Search

CN-121544414-B - Intelligent financial wind control optimization method and system for trusted multisource data fusion

CN121544414BCN 121544414 BCN121544414 BCN 121544414BCN-121544414-B

Abstract

The invention discloses an intelligent financial wind control optimization method and system for trusted multi-source data fusion, which relate to the technical field of data processing and comprise the steps of establishing operable authority personnel corresponding to a financial data source set and establishing a multidimensional dynamic authority portrait of each operable authority personnel; the method comprises the steps of clustering according to multidimensional dynamic authority figures of each financial data source to obtain multi-cluster financial data source groups, cross-verifying and fusing intra-cluster data sources of each cluster financial data source group to obtain multi-cluster fused data sources and a plurality of data source loss rates corresponding to the multi-cluster fused data sources, inputting the multi-cluster fused data sources and the plurality of data source loss rates into a wind control monitoring module, and performing authority reallocation on operators of abnormal cluster financial data source groups according to the plurality of data source loss rates. The invention solves the technical problem of low reliability of multi-source financial data fusion in the prior art, and achieves the technical effects of improving the reliability of multi-source financial data fusion results and the accuracy of financial wind control.

Inventors

  • WANG XIAOMING

Assignees

  • 迈思诚(大连)信息技术有限公司

Dates

Publication Date
20260508
Application Date
20260119

Claims (8)

  1. 1. An intelligent financial wind control optimization method for trusted multisource data fusion, which is characterized by comprising the following steps: identifying a financial data source set, and establishing an operable authority personnel corresponding to the financial data source set; Establishing a multidimensional dynamic rights portraits of each operable rights personnel, comprising rights level, rights operation type, rights time range and rights data range, aiming at the operable rights personnel corresponding to each financial data source; Clustering according to the multidimensional dynamic rights expression of each financial data source to obtain a multi-cluster financial data source group; cross-verifying and fusing the intra-cluster data sources of each cluster of financial data source group to obtain multi-cluster fused data sources and a plurality of data source loss rates corresponding to the multi-cluster fused data sources; Inputting the multi-cluster fusion data source and the data source loss rates into a wind control monitoring module, and performing authority reallocation on the operators with the operating authorities of the abnormal cluster financial data source group according to the data source loss rates; cross-verifying and fusing the intra-cluster data sources of each cluster of financial data source group to obtain multi-cluster fused data sources, wherein the method comprises the following steps: constructing a business rule base of each financial data source; Pre-screening the intra-cluster data sources of each cluster of financial data source group by utilizing the business rule base to obtain each cluster of screened financial data source group; constructing a pair-by-pair and a group of cross verification pairs for each screened group of financial data source groups, executing data consistency verification on the cross verification pairs pair by utilizing a consistency verification rule base, and taking the data passing the consistency verification as a fusion data source output by each group of financial data source groups; The method for calculating the loss rate of the multiple data sources corresponding to the multi-cluster fusion data source comprises the following steps: extracting an initial data space of each cluster of financial data sources in the multi-cluster financial data sources; Extracting a fusion data space of each cluster of fusion data sources in the multi-cluster fusion data sources; And calculating a plurality of data source loss rates corresponding to the multi-cluster fusion data sources according to the proportion of the fusion data space to the initial data space.
  2. 2. The method of claim 1, wherein the cross-validation pairs are constructed for each group of screened financial data source groups, the method further comprising: Extracting data credibility features of data sources in clusters in each cluster of financial data source group, wherein the data credibility features comprise historical trust points of the data sources in the clusters, recent accuracy of the data sources and trust point network consistency; calculating a data source dominant index according to the data credible characteristics, and extracting the data sources in the identification clusters, which are larger than a preset dominant index, in each group of financial data sources; Determining a first identification intra-cluster data source from the identification intra-cluster data sources; And constructing pairwise cross verification pairs based on the first identification intra-cluster data source and the residual intra-cluster data source, wherein the first identification intra-cluster data source is used for assisting in verifying the residual intra-cluster data source.
  3. 3. The method of claim 2, wherein determining a first intra-identification cluster data source from the intra-identification cluster data sources, the method comprising: Acquiring data storage attribute information of data sources in clusters in each cluster of financial data source group; Calculating a correlation index of the identified cluster data sources and the rest cluster data sources in the cluster data sources according to the data storage attribute information; And determining the data source in the first identification cluster from the data sources in the identification clusters according to the correlation index.
  4. 4. The method of claim 1, wherein data consistency checking is performed pair-wise for the cross-validation pair using a consistency check rule base comprising a numerical logic consistency check rule, a spatiotemporal logic consistency check rule, and a state logic consistency check rule.
  5. 5. The method of claim 1, wherein the operational authority personnel of the abnormal cluster financial data source group are re-authorized according to the plurality of data source loss rates, the method comprising: the abnormal cluster financial data source group is a financial data source group with the data source loss rate larger than a preset loss rate threshold value; and tracing the abnormal data sources in the abnormal cluster financial data source group, acquiring the operable authority personnel corresponding to the first financial data source, and performing authority reallocation on the operable authority personnel.
  6. 6. The method of claim 5, wherein rights are redeployed for the operable rights personnel, the method comprising: Extracting a multidimensional dynamic rights portrait of each operable rights personnel of the operable rights personnel; and calculating the influence contribution degree of each authority dimension in the multi-dimensional dynamic authority portrait to influence the loss rate of the data source, and carrying out authority reallocation according to the influence contribution degree.
  7. 7. The method of claim 6, wherein the permission re-allocation is performed according to the influence contribution degree, wherein the manner of permission re-allocation includes a reduced allocation of any one of a permission level, a permission operation type, a permission time range, and a permission data range.
  8. 8. An intelligent financial wind control optimization system for trusted multisource data fusion, wherein the system is configured to perform an intelligent financial wind control optimization method for trusted multisource data fusion according to any one of claims 1 to 7, the system comprising: The identification module is used for identifying the financial data source set and establishing an operable authority person corresponding to the financial data source set; The portrait creation module is used for creating a multidimensional dynamic rights portrait of each operable rights personnel, comprising a rights level, a rights operation type, a rights time range and a rights data range, aiming at the operable rights personnel corresponding to each financial data source; the clustering module is used for clustering according to the multidimensional dynamic rights portraits of each financial data source to obtain a multi-cluster financial data source group; The fusion module is used for carrying out cross-validation fusion on the intra-cluster data sources of each cluster of financial data source group to obtain a multi-cluster fusion data source and a plurality of data source loss rates corresponding to the multi-cluster fusion data source; The authority re-allocation module is used for inputting the multi-cluster fusion data source and the data source loss rates into the wind control monitoring module, and performing authority re-allocation on the operable authority personnel of the abnormal cluster financial data source group according to the data source loss rates; Further, the system is further configured to implement the following functions: The business rule base of each financial data source is constructed, the business rule base is utilized to pre-screen the intra-cluster data sources of each cluster of financial data source groups to obtain each cluster of screened financial data source groups, each cluster of screened financial data source groups is constructed into a pair-by-pair cross verification pair, the cross verification pair is utilized to execute data consistency verification by the consistency verification rule base, and the data passing through the consistency verification are used as fusion data sources output by each cluster of financial data source groups; Extracting initial data space of each cluster of financial data sources in the multi-cluster financial data sources, extracting fusion data space of each cluster of fusion data sources in the multi-cluster fusion data sources, and calculating a plurality of data source loss rates corresponding to the multi-cluster fusion data sources according to the proportion of the fusion data space to the initial data space.

Description

Intelligent financial wind control optimization method and system for trusted multisource data fusion Technical Field The invention relates to the technical field of data processing, in particular to an intelligent financial wind control optimization method and system for trusted multi-source data fusion. Background In a financial wind control scene, financial data is usually sourced from a plurality of heterogeneous systems and business bodies, and different data sources have differences in the aspects of acquisition modes, business rules, update frequencies, management authorities and the like, so that insufficient data consistency and uneven reliability are easily caused. Because of the lack of effective evaluation and constraint on the credibility of the data sources, the multisource data is often summarized or simply checked directly in the data fusion process, so that abnormal or low-quality data sources are difficult to identify in time, and the accuracy of fusion results is affected. Meanwhile, the operation authority configuration corresponding to the data source is usually relatively static, and cannot be dynamically adjusted according to the data quality change, so that the abnormal data source continuously participates in fusion, and the reliability of integral financial data fusion and the reliability of wind control analysis are reduced. Disclosure of Invention The application provides an intelligent financial wind control optimization method and system for trusted multi-source data fusion, which are used for solving the technical problem of low reliability of multi-source financial data fusion in the prior art. In view of the above problems, the application provides an intelligent financial wind control optimization method and system for trusted multi-source data fusion. In a first aspect of the present application, there is provided an intelligent financial wind control optimization method for trusted multisource data fusion, the method comprising: The method comprises the steps of identifying a financial data source set, establishing an operable authority person corresponding to the financial data source set, establishing a multi-dimensional dynamic authority portrait of each operable authority person, including authority level, authority operation type, authority time range and authority data range, for each operable authority person, clustering according to the multi-dimensional dynamic authority portrait of each financial data source to obtain multi-cluster financial data source groups, cross-verifying and fusing intra-cluster data sources of each cluster financial data source group to obtain multi-cluster fused data sources and a plurality of data source loss rates corresponding to the multi-cluster fused data sources, inputting the multi-cluster fused data sources and the plurality of data source loss rates into a wind control monitoring module, and performing authority reconfiguration on the operable authority person of an abnormal cluster financial data source group according to the plurality of data source loss rates. In a second aspect of the present application, there is provided an intelligent financial wind control optimization system for trusted multisource data fusion, the system comprising: The system comprises a finance data source set, a figure establishing module, a clustering module, a fusion module, a wind control monitoring module and a rights re-allocation module, wherein the finance data source set is used for identifying the finance data source set, establishing operable rights personnel corresponding to the finance data source set, the figure establishing module is used for establishing multidimensional dynamic rights figures of each operable rights personnel, the multidimensional dynamic rights figures comprise rights levels, rights operation types, rights time ranges and rights data ranges, the clustering module is used for clustering according to the multidimensional dynamic rights figures of each finance data source to obtain multi-cluster finance data source groups, the fusion module is used for carrying out cross verification fusion on data sources in each cluster of the finance data source groups to obtain multi-cluster fusion data sources, the rights re-allocation module is used for inputting the multi-cluster fusion data sources and the multi-data source loss rates into the wind control monitoring module, and carrying out rights re-allocation on the operable rights personnel of an abnormal cluster finance data source group according to the multi-cluster data source loss rates. One or more technical schemes provided by the application have at least the following technical effects or advantages: The method comprises the steps of identifying a financial data source set, establishing an operable authority person corresponding to the financial data source set, establishing a multi-dimensional dynamic authority portrait of each operable authority person, includin