CN-121997919-A - Personnel resume data processing method and system

CN121997919ACN 121997919 ACN121997919 ACN 121997919ACN-121997919-A

Abstract

The embodiment of the invention discloses a personnel resume data processing method and a system, which relate to the technical field of data processing, wherein the method comprises the steps of obtaining an organization architecture file and an unstructured appoint and remove table file of the current personnel; the method comprises the steps of collecting a file analysis table file, collecting a file analysis agent, analyzing the file in the file analysis table, extracting and outputting structured data containing a plurality of preset fields, collecting a plurality of semantic analysis agents, carrying out parallel semantic analysis on the structured data from multiple dimensions in a parallel processing mode, calculating semantic analysis results of the semantic analysis agents according to preconfigured historical data quantization calculation logic through a dynamic quantization rule engine, and generating a historical data quantization value and/or an overall historical data quantization value of each dimension of a current person. The embodiment of the invention can solve the problem that the extraction precision of key information is low because the unstructured stem appoint and remove table document cannot be automatically analyzed and the specialized collaborative semantic analysis capability combined with the organization architecture is lacked in the prior art.

Inventors

CHEN GUANGJUN
YU BING
LI LINGZHI
LIU WEI
WU JINGXIANG

Assignees

中智国研(北京)数字科技有限公司

Dates

Publication Date: 20260508
Application Date: 20251104

Claims (10)

1. A personnel history data processing method, comprising: obtaining an organization architecture file and appoint and remove table files of current personnel, wherein the appoint and remove table files are unstructured documents; invoking a document analysis agent to analyze the appoint and remove table file so as to extract and output structured data containing a plurality of preset fields; Invoking a plurality of semantic analysis agents, and performing semantic analysis on the structured data from multiple dimensions in a parallel processing mode, wherein the semantic analysis comprises the steps of analyzing and matching a work unit name field in the structured data by using the organization architecture file; Calculating semantic analysis results of the plurality of semantic analysis agents according to preconfigured historical data quantization calculation logic through a dynamic quantization rule engine, and generating historical data quantization values of each dimension and/or overall historical data quantization values of the current personnel.
2. The method of claim 1, wherein invoking a document parsing agent to parse the appoint and remove table file comprises: Invoking dynamic analysis service based on a large language model through an application programming interface, carrying out semantic understanding and context reasoning on the appoint and remove table file, and identifying and extracting data of the plurality of preset fields; organizing the extracted data into a JavaScript object notation format conforming to a predefined schema; splitting the structured data in the JavaScript object notation format according to a physical entity, and storing the structured data in different database entities.
3. The method of claim 1, wherein invoking a plurality of semantic analysis agents to semantically analyze the structured data from multiple dimensions by parallel processing comprises: Invoking an organization architecture analysis agent, matching a work unit name field in the structured data to an organization architecture tree in the organization architecture file, and outputting a belonging level and a unique number; Calling a professional striping recognition agent, and outputting one or more professional labels and function types based on post description text fields in the structured data by combining a predefined professional striping dictionary and a large language model; Calling a diversified experience judging agent, and judging whether each section of working experience in a working resume field in the structured data belongs to a special experience defined by a policy or not according to a configuration rule; Calling a basic level experience analysis agent, analyzing the affiliated hierarchy output by the agent in combination with the organization architecture, and judging whether the experience in the working resume field meets the preset basic level experience condition; And invoking qualification and honor analysis agents, and carrying out standardized mapping and grade judgment on professional technical job fields, academic institution fields and reward and punishment condition fields in the structured data.
4. The method of claim 1, wherein the method further comprises: Configuring resume data quantification calculation logic of personnel to be evaluated in each item in a man-machine interaction mode; And determining the semantic analysis agents to be called and/or classification or identification granularity of a semantic analysis model inside the semantic analysis agents according to the configured historical data quantization calculation logic of the current personnel.
5. The method of claim 4, wherein configuring the log data quantization calculation logic of the person to be evaluated in each item in a human-computer interaction manner comprises: Configuring resume data quantification calculation logic for personnel to be evaluated in each project through a visual interface; The configuration content of the quantization calculation logic comprises various indexes used by the personnel record data and logic rules for calculating quantization values of the indexes, wherein the logic rules comprise types of factors corresponding to the various indexes, logic for calculating the quantization values of the indexes based on the factors and parameters required for executing the logic.
6. The method of claim 5, wherein configuring the historical data quantization calculation logic for the personnel to be evaluated in each item comprises; creating a project quantification record table; configuring a corresponding personnel set to be evaluated, and creating a personnel quantitative record list for each item in the item quantitative record list; Configuring a corresponding index set and creating an index quantitative record table for each person to be evaluated in the personnel quantitative record table; Configuring a corresponding measurement standard set and creating a measurement standard quantization record table for each index in the index quantization record table; quantifying each metric in the record table for the metric by configuring a corresponding logic rule; The project quantization record table, the personnel quantization record table, the index quantization record table and the measurement standard quantization record table together form a hierarchical quantization data structure, and the hierarchical quantization data structure is used for storing quantization values of each level, which are calculated and generated by the dynamic quantization rule engine.
7. The method of claim 6, wherein performing the calculation by the dynamic quantization rules engine further comprises: polling the project quantization record table and the personnel quantization record table according to the record data quantization task identification, and monitoring the record data quantization task state; when finding the resume data quantization task of failure, updating the state of the associated index quantization record table and the measurement standard quantization record table as failure, and recording the failure reason; Scheduling and executing the record data quantization tasks to be executed according to the index priority order, converting the logic rules into executable sentences for calculation, and correspondingly updating the states of the quantization record tables of all levels; after the resume data quantization task is completed, the quantized values are summarized and the personnel quantized record state is updated.
8. A personnel history data processing system, comprising: the file acquisition module is used for acquiring an organization architecture file and appoint and remove table files of current personnel, wherein the appoint and remove table files are unstructured documents; The document analysis agent is used for analyzing the appoint and remove table file, extracting and outputting structured data containing a plurality of preset fields; The semantic analysis agents are used for carrying out semantic analysis on the structured data from multiple dimensions in a parallel processing mode, wherein the semantic analysis agents are used for analyzing and matching the name field of the working unit in the structured data by using the organization architecture file; The central coordinator is used for scheduling the concurrent execution of the plurality of semantic analysis agents; and the dynamic quantization rule engine is used for calculating the output of the semantic analysis agents according to preconfigured historical data quantization calculation logic to generate the historical data quantization value of each dimension of the current personnel and/or the overall historical data quantization value.
9. The system of claim 8, wherein the document parsing agent is specifically configured to: Invoking dynamic analysis service based on a large language model through an application programming interface, carrying out semantic understanding and context reasoning on the appoint and remove table file, and identifying and extracting data of the plurality of preset fields; organizing the extracted data into a JavaScript object notation format conforming to a predefined schema; splitting the structured data in the JavaScript object notation format according to a physical entity, and storing the structured data in different database entities.
10. The system of claim 8, wherein the plurality of semantic analysis agents comprises: The organization architecture analysis agent is used for matching the name field of the working unit in the structured data to an organization architecture tree in the organization architecture file and outputting the belonging hierarchy and the unique number; The professional striping recognition agent is used for outputting one or more professional labels and function types based on post description text fields in the structured data by combining a predefined professional striping dictionary and a large language model; The diversified experience judging agent is used for judging whether each section of working experience in the working resume field in the structured data belongs to a special experience defined by a policy or not according to the configuration rule; the basic level experience analysis agent is used for analyzing the hierarchy of the agent output in combination with the organization architecture, and judging whether the experience in the working resume field meets the preset basic level experience condition; and the qualification and honor analysis agent is used for carrying out standardized mapping and grade judgment on professional technical job fields, academic institution fields and reward and punishment situation fields in the structured data.

Description

Personnel resume data processing method and system Technical Field The present invention relates to the field of data processing technologies, and in particular, to a method and a system for processing personnel resume data. Background In the aspect of organizing personnel management, analysis and evaluation of the histories of the trunk are key links of trunk selection and talent echelon construction, and the efficiency and the precision of the analysis and evaluation directly influence the scientificity of personnel decision. At present, the talent record analysis of the main part has remarkable limitation on the technical aspect, and is difficult to meet the demands of personnel departments of organization on efficient and accurate record data processing, and the specific problems are as follows: First, the existing method has low automation degree, and is difficult to process unstructured original documents. The current mainstream history analysis mode still relies on manual review or semi-automatic form entry, and part of human resource information systems support importing data from a structured Excel form and simple scoring based on static rules (such as age, academic, etc.), but generally lack automatic parsing capability for unstructured stem appoint and remove form documents (such as PDF, word format). The method has the advantages that a great deal of manpower is required to manually convert the resume information in the paper or electronic unstructured document into the structured data, the processing efficiency is low (the manual processing of single resume usually takes more than 30 minutes), and the accuracy of the subsequent analysis results is easily affected due to the fact that input errors are introduced by manual operation. Secondly, the prior art scheme has serious shortcomings in semantic understanding and information extraction dimensions. Even if unstructured documents are converted into texts through Optical Character Recognition (OCR) and other technologies, the traditional Natural Language Processing (NLP) method or the analysis technology based on a fixed template is difficult to accurately understand and extract complex semantic information contained in the stem histories, namely, the stem histories cannot be combined with organizational structure information (such as hierarchical relations of a group headquarters, a secondary unit and a tertiary unit) to carry out attribution matching on the names of the working units, and professional lines (such as informatization management, financial management), basic experience, cross-unit chores and other key histories with policy guidance performance cannot be accurately identified, so that the history information is single in extraction dimension, and comprehensive data support cannot be provided for stem compound capability assessment. Disclosure of Invention In view of the above, the embodiment of the invention provides a personnel resume data processing method and system, which are used for solving the problems that the prior art cannot automatically analyze unstructured stem appoint and remove table documents and lacks specialized collaborative semantic analysis capability combined with organization architecture, so that the extraction precision of key information is low. In a first aspect, an embodiment of the present invention provides a method for processing personnel resume data, including: obtaining an organization architecture file and appoint and remove table files of current personnel, wherein the appoint and remove table files are unstructured documents; invoking a document analysis agent to analyze the appoint and remove table file so as to extract and output structured data containing a plurality of preset fields; Invoking a plurality of semantic analysis agents, and performing semantic analysis on the structured data from multiple dimensions in a parallel processing mode, wherein the semantic analysis comprises the steps of analyzing and matching a work unit name field in the structured data by using the organization architecture file; Calculating semantic analysis results of the plurality of semantic analysis agents according to preconfigured historical data quantization calculation logic through a dynamic quantization rule engine, and generating historical data quantization values of each dimension and/or overall historical data quantization values of the current personnel. Further, invoking a document parsing agent to parse the appoint and remove table file, including: Invoking dynamic analysis service based on a large language model through an application programming interface, carrying out semantic understanding and context reasoning on the appoint and remove table file, and identifying and extracting data of the plurality of preset fields; organizing the extracted data into a JavaScript object notation format conforming to a predefined schema; splitting the structured data in the JavaScript object notation format