Search

CN-115713404-B - Credit evaluation method for construction industry enterprises

CN115713404BCN 115713404 BCN115713404 BCN 115713404BCN-115713404-B

Abstract

A credit evaluation method for building enterprises relates to the field of informatization and is characterized by comprising the following steps of standard modeling, data aggregation, final evaluation data and special evaluation, wherein the standard modeling is used for extracting data in indexes to form a standard model library, the data aggregation is used for carrying out real-time detection on aggregate qualification of evaluation objects and evaluation users passing unified authentication to form final evaluation data, the data aggregation is carried out in the evaluation period, and the special evaluation is used for carrying out process log record, operation result acquisition and persistence storage on the evaluation objects and evaluation users defined by evaluation and taking data sources such as the final evaluation data and index models as parameters to be brought into a calculation engine. The method has the advantages that unified standard index analysis is provided by accessing different construction enterprise qualification credit evaluation standards, and analysis is mainly performed in aspects of type, traceability duration, data source, collection mode and the like. Through index data analysis, a data model of index extraction data is formed, and standard index information is formed by taking information of one type as an example.

Inventors

  • YANG WENBO
  • WANG QIANG
  • LIU QINGHUA
  • LIU YANG
  • MA AOHUI
  • XIA JUNHUI

Assignees

  • 星际空间（天津）科技发展有限公司

Dates

Publication Date
20260508
Application Date
20221121

Claims (7)

  1. 1. The credit evaluation method for the building industry enterprises is characterized by comprising the following steps: Standard modeling, namely extracting data type, duration, source and collection mode characteristics in the index, completing index data extraction and analysis, refining calculation rules of specific indexes according to the index characteristics, establishing index nodes, observation nodes, grading rules, aggregation forms, repeated verification and grading algorithm configuration in the index model, and forming a standard model library; the standard modeling comprises the following steps: The standard index extraction analysis comprises the steps of combing and grouping evaluation standards to form data types, performing traceability duration analysis on valid period keywords in the data types, performing accuracy judgment analysis on data sources to form rating data sources, and collecting the rating data sources according to enterprise reporting data, shared data and data crawling types to form a front-stage reserve for data aggregation; The method comprises the steps of index data extraction analysis, data source analysis, data collection and index data extraction rule formation, wherein the index data extraction analysis comprises the steps of analyzing types in index data to obtain designated index data items, and performing trace-back duration analysis on the designated index data according to real-time data and designated time efficiency analysis; the index calculation rule analysis comprises the steps of obtaining logic extraction of extracted designated index data through an operation formula to form a general calculation rule, and then modularization and subdivision of the calculation of the whole index through grouping, splitting and limiting calculation logic to form a set of configurable grading operation model; the index model configuration comprises the steps of creating index nodes according to an operation formula and a final operation result in a scoring operation model, configuring one or more observation nodes for the index nodes after the creation, defining information items in the existing observation point nodes to complete scoring rule configuration, forming names in scoring rules, specific scoring rule configuration to the lowest score, defining a data drawing form to form scoring data items, checking data content and repetition rate of the same observation point submitted by the same evaluation object, integrating the scoring rules and the data content of the observation point nodes through an association algorithm group, and performing operation to obtain the evaluation result of a single observation node; the standard model library management, namely configuring the scoring rules, generating forms, evaluating and executing rules and repeated checking rules to form an index model and storing the index model into a database; Data aggregation, namely carrying out real-time detection on the aggregation qualification of the evaluation object and the evaluation user passing through unified authentication, automatically providing the aggregation qualification after the detection is passed, and carrying out automatic extraction and interactive aggregation of an evaluation index model to complete construction of an evaluation data source and form final evaluation data; The data aggregation comprises the following steps: providing an authentication mode according to the type of the evaluation object, performing qualification verification authentication according to the built-in data and the identity of the evaluation object, and performing qualification verification authentication through the information of the identity related to the third party authentication data source and the evaluation user; Data aggregation, namely sorting and aggregating the evaluation objects and the evaluation users passing through the unified authentication into an evaluation qualification queue, and automatically extracting and interactively aggregating index data; An evaluation data source is constructed, namely, each information item of the data aggregation form is called from index model configuration, and the information items are processed and removed from interference items, combined with information models and operation information models with the information models stored in a large data warehouse to obtain final evaluation data; Performing special evaluation, namely carrying out calling of final evaluation data and index model data sources as parameters on evaluation objects limited by the evaluation and evaluation users to carry out process log recording, obtaining operation results and persistence storage on the evaluation objects and the evaluation users; the special evaluation comprises the following steps: Setting an evaluation prohibition term according to an evaluation rule, comparing characteristic data of an evaluation object and an evaluation user one by one according to the prohibition term, checking a flow through a restriction term when judging that the prohibition term does not enter the evaluation object and the evaluation user list and checking the flow through the restriction term when judging that the prohibition term is not satisfied; The evaluation calculation comprises the steps of taking a user list of each evaluation object and evaluation data and index models matched with the user list as parameters of a calculation engine to carry out multithread parallel calculation to obtain an operation result and a process log, pushing abnormal log information, and storing the whole process log and the operation result by adopting unstructured distributed storage; the result pushing step of defining one or more pushing tasks, configuring information items, pushing time periods, network environments and appointed encryption/decryption rules contained in the tasks, and pushing the configured pushing tasks; wherein, the result support includes the following steps: The information publicizing step of retrieving a good information model from a big data warehouse and incorporating the good information model into an automatic publicizing flow, and judging to incorporate an evaluation data source when no objection is received in a designated publicizing period; and sharing the result by sharing the evaluation reference information and the evaluation result information to the public database through the data docking authentication interface.
  2. 2. The method for evaluating credit for construction enterprises according to claim 1, wherein the standard index extraction analysis comprises the steps of: Analyzing the data to obtain data types, namely combing the evaluation content and the scoring standard in the evaluation standard to obtain the combed information types, and grouping the related information according to the types to obtain 4 data types including basic information, management performance, good information and bad behavior information; the traceability duration analysis is carried out, namely the validity duration analysis is carried out on the data about timeliness in each data type according to the keywords; The data source analysis, namely judging and analyzing the accuracy of the sources of various data in each evaluation index to obtain an evaluation data source, and storing data collection in the early stage; And (3) data collection mode analysis, namely caching the data subjected to classification collection and format unification treatment by the evaluation data source formed by data source analysis through enterprise reporting data, shared data and data crawling, and storing the data in advance.
  3. 3. The method for evaluating credit for construction enterprises according to claim 1, wherein the analysis of the index calculation rule comprises the steps of: The same group calculation logic is formed by the rule that the calculation logic is regulated to the data in the group after the specified index data is divided into the same group according to the index data extraction rule, and the formula is as follows: Wherein Represents a grouping calculation rule for the evaluation data, Representing specific calculation rules among groups; The index data extraction rule is that one group of data extraction rules and more than one group of data extraction rules in assignment, disassembly, classification and overtime are operated on the appointed index data in the same group of calculation logic according to the actual demands to obtain index data extraction rules and operation results meeting the requirements; The boundary calculation logic is used for obtaining boundary calculation logic and a final operation result which meet the requirements by comparing and analyzing the operation result in the group with the specified boundary data, wherein the formula is as follows: Where MIN represents the boundary minimum, MAX represents the boundary maximum, Representing and; A scoring operation model is built, namely a scoring operation model meeting the requirements is formed by physically matching the final operation result with the obtained general same group of calculation logic, index data extraction rules and boundary calculation logic, and the model contains one or more scoring operation formulas; The section judgment is obtained by formula operation when the section judgment comprises upper limit operation, and the operation formula is as follows: wherein a, b, c represent values taken after comparison, x, y, z represent values of comparison The section judgment is obtained by formula operation when the section judgment comprises lower limit operation, and the operation formula is as follows: wherein a, b, c represent values taken after comparison, x, y, z represent values of comparison When the comparison operation is carried out, the method is obtained through formula operation, and the operation formula is as follows: where x is the value to be compared, b is the value compared, n is the value that the comparison was successful, m is the value that the comparison was unsuccessful,; when exceeding and not enough operation is performed after comparison, the method is obtained through an operation formula, and the operation formula is as follows: Wherein x is a value to be compared, b is a value to be compared, n is a value to be compared successfully, m is a value to be compared failed, Is a comparator, is another case, Representing an add or subtract operation, c represents the radix of the add score, and i represents taking the absolute value.
  4. 4. The method for evaluating credit for construction enterprises according to claim 1, wherein the index model configuration comprises the following steps: Creating index nodes, namely creating the index nodes according to an operation formula and a final operation result in a scoring operation model, wherein the created index nodes comprise names, codes, scores and indexes; The method comprises the steps of configuring one or more than one observation node for the index node after creation, defining information items in the existing observation nodes to complete grading rule configuration, forming names in grading rules, specific grading rule configuration from the lowest score, defining a data drawing form to form grading data items, checking data content and repetition rate of the same observation point submitted by the same evaluation object, and integrating the grading rules and the data content of the observation point nodes through an association algorithm group to obtain evaluation execution rules of the single observation node.
  5. 5. The method for evaluating credit for construction enterprises according to claim 4, wherein the observation point node comprises the steps of: the grading rule configuration comprises the steps of configuring one or more than one observation node for the index node after being established, defining information items in the existing observation point nodes to complete grading rule configuration, and forming specific grading rule configuration from the name in the grading rule to the lowest score; selecting whether to automatically generate a form, generating a customized form according to specified parameters when the automatic form is judged to be generated, and selecting information items participating in evaluation according to index data to perform standardized configuration to generate the form and configured data when the automatic form is judged to be generated; Setting one or more groups of keywords for the data content at the same observation point submitted by the same evaluation object, and carrying out semantic association matching on the keywords and the approximate words and synonyms thereof to obtain a repetition rate supporting data collection and an evaluation result; and grading, namely integrating grading rules of the observation point nodes and data content through the association algorithm group to obtain evaluation execution rules of the single observation node.
  6. 6. The credit evaluation method for construction enterprises according to claim 1, wherein the data aggregation comprises the specific steps of collecting the evaluation objects and evaluation users passing through the unified authentication into an evaluation qualification queue, and automatically extracting and interactively collecting index data; The method comprises the steps of collecting qualification, namely collecting and entering the uniformly authenticated evaluation object and evaluation user into an evaluation qualification queue, and carrying out real-time detection on the evaluation qualification queue; The method comprises the steps of automatically extracting evaluation objects meeting convergence qualification, automatically incorporating the evaluation users into a data extraction queue, extracting evaluation index model data, carrying out authentication and butt joint of data interfaces through a public database, carrying out data comparison on the evaluation objects, the evaluation users and the model data, carrying out analysis and comparison on the difference information when the difference information appears in the data comparison, obtaining the difference information, carrying out the next process when the data comparison result is consistent, carrying out retrieval and matching on the objects of the evaluation object information one by one with the data model in the public database, synchronizing the matching result into a local database, and carrying out timing update on the history version of the local database, wherein the evaluation index model data comprises enterprise data, personnel data, qualification data and performance data; And the auditor carries out negative data model search on the evaluation object and the evaluation user through supervision records, when the relevant negative data model is searched, the negative data of the relevant evaluation object and the evaluation user is extracted from the supervision records and filled into the negative data model, the negative data model is released, the evaluation object and the evaluation user carry out consulting confirmation, and the flow is ended when the relevant data model cannot be searched.
  7. 7. The construction industry enterprise-oriented credit assessment method according to claim 1, wherein the assessment calculation comprises the steps of: acquiring a data source, namely respectively calling final evaluation data, an index model and an evaluation user list of an evaluation object to form a parameter set to participate in operation; The computing engine is used for sequentially grouping the evaluation user lists of the evaluation objects, distributing more than one data processor according to actual demands, carrying out multi-line Cheng Yunsuan mode on the data processors, carrying out asynchronous operation on the data in the sequences according to the evaluation data parameters and the index model parameters in a circulating way, and carrying out real-time monitoring on progress, information, states and anomalies in the asynchronous process to form an overall process log and an operation result; The formula: ; ; ; Wherein score represents the final calculated score value source (i) is the data source, company is the enterprise data, score (i) is the calculation engine, arrow represents the data flow direction, base (i) is the basic information, good (i) is the good information, bad (i) is the bad information, MANAGEMENT (I) is the business information; and (3) persistent storage, namely storing the whole process log and the operation result by adopting an unstructured distributed storage mode.

Description

Credit evaluation method for construction industry enterprises Technical Field The invention relates to the field of informatization, in particular to a credit evaluation method for construction enterprises, which is characterized in that unified standard index analysis is provided by accessing different construction enterprise qualification credit evaluation standards, analysis is mainly carried out from aspects of types, tracing duration, data sources, collection modes and the like to form a data model of index extraction data, standard index information is formed by taking one type of information as an example, and general calculation rules are formed by carrying out logic extraction on the extracted index data, and calculation logics such as grouping, splitting, limiting and the like are used for modularization and subdivision of the calculation of the whole index to form a set of configurable and modularized calculation rule system. Background Along with the development of society, informatization gradually enters various industries, and a set of more perfect management methods are required for the credit evaluation system of each enterprise in the building industry at present to aim at forming, namely, by accessing different building enterprise qualification credit evaluation standards, unified standard index analysis is provided, analysis is mainly carried out from aspects of types, tracing duration, data sources, collection modes and the like, a data model of index extraction data is formed, standard index information is formed by taking one type of information as an example, and general calculation rules are formed by carrying out logic extraction on the calculation aspects of the extracted index data, and calculation logics such as grouping, splitting, limiting and the like modularize and divide the calculation of the whole index, so that a set of configurable and modularized calculation rule system is formed. Disclosure of Invention The embodiment of the invention provides a credit evaluation method for construction enterprises, which provides unified standard index analysis by accessing different construction enterprise qualification credit evaluation standards, and mainly analyzes the type, the traceability time length, the data source, the collection mode and the like. Through index data analysis, a data model of index extraction data is formed, and standard index information is formed by taking information of one type as an example. The calculation of the whole index is modularized and subdivided through calculation logics such as grouping, splitting, limiting and the like, so as to form a set of configurable and modularized calculation rule system. The method comprises the steps of establishing a set of index model configuration based on organization data based on an index model formed by analyzing index rules, dividing the configuration of the index model into scoring rules, data collection forms, repeated verification and scoring algorithms based on a modularized programming technology, realizing configurable programming, enabling subsequent standard modification to be achieved only by adding corresponding modules, gradually realizing a module for automatically optimizing the index model by training self-learning capacity of the existing index analysis through modeling of the existing index analysis by using a deep learning technology, and realizing functions of dynamically generating an input form, globally verifying input data, universally configuring the scoring rules, refining the scoring algorithms and the like based on a modularized thought and combining a single responsibility principle. And providing two data gathering modes of data extraction and autonomous data gathering. And the autonomous data aggregation performs data validity verification in a man-machine interaction mode. The human review provides three levels of review (initial review, validation), with the initial or complex information requiring human intervention for review. The machine audit is carried out by a built-in machine learning module, continuous learning of human audit data and a data model and accumulation of sample data, and a mode of machine audit and manual rechecking is gradually realized in the later period. A self-lapping score calculation engine based on Java pre-lapping. The multi-thread technology realizes the real-time calculation of millions of data, greatly improves the calculation efficiency of scoring by using distributed calculation, realizes the full asynchronization of the calculation process, avoids the blockage of threads, including asynchronous tasks, asynchronous messages, asynchronous exception handling and the like, realizes the breakpoint continuous evaluation function, can realize the breakpoint continuous evaluation function similar to breakpoint continuous transmission when the scoring is broken due to nonresistance factors, and provides traceable whole process data including real-t