp mark to obtain real-time logistics data, acquiring service interaction data based on an after-sales service platform interface, adding a time stamp mark to obtain real-time service data, acquiring financial interaction data based on a financial settlement system interface, and adding a time stamp mark to obtain real-time financial data; And carrying out time window alignment and data association on the real-time order data, the real-time quality data, the real-time logistics data, the real-time service data and the real-time financial data to obtain real-time multisource fusion data, and carrying out cross check on the real-time multisource fusion data based on a preset check rule to obtain the real-time operation data of the suppliers.
  4. 4. The method of claim 3, wherein the process of extracting dynamic performance characteristics from the provider real-time operation data through a time-series attention mechanism comprises: based on a time sequence attention mechanism, extracting real-time dynamic data from the real-time operation data of the provider, and sequencing the real-time dynamic data according to time sequence to obtain a real-time sequence data sequence; And calculating the attention weight of each time step in the local time sequence feature based on the time sequence attention coding layer to obtain a weighted time sequence feature, and carrying out global average pooling on the weighted time sequence feature to obtain a dynamic performance feature.
  5. 5. The method for comprehensively managing suppliers based on real-time AI analysis according to claim 4, wherein the step of comprehensively evaluating the dynamic performance characteristics based on the core capacity characteristics of the suppliers to obtain the real-time grading results of the suppliers comprises the steps of: The method comprises the steps of carrying out feature stitching on the core capability features of the suppliers and the corresponding dynamic performance features to obtain comprehensive evaluation feature vectors, inputting the comprehensive evaluation feature vectors into a pre-trained comprehensive evaluation model of the suppliers to obtain initial evaluation scores, and grading the initial evaluation scores based on a preset grading threshold to obtain real-time grading results of the suppliers.
  6. 6. The method for vendor integration management based on real-time AI analysis of claim 5, wherein the training process of the vendor integration assessment model comprises: Acquiring a plurality of groups of historical acquisition period historical cooperation data, namely acquiring the core capability characteristics of suppliers and the dynamic performance characteristics of corresponding periods in the historical acquisition period historical cooperation data, and taking the cooperation satisfaction degree as a label value; Grouping and marking the core capability features of the suppliers, the dynamic performance features of the corresponding periods and the tag values in a plurality of groups of historical acquisition period historical cooperation data, and marking as Is a natural number; Will be The group history collects the supplier core capability features and the dynamic performance features of the corresponding period and the tag values in the period history cooperative data as sample data, and Is smaller than The natural number of the sample data is utilized to obtain a sample data average value, and the sample data average value is recorded as a sample set; The method comprises the steps of taking the core capability features of suppliers in the historical cooperation data of the historical acquisition period of the rest groups, the dynamic performance features of the corresponding period and the label values as test sets, and forming a training sample set according to the sample set and the test set; Training the GBDT model and the neural network model according to the training sample set, and marking the model after training as a comprehensive evaluation model of the supplier.
  7. 7. The method for comprehensively managing suppliers based on real-time AI analysis according to claim 6, wherein the process of generating the supplier ranking list includes: Acquiring current purchasing demand information, extracting purchasing demand priority characteristics, and calculating weight coefficients of the purchasing demand priority characteristics based on a hierarchical analysis method; adjusting the initial evaluation score of the real-time grading result of the provider according to the weight coefficient to obtain an adjusted evaluation score; and re-grading based on the adjusted evaluation scores, and sorting according to the grading order and the adjusted scores to generate a provider grading directory.
  8. 8. The method for comprehensively managing suppliers based on real-time AI analysis of claim 7, wherein constructing a digital representation hierarchy of suppliers based on the hierarchical directory of suppliers comprises: extracting the name, the rating level, the evaluation score after adjustment, the core advantage and the adaptation requirement type of each supplier from the supplier rating list to construct a core capability image, acquiring the dynamic performance characteristics of the corresponding suppliers to construct a dynamic performance image, extracting semantic characteristics in a cooperative evaluation text or a complaint record based on a natural language processing technology to construct a cooperative adaptation image; And fusing the core capability image, the dynamic performance image and the cooperation adaptation image of the corresponding suppliers to form a complete supplier digital image of the corresponding suppliers.
  9. 9. The method for comprehensively managing suppliers based on real-time AI analysis according to claim 8, wherein the process of clustering suppliers in the image system for cooperation suitability by a density clustering algorithm to obtain the supplier classification cooperation model comprises the steps of: based on Z-score standardization, carrying out feature standardization on the collaborative adaptation feature set to obtain a standardized adaptation feature set; performing feature analysis on each cluster obtained by clustering to extract cluster core features, wherein the cluster core features comprise a large-batch standardized supply cluster, a small-batch customized supply cluster and an emergency order response cluster; and setting a cooperation scene label and adapting to the purchase demand type for each cluster group to obtain a vendor classification cooperation model.
  10. 10. The provider comprehensive management system based on real-time AI analysis realizes the provider comprehensive management method based on real-time AI analysis as set forth in any one of the claims 1 to 9, and is characterized by comprising a feature extraction module, a real-time acquisition module, a comprehensive evaluation module, a digital portrait module and a dynamic cooperation module; The feature extraction module is used for carrying out data standardization processing on the pre-collected provider history cooperation data and extracting the core capability features of the provider from the standardized provider history cooperation data based on the depth factor decomposition model; The real-time acquisition module is used for carrying out multidimensional data acquisition and real-time verification on the whole process of supplier cooperation based on the multisource real-time data acquisition terminal to obtain supplier real-time operation data, and extracting dynamic performance characteristics from the supplier real-time operation data through a time sequence attention mechanism; The comprehensive evaluation module is used for carrying out AI comprehensive evaluation on the dynamic performance characteristics based on the core capability characteristics of the suppliers to obtain real-time grading results of the suppliers, carrying out weight adjustment on the real-time grading results of the suppliers by combining purchasing demand priority, and generating a grading directory of the suppliers; The digital portrait module constructs a provider digital portrait system based on the provider hierarchical directory, and performs cooperation suitability clustering on providers in the portrait system through a density clustering algorithm to obtain a provider classification cooperation model; And the dynamic cooperation module is used for carrying out intelligent matching and cooperation strategy optimization on purchasing demands based on the supplier classification cooperation model, generating a customized compound operation scheme, and carrying out dynamic iterative updating on the cooperation scheme by combining with real-time market dynamics so as to realize comprehensive management of the whole life cycle of the supplier.

Description

Supplier comprehensive management method and system based on real-time AI analysis Technical Field The invention relates to the technical field of comprehensive supplier management, in particular to a comprehensive supplier management method and system based on real-time AI analysis. Background At present, suppliers manage multi-dependency historical cooperation data to perform static evaluation, and the problems of single evaluation dimension and insufficient real-time performance exist. The core capability and performance of the provider are difficult to be comprehensively captured, and multi-source data (such as order execution, quality detection, logistics service and the like) in the whole cooperation process are scattered and independent, and an effective integration and real-time analysis mechanism is lacked, so that the operation state change of the provider cannot be perceived in time. Meanwhile, the evaluation result is not fully combined with purchasing demand priority and market dynamics to dynamically adjust, so that the matching and cooperation strategies of suppliers lack pertinence, complex and changeable purchasing scenes are difficult to adapt, and the fine management of the whole life cycle of the suppliers cannot be realized, so that purchasing efficiency and cooperation stability are affected. Therefore, a method and a system for comprehensively managing suppliers based on real-time AI analysis are provided. Disclosure of Invention In order to solve the technical problems, the invention aims to provide a provider comprehensive management method and system based on real-time AI analysis. In order to achieve the above purpose, the invention provides a provider integrated management method based on real-time AI analysis, which comprises the following steps: performing data standardization processing on the pre-collected provider history cooperative data, and extracting provider core capability features from the standardized provider history cooperative data based on a depth factor decomposition model; Carrying out multidimensional data acquisition and real-time verification on the whole supplier cooperation process based on the multisource real-time data acquisition terminal to obtain supplier real-time operation data, and extracting dynamic performance characteristics from the supplier real-time operation data through a time sequence attention mechanism; Performing AI comprehensive evaluation on the dynamic performance characteristics based on the core capacity characteristics of the suppliers to obtain real-time grading results of the suppliers, and performing weight adjustment on the real-time grading results of the suppliers by combining purchasing demand priorities to generate a grading directory of the suppliers; Constructing a provider digital portrait system based on the provider hierarchical directory, and carrying out cooperation suitability clustering on providers in the portrait system through a density clustering algorithm to obtain a provider classification cooperation model; and carrying out intelligent matching and cooperation strategy optimization on purchasing requirements based on the supplier classification cooperation model to generate a customized compound operation scheme, and carrying out dynamic iterative updating on the customized compound operation scheme by combining with real-time market dynamics to realize comprehensive management of the whole life cycle of the supplier. Further, the process of extracting vendor core capability features from normalized vendor historical collaboration data based on the depth factoring model includes: Extracting multi-dimensional historical feature data from standardized provider historical cooperation data, and performing feature discretization and feature normalization on the multi-dimensional historical feature data based on a preset feature dimension system to obtain a provider basic feature set; Extracting cross feature data from the multi-dimensional historical feature data, and performing feature embedding on the cross feature data to obtain a provider cross feature set; Inputting the effective basic feature set and the provider cross feature set into a depth factor decomposition model to obtain a first-order feature interaction vector and a high-order feature interaction vector; calculating the attention weight distribution of the first-order feature interaction vector and the high-order feature interaction vector based on a feature interaction attention layer to obtain feature interaction weights; the first-order feature interaction vector and the high-order feature interaction vector are subjected to weighted fusion based on the feature interaction weight, and a fusion feature vector is obtained; and performing feature dimension reduction on the fusion feature vector to obtain the core capability feature of the provider. Further, the process of acquiring the real-time operation data of the suppliers by carrying out mul