A self-service interactive platform for skilled talent recruitment

Abstract

The present invention relates to the technical field of talent recruitment platform, specifically a self-service interactive platform for skilled talent recruitment, which is composed of a data capture module, a three-level semantic network model module, a person-job matching module, a skilled talent selection module, a skilled talent big data map generation module, a skilled talent directory generation module, and a skilled talent portrait generation module. By using the three-level semantic network model module, the captured raw data can be analyzed and learned at a deeper level to generate knowledge graphs and person-job portraits. This in-depth data processing provides the platform with a comprehensive understanding of talents and positions, thereby improving the accuracy of matching. The explicit features, potential features, ripple features, and intersection features, these four dimensions conduct a comprehensive analysis of talents, not only limited to the experience and skills on the resume, but also to tap the potential capabilities of talents and the ripple effects that may be caused to the team.

Description

Technical Field

[0001] The present invention relates to the technical field of talent recruitment platforms, and in particular to a self-service interactive platform for skilled talent recruitment. Background Art

[0002] A talent recruitment platform is an online platform designed to help companies and organizations recruit and manage talent more effectively. This platform provides a centralized digital environment that enables companies to post job requirements, screen resumes, communicate and interview candidates, and ultimately select suitable candidates for employment. At the same time, a talent recruitment platform also provides job seekers with a convenient way to browse and apply for jobs.

[0003] In the actual use of talent recruitment platforms, many existing recruitment platforms only capture data from a single or limited data source, resulting in incomplete information and the possibility of missing out on some outstanding talents or important information. The lack of in-depth data analysis models leads to inaccurate matching between talents and positions, and there may be a large number of inappropriate matching results. Most existing platforms rely too much on explicit information on resumes, such as educational background, work experience, etc., while ignoring more in-depth analysis of potential characteristics, ripple characteristics, etc. The data of some recruitment platforms are not updated frequently enough to capture industry changes and new talent information in a timely manner. Many platforms lack effective interactive mechanisms, making communication between recruiters and job seekers a major obstacle. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a self-service interactive platform for skilled talent recruitment.

[0005] In order to achieve the above-mentioned object, the present invention adopts the following technical scheme: a self-service interactive platform for skilled talent recruitment is composed of a data capture module, a three-level semantic network model module, a person-job matching module, a skilled talent selection module, a skilled talent big data map generation module, a skilled talent catalog generation module, and a skilled talent portrait generation module;

[0006] The data capture module captures data from network platforms such as vertical platforms, regional policy release platforms, corporate public data and social media platforms, and names it as raw data;

[0007] The three-level semantic network model module parses, learns, and purifies the original data through the three-level semantic network model to generate a knowledge graph and a person-job portrait;

[0008] The person-job matching module uses a talent in-depth analysis model based on four-dimensional features to match person-job portraits with job requirements and generate person-job matching results. The four-dimensional features include explicit features, latent features, ripple features, and intersection features.

[0009] The skilled talent selection module screens and evaluates skilled talents based on the person-job matching results and generates a skilled talent list;

[0010] The skilled talent big data map generation module generates a skilled talent big data map using the knowledge graph and the person-job matching results;

[0011] The skilled talent catalog generation module generates a classification catalog of skilled talents based on the skilled talent list and names it as a skilled talent catalog;

[0012] The skilled talent portrait generation module generates a detailed portrait of the skilled talent based on the person-job matching results and ripple characteristics, and names it as the skilled talent portrait.

[0013] As a further solution of the present invention: the data capture module includes a data source positioning submodule, a data crawling submodule, a data cleaning submodule, and a data standardization submodule;

[0014] The three-level semantic network model module includes a data parsing submodule, a data learning submodule, and a data purification submodule;

[0015] The person-job matching module includes a feature extraction submodule, a depth analysis submodule, and a matching result generation submodule;

[0016] The skilled talent selection module includes a screening submodule, an evaluation submodule, and a list generation submodule;

[0017] The skilled talent big data map generation module includes a data integration submodule, a map design submodule, and a map generation submodule;

[0018] The skilled talent catalog generation module includes a classification submodule, a catalog design submodule, and a catalog generation submodule;

[0019] The skilled talent portrait generation module includes a portrait design submodule and a portrait generation submodule.

[0020] As a further solution of the present invention: the data source positioning submodule adopts web crawler technology to locate the data source through preset keywords and rules, and determines the network platform for data crawling, and the network platform includes vertical platforms, regional policy release platforms, corporate public data and social media platforms;

[0021] The data crawling submodule uses Python's Scrapy framework to crawl data from the network platform to obtain original data;

[0022] The data cleaning submodule uses Python to call the data cleaning function in the Pandas library to clean the original data, remove invalid and redundant information, and obtain cleaned data;

[0023] The data standardization submodule uses Python to call the data standardization function in the Scikit-learn library to perform standardization processing on the cleaned data to generate standardized data.

[0024] As a further solution of the present invention: the data parsing submodule is responsible for parsing the standardized data, calling the data parsing function in the Numpy library, parsing the data, and obtaining parsing information;

[0025] The data learning submodule calls the deep learning models in the TensorFlow library and the PyTorch library based on the parsed information to learn the data and form a preliminary knowledge graph and person-job portrait;

[0026] The data purification submodule uses a data purification algorithm to purify the preliminary knowledge graph and person-job portraits to generate the final knowledge graph and person-job portraits.

[0027] As a further solution of the present invention: the feature extraction submodule calls the feature extraction function in the Scikit-learn library to extract four-dimensional features from the knowledge graph and the person-job portrait;

[0028] The deep analysis submodule uses a deep learning algorithm to perform a deep analysis on the four-dimensional features to form a preliminary result of person-job matching;

[0029] The matching result generation submodule calls the result generation function in the Pandas library to integrate the preliminary results and generate person-job matching results.

[0030] As a further solution of the present invention: the screening submodule calls the screening function in the Pandas library based on the person-job matching result to screen the skilled talents and obtain the screening result;

[0031] The evaluation submodule calls the evaluation function in the Scikit-learn library to evaluate the skilled talents in the screening results to form an evaluation result;

[0032] The list generation submodule generates a skilled talent list based on the evaluation results using a list generation algorithm.

[0033] As a further solution of the present invention: the data integration submodule uses a data integration algorithm to integrate data based on the knowledge graph and the person-job matching result to obtain a data integration result;

[0034] The map design submodule calls the map design function in the Matplotlib library based on the data integration result to perform map design and design the structure and content of the skilled talent big data map;

[0035] The map generation submodule uses a map generation algorithm to generate a big data map of skilled personnel according to the map design.

[0036] As a further solution of the present invention: the classification submodule calls the classification function in the Scikit-learn library according to the skilled talent list to classify the skilled talents and obtain the classification results;

[0037] The catalog design submodule calls the catalog design function in the Pandas library based on the classification results to design the catalog framework of the skilled talent classification catalog;

[0038] The directory generation submodule calls the directory generation function in the Pandas library to generate a skilled talent directory based on the directory framework and classification results.

[0039] As a further solution of the present invention: the portrait design submodule calls the portrait design function in the Matplotlib library according to the person-job matching result and the ripple characteristics to design a detailed portrait of the skilled talent;

[0040] The portrait generation submodule is based on detailed portraits, uses Django as the backend framework, combines front-end frameworks including Vue.js and React to build user interfaces and interactions, uses Echarts components to draw maps, and integrates with JavaScript frameworks to create charts and talent information visualization.

[0041] Compared with the prior art, the advantages and positive effects of the present invention are:

[0042] In the present invention, by using a three-level semantic network model module, the platform can perform a deeper analysis and learning of the captured raw data, and generate a knowledge graph and a person-job portrait based on massive, high-quality dual-volume data. This in-depth data processing provides the platform with a comprehensive understanding of talents and positions, thereby improving the accuracy of matching. The four dimensions of explicit features, potential features, ripple features, and intersection features conduct a comprehensive analysis of talents, not only limited to the experience and skills on the resume, but also to tap the potential capabilities of talents and the ripple effects that may be caused to the team. The data capture module can obtain data from multiple network platforms, including vertical platforms, policy release platforms, and social media, which means that the platform can obtain more comprehensive and real talent information. The skilled talent selection module can perform more accurate screening based on in-depth person-job matching results, improving the effectiveness of screening. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 The present invention proposes a system main framework diagram of a self-service interactive platform for skilled talent recruitment;

[0044] Figure 2 The present invention proposes a data capture module flow chart for a self-service interactive platform for skilled talent recruitment;

[0045] Figure 3 The present invention proposes a three-level semantic network model module flow chart of a self-service interactive platform for skilled talent recruitment;

[0046] Figure 4 The present invention proposes a flow chart of a person-job matching module of a self-service interactive platform for skilled talent recruitment;

[0047] Figure 5 The present invention proposes a flow chart of a skilled talent selection module of a self-service interactive platform for skilled talent recruitment;

[0048] Figure 6 The present invention proposes a flow chart of a skilled talent big data map generation module of a self-service interactive platform for skilled talent recruitment;

[0049] Figure 7 The present invention proposes a flow chart of a skilled talent catalog generation module of a self-service interactive platform for skilled talent recruitment;

[0050] Figure 8 The present invention proposes a flow chart of a skilled talent portrait generation module of a self-service interactive platform for skilled talent recruitment. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0052] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0053] Embodiment 1

[0054] See also Figure 1 ,A self-service interactive platform for skilled talent recruitment is composed of a data crawling module, a three-level semantic network model module, a person-job matching module, a skilled talent selection module, a skilled talent big data map generation module, a skilled talent catalog generation module, and a skilled talent portrait generation module;

[0055] The data capture module captures data from online platforms such as vertical platforms, regional policy release platforms, corporate public data, and social media platforms, and names it as raw data;

[0056] The three-level semantic network model module parses, learns, and purifies the original data through the three-level semantic network model to generate knowledge graphs and job profiles.

[0057] The person-job matching module uses a talent in-depth analysis model based on four-dimensional features to match person-job profiles with job requirements and generate person-job matching results. The four-dimensional features include explicit features, latent features, ripple features, and intersection features.

[0058] The skilled talent selection module screens and evaluates skilled talents based on the person-job matching results and generates a skilled talent list;

[0059] The skilled talent big data map generation module uses the knowledge graph and person-job matching results to generate a skilled talent big data map;

[0060] The skilled talent catalog generation module generates a classification catalog of skilled talents based on the skilled talent list and names it as the skilled talent catalog;

[0061] The skilled talent portrait generation module generates a detailed portrait of skilled talents based on the person-job matching results and ripple characteristics, and names it as skilled talent portrait.

[0062] First, the data capture module obtains data from various network platforms and provides a comprehensive source of information, making the recruitment process more accurate and comprehensive. Secondly, the three-level semantic network model is used to parse and learn the original data to generate knowledge graphs and person-job portraits, which improves the semantic understanding and matching accuracy of the data. Then, the person-job matching module uses a talent in-depth analysis model based on four-dimensional features for matching, comprehensively considering explicit features, potential features, ripple features, and intersection features to achieve accurate person-job matching results. The skilled talent selection module screens and evaluates according to the matching results, greatly simplifying the recruitment process and the time cost of talent screening. In addition, the skilled talent big data map and directory generation module provides an intuitive way to browse and manage skilled talent resources. Finally, the skilled talent portrait generation module generates detailed skilled talent portraits based on matching results and ripple features to provide employers with more accurate candidate information.

[0063] See also Figure 1 ,The data crawling module includes a data source locating submodule, a data crawling submodule, a data cleaning submodule, and a data standardization submodule;

[0064] The three-level semantic network model module includes data parsing submodule, data learning submodule, and data purification submodule;

[0065] The person-job matching module includes a feature extraction submodule, a depth analysis submodule, and a matching result generation submodule;

[0066] The skilled talent selection module includes a screening submodule, an evaluation submodule, and a list generation submodule;

[0067] The skill talent big data map generation module includes data integration submodule, map design submodule, and map generation submodule;

[0068] The skilled talent catalog generation module includes a classification submodule, a catalog design submodule, and a catalog generation submodule;

[0069] The skilled talent portrait generation module includes a portrait design sub-module and a portrait generation sub-module.

[0070] The data capture module includes a data source positioning submodule, a data crawling submodule, a data cleaning submodule, and a data standardization submodule. First, the data source positioning submodule determines the network platform for data capture, including vertical platforms, regional policy release platforms, corporate public data, and social media platforms. Then, the data crawling submodule will capture data from these platforms to obtain the original data. Next, the data cleaning submodule will clean the original data to remove invalid and redundant information. Finally, the data standardization submodule will standardize the cleaned data to generate standardized data.

[0071] The three-level semantic network model module includes a data analysis submodule, a data learning submodule, and a data purification submodule. First, the data analysis submodule will analyze the standardized data and extract useful information. Then, the data learning submodule will learn this information to form a preliminary knowledge graph and person-job portrait. Finally, the data purification submodule will purify these knowledge graphs and person-job portraits to generate refined knowledge graphs and person-job portraits.

[0072] The person-job matching module includes a feature extraction submodule, a deep analysis submodule, and a matching result generation submodule. The feature extraction submodule extracts four-dimensional features from the refined knowledge graph and person-job portrait, including explicit features, potential features, ripple features, and intersection features. These features reflect the relationship and characteristics between skilled talents and job requirements. The deep analysis submodule performs in-depth analysis on the extracted features and uses deep learning algorithms and other technologies for model training and data analysis to achieve more accurate person-job matching. By analyzing the relationship and mutual influence between features, the deep analysis submodule can dig out deeper information hidden in the data and improve the accuracy and precision of matching. The matching result generation submodule is the person-job matching module that synthesizes the features and data that have been deeply analyzed to generate the final person-job matching results. This process involves algorithms and methods for result processing and integration to ensure the reliability and consistency of matching results.

[0073] Speed ​​and accuracy are two key considerations in the job-person matching process. The self-service interactive nature of the platform requires that the matching process can be completed quickly to provide timely recruitment and talent screening services. At the same time, high-precision matching results can ensure the best match between job requirements and skilled talents, reduce mismatching and waste of resources, and the job-person matching module uses advanced technologies and algorithms, such as methods in the fields of deep learning, data analysis, and model training. These technologies not only improve the speed and accuracy of matching, but also provide more comprehensive job-person matching results, providing a satisfactory solution for recruitment and talent management.

[0074] The skilled talent selection module includes a screening submodule, an evaluation submodule, and a list generation submodule. First, the screening submodule will screen skilled talents based on the person-job matching results. Then, the evaluation submodule will evaluate the screened skilled talents and form an evaluation result. Finally, the list generation submodule will generate a skilled talent list based on the evaluation results.

[0075] The skill talent big data map generation module includes a data integration submodule, a map design submodule, and a map generation submodule. First, the data integration submodule integrates the knowledge graph and the person-job matching results. Then, the map design submodule designs the structure and content of the skill talent big data map. Finally, the map generation submodule generates the skill talent big data map based on the design.

[0076] The skilled talent catalog generation module includes a classification submodule, a catalog design submodule, and a catalog generation submodule. First, the classification submodule will classify skilled talents according to the skilled talent list. Then, the catalog design submodule will design a classification catalog for skilled talents. Finally, the catalog generation submodule will generate a skilled talent catalog based on the design.

[0077] The skilled talent portrait generation module includes a portrait design submodule and a portrait generation submodule. First, the portrait design submodule will design a detailed portrait of the skilled talent based on the person-job matching results and ripple characteristics. Then, the portrait generation submodule will generate a skilled talent portrait based on the design.

[0078] See also Figure 2 ,The data source location submodule uses web crawler technology to locate the data source through ,preset keywords and rules, and determine the network platform for data ,capture. The network platform includes vertical platforms, regional policy ,release platforms, corporate public data and social media platforms;

[0079] The data crawling submodule uses Python's Scrapy framework to crawl data from the network platform and obtain the original data;

[0080] The data cleaning submodule uses Python to call the data cleaning function in the Pandas library to clean the original data, remove invalid and redundant information, and obtain the cleaned data;

[0081] The data standardization submodule uses Python to call the data standardization function in the Scikit-learn library to standardize the cleaned data and generate standardized data.

[0082] First, through the data source positioning submodule, multiple network platforms can be accurately located, including vertical platforms, regional policy release platforms, corporate public data and social media platforms, so as to obtain comprehensive data sources. Secondly, using the data crawling submodule, data can be efficiently captured from these network platforms to ensure the timeliness and accuracy of the data. Then, the data cleaning submodule can remove invalid and redundant information, making the data clean and standardized, and improving the quality of the data. In addition, the data standardization submodule can standardize the cleaned data, eliminate the dimensional differences between different data sources, and improve the comparability of the data.

[0083] See also Figure 3 ,The data parsing submodule is responsible for parsing the standardized data, calling the data parsing function in the Numpy library, parsing the data and obtaining the parsing information;

[0084] The data learning submodule uses the deep learning models in the TensorFlow library and the PyTorch library based on the parsed information to learn the data and form a preliminary knowledge graph and job profile.

[0085] The data purification submodule uses a data purification algorithm to purify the preliminary knowledge graph and person-job portraits to generate the final knowledge graph and person-job portraits.

[0086] First, the data parsing submodule can extract key information from standardized data, providing important input for subsequent data learning and purification. Secondly, the data learning submodule uses the parsed information to call the deep learning model for learning, forming a preliminary knowledge graph and job portrait. Such a learning process can tap deeper knowledge and insights, and improve the accuracy of data analysis and matching. Finally, the data purification submodule uses the data purification algorithm to purify the preliminary knowledge graph and job portrait to generate the final accurate and useful knowledge graph and job portrait. Through the three-level semantic network model, the platform's semantic understanding and data structuring capabilities are improved. They can extract key information from standardized data, learn through deep learning models, form preliminary knowledge graphs and job portraits, and purify these results to generate the final accurate and useful information.

[0087] See also Figure 4 , the feature extraction submodule calls the feature extraction function in the Scikit-learn library to extract four-dimensional features from the knowledge graph and person-job portraits;

[0088] The deep analysis submodule uses deep learning algorithms to conduct in-depth analysis of four-dimensional features to form preliminary results of person-job matching;

[0089] The matching result generation submodule calls the result generation function in the Pandas library, integrates the preliminary results, and generates person-job matching results.

[0090] The feature extraction submodule uses the feature extraction function in the Scikit-learn library to extract key features from the knowledge graph and person-job portraits, providing important input for subsequent matching. The deep analysis submodule uses a deep learning algorithm to perform a deep analysis on the extracted features to obtain preliminary person-job matching results and accurately capture the matching degree between talents and positions. Finally, the matching result generation submodule calls the result generation function in the Pandas library to integrate and filter the preliminary results to generate the final person-job matching results. In summary, the introduction of these submodules, feature extraction, deep analysis, and matching result generation, further enhances the accuracy and efficiency of the person-job matching process. Through feature extraction, key features can be extracted, deep analysis can discover the implicit relationship between features with the help of deep learning algorithms, and matching result generation ensures the visualization and clarity of the final matching results. This will improve the matching accuracy, efficiency, and user experience of the skilled talent recruitment platform, and provide reliable and accurate person-job matching results for recruitment decisions.

[0091] See also Figure 5 The screening submodule calls the screening function in the Pandas library based on the person-job matching results to screen skilled talents and obtain the screening results;

[0092] The evaluation submodule calls the evaluation function in the Scikit-learn library to evaluate the skilled talents in the screening results and form an evaluation result;

[0093] The list generation submodule generates a list of skilled talents based on the evaluation results using a list generation algorithm.

[0094] The screening submodule uses the screening function in the Pandas library to screen skilled talents according to the job matching results and obtain qualified candidates. The evaluation submodule uses the evaluation function in the Scikit-learn library to evaluate the skilled talents in the screening results and form an evaluation result. The list generation submodule generates a list of skilled talents based on the evaluation results using the list generation algorithm. In general, the introduction of these submodules further enhances the accuracy and efficiency of the skilled talent recruitment process. The screening submodule can accurately screen skilled talents according to specific requirements and conditions, the evaluation submodule objectively and accurately evaluates the screening results through the evaluation function, and the list generation submodule generates a clear and visual list of skilled talents.

[0095] See also Figure 6 ,The data integration submodule integrates the data based on the knowledge graph and the ,person-job matching results, using the data integration algorithm to integrate the ,data and obtain the data integration results;

[0096] Based on the data integration results, the map design submodule calls the map design function in the Matplotlib library to perform map design and design the structure and content of the skill talent big data map; uses Django as the back-end framework, combined with front-end frameworks including Vue.js and React, to build user interfaces and interactions, uses Echarts components to draw maps, and integrates with the JavaScript framework to create charts and talent information visualization.

[0097] The map generation submodule uses a map generation algorithm to generate a big data map of skilled personnel based on the map design.

[0098] The data integration submodule uses data integration algorithms based on knowledge graphs and job matching results to integrate information from different data sources and obtain a more comprehensive and consistent data foundation. The map design submodule uses the map design function in the Matplotlib library based on the data integration results to visualize the data into the structure and content of the skilled talent big data map. The map generation submodule uses map generation algorithms based on map design specifications and requirements to generate a skilled talent big data map with good visualization effects.

[0099] See also Figure 7 ,The classification submodule calls the classification function in the Scikit-learn library according to the skilled talent list, classifies the skilled talents, and obtains the classification results;

[0100] The catalog design submodule calls the catalog design function in the Pandas library based on the classification results to design the catalog framework of the skilled talent classification catalog;

[0101] The directory generation submodule calls the directory generation function in the Pandas library to generate a skilled talent directory based on the directory framework and classification results.

[0102] The classification submodule calls the classification function in the Scikit-learn library to accurately classify skilled talents according to the skilled talent list and obtain the classification results. The catalog design submodule uses the catalog design function in the Pandas library to design the hierarchical structure of the skilled talent classification catalog based on the classification results. The catalog generation submodule uses the catalog generation function in the Pandas library to generate a skilled talent catalog based on the catalog framework and classification results. The classification submodule can accurately classify skilled talents, allowing recruitment decision makers to better understand and compare talents of different classifications. The catalog design submodule can design a clear hierarchical catalog based on the classification results, providing users with convenient navigation and browsing methods. The catalog generation submodule can generate a skilled talent catalog based on the catalog framework and classification results, allowing users to quickly find the required talent classification information.

[0103] See also Figure 8 ,The portrait design submodule calls the portrait design function in the Matplotlib library to design a detailed portrait of skilled talents based on the person-job matching results and ripple features;

[0104] The portrait generation submodule is based on detailed portraits, uses Django as the backend framework, combines front-end frameworks including Vue.js and React to build user interfaces and interactions, uses Echarts components to draw maps, and integrates with JavaScript frameworks to create charts and talent information visualization.

[0105] The portrait design submodule calls the portrait design function in the Matplotlib library to design a detailed portrait of skilled talents based on the person-job matching results and ripple characteristics. Through this submodule, the platform can design detailed and rich skilled talent portraits based on the degree of skilled talent's person-job matching and other ripple characteristics. This helps recruitment decision makers to more accurately understand the characteristics of skilled talents, including skill level, experience, educational background, etc., so as to make more targeted recruitment decisions. The portrait generation submodule calls the portrait generation function in the Matplotlib library based on the detailed portrait to design the skilled talent portrait. This submodule can convert the detailed skilled talent portrait into a visual form, making it easier to understand and communicate. Through portrait generation, the platform can provide an intuitive display of skilled talent information, including charts, images, and other visual elements, so that users can quickly grasp the key information of skilled talents.

[0106] Working principle: First, the data capture module is responsible for capturing data from multiple network platforms (such as vertical platforms, regional policy release platforms, corporate public data and social media platforms) and naming it as raw data. The data capture module includes a data source positioning submodule, a data crawling submodule, a data cleaning submodule and a data standardization submodule. The data source positioning submodule uses web crawler technology to locate the data source and determine the network platform for data crawling; the data crawling submodule uses the Scrapy framework to crawl data from the network platform; the data cleaning submodule uses the Pandas library to clean the raw data and remove invalid and redundant information; the data standardization submodule uses the Scikit-learn library to standardize the cleaned data and generate standardized data.

[0107] Next, the three-level semantic network model module parses, learns and purifies the original data through the three-level semantic network model to generate a knowledge graph and a person-job portrait. The three-level semantic network model module includes a data parsing submodule, a data learning submodule and a data purification submodule. The data parsing submodule is responsible for parsing the standardized data and extracting parsing information; the data learning submodule uses a deep learning model (such as TensorFlow) to learn the data based on the parsing information to form a preliminary knowledge graph and a person-job portrait; the data purification submodule uses a purification algorithm to purify the preliminary knowledge graph and the person-job portrait to generate the final knowledge graph and the person-job portrait.

[0108] Then, the person-job matching module uses a talent deep analysis model based on four-dimensional features to match person-job profiles with job requirements and generate person-job matching results. Four-dimensional features include explicit features, latent features, ripple features, and intersection features. The person-job matching module includes a feature extraction submodule, a deep analysis submodule, and a matching result generation submodule. The feature extraction submodule extracts four-dimensional features from the knowledge graph and person-job profiles; the deep analysis submodule uses a deep learning algorithm to analyze the four-dimensional features to obtain preliminary results of person-job matching; the matching result generation submodule integrates the preliminary results to generate the final person-job matching results.

[0109] Next, the skilled talent selection module screens and evaluates skilled talents based on the job matching results and generates a skilled talent list. The skilled talent selection module includes a screening submodule, an evaluation submodule, and a list generation submodule. The screening submodule uses a screening function based on the job matching results to screen skilled talents; the evaluation submodule uses an evaluation function to evaluate the skilled talents in the screening results; and the list generation submodule uses a list generation algorithm based on the evaluation results to generate a skilled talent list.

[0110] Then, the skilled talent big data map generation module generates a skilled talent big data map using the knowledge graph and the person-job matching results. The skilled talent big data map generation module includes a data integration submodule, a map design submodule, and a map generation submodule. The data integration submodule uses a data integration algorithm to integrate data based on the knowledge graph and the person-job matching results; the map design submodule uses map design functions (such as functions in the Matplotlib library) to design the structure and content of the skilled talent big data map; the map generation submodule uses a map generation algorithm to generate a skilled talent big data map based on the map design.

[0111] Next, the skilled talent catalog generation module generates a classification catalog of skilled talents based on the skilled talent list. The skilled talent catalog generation module includes a classification submodule, a catalog design submodule, and a catalog generation submodule. The classification submodule uses classification functions (such as functions in the Scikit-learn library) to classify skilled talents; the catalog design submodule designs the catalog framework of the skilled talent classification catalog based on the classification results; the catalog generation submodule uses catalog generation functions (such as functions in the Pandas library) to generate a skilled talent catalog based on the catalog framework and classification results.

[0112] Finally, the skilled talent portrait generation module generates a detailed portrait of the skilled talent based on the person-job matching results and ripple characteristics. The skilled talent portrait generation module includes a portrait design submodule and a portrait generation submodule. The portrait design submodule calls the portrait design function (such as the function in the Matplotlib library) based on the person-job matching results and ripple characteristics to design a detailed portrait of the skilled talent; the portrait generation submodule calls the portrait generation function (such as the function in the Matplotlib library) based on the detailed portrait to generate a skilled talent portrait.

[0113] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A self-service interactive platform for skilled talent recruitment, characterized by: The self-service interactive platform for skilled talent recruitment is composed of a data capture module, a three-level semantic network model module, a person-job matching module, a skilled talent selection module, a skilled talent big data map generation module, a skilled talent catalog generation module, and a skilled talent portrait generation module; The data capture module captures data from network platforms such as vertical platforms, regional policy release platforms, corporate public data and social media platforms, and names it as raw data; The data capture module includes a data source positioning submodule, a data crawling submodule, a data cleaning submodule, and a data standardization submodule; The data source positioning submodule adopts web crawler technology to locate the data source through preset keywords and rules, and determines the network platform for data crawling, which includes vertical platforms, regional policy release platforms, corporate public data and social media platforms; The data crawling submodule uses Python's Scrapy framework to crawl data from the network platform to obtain original data; The data cleaning submodule uses Python to call the data cleaning function in the Pandas library to clean the original data, remove invalid and redundant information, and obtain cleaned data; The data standardization submodule uses Python to call the data standardization function in the Scikit-learn library to standardize the cleaned data and generate standardized data; The three-level semantic network model module parses, learns, and purifies the original data through the three-level semantic network model to generate a knowledge graph and a person-job portrait; The three-level semantic network model module includes a data parsing submodule, a data learning submodule, and a data purification submodule; The data analysis submodule is responsible for analyzing the standardized data, calling the data analysis function in the Numpy library, analyzing the data, and obtaining analysis information; The data learning submodule calls the deep learning models in the TensorFlow library and the PyTorch library based on the parsed information to learn the data and form a preliminary knowledge graph and person-job portrait; The data purification submodule uses a data purification algorithm to purify the preliminary knowledge graph and person-job portrait to generate a final knowledge graph and person-job portrait; The person-job matching module uses a talent in-depth analysis model based on four-dimensional features to match person-job portraits with job requirements and generate person-job matching results. The four-dimensional features include explicit features, latent features, ripple features, and intersection features. The person-job matching module includes a feature extraction submodule, a depth analysis submodule, and a matching result generation submodule; The feature extraction submodule calls the feature extraction function in the Scikit-learn library to extract four-dimensional features from the knowledge graph and the person-job portrait; The deep analysis submodule uses a deep learning algorithm to perform a deep analysis on the four-dimensional features to form a preliminary result of person-job matching; The matching result generation submodule calls the result generation function in the Pandas library to integrate the preliminary results and generate person-job matching results; The skilled talent selection module screens and evaluates skilled talents based on the person-job matching results and generates a skilled talent list; The skilled talent selection module includes a screening submodule, an evaluation submodule, and a list generation submodule; The screening submodule calls the screening function in the Pandas library based on the person-job matching result to screen the skilled talents and obtain the screening results; The evaluation submodule calls the evaluation function in the Scikit-learn library to evaluate the skilled talents in the screening results to form an evaluation result; The list generation submodule generates a skilled talent list based on the evaluation results using a list generation algorithm; The skilled talent big data map generation module generates a skilled talent big data map using the knowledge graph and the person-job matching results; The skilled talent big data map generation module includes a data integration submodule, a map design submodule, and a map generation submodule; The data integration submodule integrates the data based on the knowledge graph and the person-job matching results using a data integration algorithm to obtain data integration results; The map design submodule calls the map design function in the Matplotlib library based on the data integration result to perform map design and design the structure and content of the skilled talent big data map; The map generation submodule uses a map generation algorithm to generate a skilled talent big data map according to the map design; The skilled talent catalog generation module generates a classification catalog of skilled talents based on the skilled talent list and names it as a skilled talent catalog; The skilled talent catalog generation module includes a classification submodule, a catalog design submodule, and a catalog generation submodule; The classification submodule calls the classification function in the Scikit-learn library according to the skilled talent list to classify the skilled talents and obtain the classification results; The catalog design submodule calls the catalog design function in the Pandas library based on the classification results to design the catalog framework of the skilled talent classification catalog; The directory generation submodule calls the directory generation function in the Pandas library to generate a skilled talent directory based on the directory framework and classification results; The skilled talent portrait generation module generates a detailed portrait of the skilled talent based on the person-job matching result and the ripple characteristics, and names it as a skilled talent portrait; The skilled talent portrait generation module includes a portrait design submodule and a portrait generation submodule; The portrait design submodule calls the portrait design function in the Matplotlib library according to the person-job matching results and ripple characteristics to design a detailed portrait of the skilled talent; The portrait generation submodule is based on detailed portraits, uses Django as the backend framework, combines front-end frameworks including Vue.js and React to build user interfaces and interactions, uses Echarts components to draw maps, and integrates with JavaScript frameworks to create charts and talent information visualization.

Images