Employment talent supply and demand distribution method and system based on data analysis

Abstract

The invention discloses an employment talent supply and demand distribution method and system based on data analysis, and belongs to the technical field of data analysis, and the method comprises the steps: obtaining talent basic information and post basic information; cleaning and sorting the talent basic information and the post basic information, identifying and deleting repeated data, and processing missing values and abnormal values; key features are extracted from the talent basic information and the post basic information by using a deep learning technology, and the key features are weighted or screened; establishing a matching model between the talent basic information and the post basic information by using a machine learning algorithm; a matching result is analyzed, model optimization is carried out according to the result, parameters or feature weights are adjusted, and the matching accuracy and efficiency are improved; the generalization ability and accuracy of the matching model are evaluated through cross validation and confusion matrix methods; and continuously optimizing the matching model, and adjusting the matching model according to feedback and experience to improve the matching effect. The method has the effect of helping to reasonably configure human resources.

Description

Technical Field

[0001] This application relates to the technical field of data analysis, and particularly to an employment talent supply and demand allocation method and system based on data analysis. Background Art

[0002] In a rapidly changing business environment, talent has become the most valuable resource for enterprises. Under the conditions of a market economy, the allocation of employed talents is mainly achieved through the market mechanism. Workers and employers make two-way choices in the labor market and match according to their respective expectations and needs.

[0003] With the continuous development of big data and artificial intelligence technologies, enterprises increasingly rely on data analysis to accurately locate and obtain the required talents. This data-driven talent strategy not only improves the recruitment efficiency but also helps enterprises maintain their advantages in the highly competitive talent market. At the same time, for job seekers, understanding the dynamic changes in the employment market and talent supply and demand trends is also the key to successful career selection. Therefore, an employment talent supply and demand allocation method based on data analysis has emerged. Summary of the Invention

[0004] In order to adapt to the changing talent market and job requirements, this application provides an employment talent supply and demand allocation method and system based on data analysis.

[0005] The employment talent supply and demand allocation method and system provided by this application adopt the following technical solutions:

[0006] In the first aspect, this application provides an employment talent supply and demand allocation method based on data analysis, including the following steps:

[0007] Obtain basic talent information and basic job information;

[0008] Clean and organize the basic talent information and basic job information, identify and delete duplicate data, and process missing values and outliers;

[0009] Use deep learning technology to extract key features from the basic talent information and basic job information, and weight or screen the key features;

[0010] Use machine learning algorithms to establish a matching model between the basic talent information and the basic job information;

[0011] Analyze the matching results, optimize the model according to the results, adjust the parameters or feature weights, and improve the accuracy and efficiency of the matching;

[0012] Evaluate the generalization ability and accuracy of the matching model through cross-validation and confusion matrix methods;

[0013] Continuously optimize the matching model, and adjust the matching model according to feedback and experience to improve the matching effect.

[0014] Furthermore, in the step of using machine learning algorithms to establish a matching model between the basic information of talents and the basic information of positions, it specifically includes:

[0015] Classify talents and positions according to the basic information of talents and the basic information of positions, and identify the common characteristics on the supply and demand sides;

[0016] Analyze the matching degree between talents and positions through the association rule algorithm to discover potential supply and demand relationships.

[0017] Furthermore, in the step of analyzing the matching degree between talents and positions through the association rule algorithm to discover potential supply and demand relationships, it specifically includes:

[0018] Collect relevant data of talents and positions according to the basic information of talents and the basic information of positions, including educational background, work experience, skills, and salary levels;

[0019] Discretize the data;

[0020] Adopt a method based on keyword or text similarity to evaluate the matching degree between the resume and the job description;

[0021] Use the Apriori algorithm to find high-frequency patterns by calculating support and confidence, and then analyze the correlation between talent characteristics and job requirements;

[0022] Establish a rule set and optimize the matching model according to the rules;

[0023] Further optimize the generalization ability of the model through cross-validation and confusion matrix;

[0024] Analyze the talent demand trends in different industries and positions, and reveal the periodic fluctuations of talent supply and demand through cluster analysis and linear programming models;

[0025] Based on association rules and machine learning algorithms, recommend suitable positions for job seekers and suitable talents for enterprises;

[0026] Combine collaborative filtering algorithms to improve the accuracy and usability of the recommendation system.

[0027] Furthermore, before the step of using the Apriori algorithm to find high-frequency patterns by calculating support and confidence, and then analyze the correlation between talent characteristics and job requirements, it also includes:

[0028] The Apriori algorithm mines association rules by constructing frequent item sets, and it is necessary to pre-determine the optimal thresholds of support and confidence to improve the accuracy of talent and job matching;

[0029] The weighted particle swarm optimization algorithm is used to dynamically adjust the support and confidence thresholds to obtain higher-quality rules;

[0030] The support and confidence thresholds are continuously adjusted through experiments to find the best combination;

[0031] Different data sets and algorithms are used for comparative analysis to verify the effectiveness of the selected thresholds;

[0032] Evaluate the quality of the generated association rules, and the quality of the association rules includes indicators such as support, confidence, and lift;

[0033] Further optimize the thresholds according to the evaluation results to ensure that the generated rules are neither too many nor too few.

[0034] Furthermore, in the step of using a machine learning algorithm to establish a matching model between the basic information of talents and the basic information of jobs, it further includes:

[0035] Obtain the employment quality evaluation target;

[0036] According to the employment quality evaluation target, construct an evaluation system including multiple evaluation indicators. The multiple evaluation indicators are independent and operable and can comprehensively reflect the employment quality;

[0037] Collect data samples related to employment quality and perform dimensionless processing on the data samples to eliminate the dimensional influence between different indicators;

[0038] Select an employment quality state as the reference sequence, and the employment quality state is the optimal value or the mean value;

[0039] Calculate the grey correlation coefficient between each evaluation indicator and the reference sequence. The larger the grey correlation coefficient, the closer the indicator is to the reference sequence, that is, the greater the impact on the employment quality;

[0040] Normalize the grey correlation coefficients of each indicator to obtain the weight values of each indicator, and these weight values reflect the importance of each indicator in the employment quality evaluation;

[0041] Define an optimization objective function to measure the degree of balance between supply and demand in the labor market;

[0042] Use the fuzzy clustering algorithm to calculate the distance between each sample point and the cluster center, so as to determine the membership degree of each sample point;

[0043] Sample points are assigned to different categories according to the membership degree, and the positions of the clustering centers are adjusted through an optimization algorithm to achieve the dynamic balance of the supply and demand in the labor market.

[0044] Furthermore, it also includes:

[0045] Classifying and analyzing the basic information of the talents by using data mining technology;

[0046] Receiving a talent query request sent by an enterprise user through an intelligent terminal, where the talent query request carries talent query information for entering the talent query interface, and the talent query information includes job requirements and talent characteristics;

[0047] Designing a multi-dimensional scoring standard according to the job requirements and talent characteristics, and the scoring standard includes professional skills, work experience, personal qualities, and development potential;

[0048] The enterprise user adjusts the specific scoring rules and weights of each dimension according to its own needs;

[0049] Dynamically adjusting the scoring system to meet the needs of different enterprise users;

[0050] Using a collaborative filtering algorithm to calculate the matching degree score between job seekers and positions, and the higher the score, the higher the matching degree;

[0051] Comprehensively scoring and ranking the talents, forming a talent recommendation list and pushing it to the enterprise user.

[0052] Furthermore, it also includes:

[0053] Receiving a job query request sent by an individual user through an intelligent terminal, where the job query request carries job query information for entering the job query interface, and the job query information includes personal information and job expectation information;

[0054] Generating a career development plan according to the job query information and pushing it to the individual user, and the career development plan includes industry and region recommendations.

[0055] In a second aspect, the present application provides an employment talent supply and demand allocation system based on data analysis, including:

[0056] A basic information acquisition module for acquiring the basic information of talents and the basic information of positions;

[0057] A data processing module for cleaning and sorting the basic information of talents and the basic information of positions, identifying and deleting duplicate data, and processing missing values and outliers;

[0058] A feature extraction module, which is used to extract key features from the basic information of talents and the basic information of positions by using deep learning technology, and weight or screen the key features;

[0059] A matching model establishment module, which is used to establish a matching model between the basic information of talents and the basic information of positions by using machine learning algorithms;

[0060] A model optimization module, which is used to analyze the matching results, optimize the model according to the results, adjust parameters or feature weights, and improve the accuracy and efficiency of matching;

[0061] A model evaluation module, which is used to evaluate the generalization ability and accuracy of the matching model through cross-validation and confusion matrix methods;

[0062] A continuous optimization module, which is used to continuously optimize the matching model, and adjust the matching model according to feedback and experience to improve the matching effect.

[0063] In a third aspect, the present application provides an intelligent terminal, including a memory and a processor, and a computer program capable of being loaded and executed by the processor is stored on the memory, and the computer program is for implementing the above-mentioned employment talent supply and demand allocation method based on data analysis.

[0064] In a fourth aspect, the present application provides a computer-readable storage medium, storing a computer program capable of being loaded and executed by the processor, and the computer program is for implementing the above-mentioned employment talent supply and demand allocation method based on data analysis.

[0065] In summary, compared with the prior art, the beneficial effects of the above technical solutions are as follows:

[0066] The employment talent supply and demand allocation method based on data analysis according to the present application identifies and deletes duplicate data, avoids matching errors caused by data redundancy, processes missing values and outliers, ensures the quality and integrity of data, and lays a solid foundation for subsequent feature extraction and model establishment. Key features are extracted from massive data by using deep learning technology, and these features can more accurately reflect the relevance between talents and positions. The key features are weighted or screened to further improve the effectiveness and pertinence of the features, and more valuable information is provided for the matching model.

[0067] A matching model is established between the basic information of talents and the basic information of positions by using machine learning algorithms, realizing an automated matching process, improving the matching efficiency. The model can perform intelligent matching according to the characteristics of talents and positions, providing a scientific basis for recruitment decisions. By analyzing the matching results, problems and deficiencies existing in the model can be found in time, providing a direction for model optimization, adjusting parameters or feature weights, continuously optimizing the matching model, and improving the accuracy and reliability of matching.

[0068] The generalization ability and accuracy of the matching model are evaluated through methods such as cross - validation and confusion matrix, ensuring the stability and reliability of the model in different scenarios. The matching model is continuously adjusted according to feedback and experience, enabling the model to continuously adapt to new data and scenarios, improving the matching effect and practicality. By combining deep learning and machine learning techniques, accurate matching of basic talent information and basic job information is achieved. Through steps such as data cleaning, feature extraction, model establishment, result analysis, and model optimization, not only the accuracy and efficiency of matching are improved, but also the generalization ability of the model is enhanced, providing strong support for the rational allocation of human resources. Brief Description of the Drawings

[0069] Figure 1 is a schematic flowchart of a method for allocating supply and demand of employed talents based on data analysis according to an embodiment of the present application. Detailed Embodiment

[0070] The following further describes the present application in detail with reference to all the drawings.

[0071] An embodiment of the present application discloses a method and system for allocating supply and demand of employed talents based on data analysis. Refer to Figure 1 , a method for allocating supply and demand of employed talents based on data analysis includes:

[0072] S101. Obtain basic talent information and basic job information.

[0073] Specifically, the allocation system obtains basic talent information (such as education background, work experience, skills, etc.) and basic job information (such as job description, salary range, skill requirements, etc.) from channels such as recruitment platforms, enterprise HR systems, and social media, and uses API interfaces, web crawler technology, or data exchange protocols (such as JSON, XML) for data collection. This ensures the comprehensiveness and diversity of the data, providing a basis for subsequent analysis, and the diversity of data sources helps to improve the breadth and accuracy of matching.

[0074] S102. Clean and organize the basic talent information and basic job information.

[0075] Specifically, the allocation system cleans and organizes the basic information of talents and the basic information of positions, identifies and deletes duplicate data, and processes missing values and outliers. First, data cleaning tools (such as Pandas, OpenRefine) are used for data preprocessing, including removing duplicate data, handling missing values (such as filling, deleting, or interpolating), and handling outliers (such as using the Z-score or IQR method). Then, the text data is standardized (such as unifying the date format, removing special characters, etc.). Thereby improving data quality, reducing the interference of noise on subsequent analysis, ensuring the integrity and consistency of the data, and providing a clean data set for feature extraction and model training.

[0076] S103. Extract key features from the basic information of talents and the basic information of positions using deep learning techniques, and weight or screen the key features.

[0077] Specifically, the allocation system uses deep learning models (such as BERT, LSTM) to extract features from text data (such as resumes, job descriptions), and identifies key skills, experience requirements, etc. The extracted features are weighted or screened, and feature selection algorithms (such as PCA, Lasso regression) are used to remove redundant features. Key features that have a greater impact on the matching result are extracted, reducing the data dimension, improving the model training efficiency, and through weighted processing, highlighting important features and enhancing the accuracy of the matching.

[0078] Convolutional neural networks (CNNs) perform excellently in the field of image processing and are applied in this embodiment to extract features from information such as talent photos and charts in resumes. Through the combination of convolutional layers, pooling layers, and fully connected layers, CNNs can automatically learn hierarchical features in images. Recurrent neural networks (RNNs) have advantages in processing sequential data and are used in this embodiment to process text information such as job descriptions and talent work experience. By capturing dependencies in the sequence, RNNs can extract key features in the text. Autoencoders are an unsupervised learning algorithm that learns the latent representation of data by reconstructing the input data and are used in this embodiment for dimensionality reduction and feature extraction of talent and position information, extracting the most representative features.

[0079] Select appropriate feature weighting and screening methods according to actual needs, specifically including: The Filter method selects features by calculating the correlation or correlation coefficient between features and the target variable. Common metrics include chi-square test, mutual information, Pearson correlation coefficient, etc. This method is simple and fast, but may not be able to capture the interactions between features. The Wrapper method evaluates the importance of features by constructing a model. Common methods include Recursive Feature Elimination (RFE) and genetic algorithm-based feature selection. This method can consider the interactions between features, but has a high computational complexity. The Embedded method combines feature selection with the model training process. Common methods include Lasso and Ridge regression, decision trees, random forests, etc. This method has high computational efficiency and can perform feature selection and model training simultaneously, but may overfit.

[0080] In deep learning tasks, by assigning different weights to different feature labels, the attention of the model to them can be adjusted, which helps to solve the problem of feature imbalance and improve the accuracy and generalization ability of the model. The attention mechanism is an important technique in deep learning. It allows the model to dynamically adjust the attention to different parts when processing input data. By introducing the attention mechanism, different weights can be automatically assigned to different features.

[0081] S104. Use machine learning algorithms to establish a matching model between the basic information of talents and the basic information of positions.

[0082] Specifically, the allocation system selects appropriate machine learning algorithms (such as collaborative filtering, decision trees, random forests, SVM, neural networks, etc.) for model training, takes talent features and position features as inputs, and matching degree as output to train the model. Establishing the matching relationship between talents and positions can recommend suitable positions according to talent features or recommend suitable talents according to position requirements. Through model training, automated matching is initially realized, reducing manual intervention.

[0083] S105. Analyze the matching results and optimize the model according to the results.

[0084] Specifically, the allocation system analyzes the matching results, optimizes the model according to the results, adjusts parameters or feature weights to improve the accuracy and efficiency of matching; analyzes the matching results to identify the reasons for inaccurate matching (such as unreasonable feature weights, inappropriate model parameters, etc.). By adjusting model parameters, feature weights or introducing new features for optimization, the accuracy and efficiency of matching can be improved, reducing the cases of mis-matching and missed-matching. Through continuous optimization, the model can better adapt to different supply and demand scenarios.

[0085] S106. Evaluate the generalization ability and accuracy of the matching model through cross-validation and confusion matrix methods.

[0086] Specifically, the allocation system uses a cross-validation method (such as K-fold cross-validation) to evaluate the generalization ability of the model, ensuring the stability of the model on different datasets. Metrics such as confusion matrix, accuracy, recall, and F1-score are used to evaluate the accuracy of the model, ensuring that the model performs consistently on different datasets, avoiding overfitting or underfitting. Through quantitative evaluation, the matching effect of the model can be intuitively understood, providing a basis for further optimization.

[0087] S107. Continuously optimize the matching model.

[0088] Specifically, the allocation system continuously optimizes the matching model, adjusting the matching model according to feedback and experience to improve the matching effect. According to the evaluation results and feedback in actual applications, continuously adjust the model parameters, feature weights, or introduce new features, using online learning or incremental learning techniques, enabling the model to dynamically adjust according to new data. Through continuous optimization, the matching effect of the model is continuously improved, enabling it to better adapt to market changes. The dynamic adjustment mechanism enables the model to respond in real-time to new supply and demand situations, improving the timeliness and accuracy of matching.

[0089] In another embodiment, S104 specifically includes the following sub-steps:

[0090] S104.11. Classify talents and positions according to the basic talent information and basic position information, and identify the common characteristics on the supply and demand sides.

[0091] Specifically, the allocation system standardizes the basic talent information and basic position information, such as unifying skill names, industry classifications, job categories, etc., and then uses text vectorization techniques (such as TF-IDF, Word2Vec) to convert text data (such as skill descriptions, job requirements) into numerical features, extracting common characteristics of talents and positions, such as skill matching degree, industry matching degree, work experience matching degree, etc. Use clustering algorithms (such as K-Means, hierarchical clustering) to classify talents and positions, grouping similar talents and positions into the same category, and use classification algorithms (such as decision trees, random forests, SVM) to classify talents and positions to identify the common characteristics on the supply and demand sides.

[0092] For example, classify talents into "technical", "management", "sales", etc., and classify positions into "technical positions", "management positions", "sales positions", etc. Through classification, the common characteristics of talents and positions can be quickly identified, providing a basis for subsequent matching. The classification results can help narrow the matching scope and improve the matching efficiency.

[0093] S104.12. Analyze the matching degree between talents and positions through association rule algorithms to discover potential supply and demand relationships.

[0094] Specifically, the allocation system converts the characteristics of talents and positions into a transaction dataset, where each record represents a combination of the characteristics of a talent or a position. For example, the skill set of a talent (Python, SQL, data analysis) or the skill requirements of a position (Python, machine learning, data mining). Association rule algorithms (such as Apriori, FP-Growth) are used to mine the association relationships between talent characteristics and position characteristics.

[0095] By setting the thresholds of support and confidence, strong association rules are filtered out. For example, the rule "Python & data analysis → data scientist position" indicates a strong association between talents with Python and data analysis skills and the data scientist position. The matching degree between talents and positions is calculated according to the association rules. The higher the matching degree, the higher the fit between the talent and the position.

[0096] Weighted matching degrees can be used to adjust the matching degree according to the importance of different characteristics (such as skill weights, experience weights). Through association rule analysis, potential supply and demand relationships can be discovered, and implicit matching patterns between talents and positions can be identified. The calculation of the matching degree provides a quantitative basis for subsequent recommendations and allocations, improving the accuracy and interpretability of the matching.

[0097] In another embodiment, S104.12 specifically includes the following sub-steps:

[0098] S104.12.11. Collect relevant data on talents and positions according to the basic information of talents and positions.

[0099] Specifically, the allocation system collects relevant data on talents and positions according to the basic information of talents and positions, including educational background, work experience, skills, and salary levels; collects data from channels such as recruitment platforms, enterprise HR systems, and social media, including: talent data such as educational background, work experience, skills, salary expectations, and geographical locations; position data such as job descriptions, skill requirements, salary ranges, industry classifications, and work locations. Data collection is carried out using web crawler technology, API interfaces, or data exchange protocols (such as JSON, XML). A comprehensive and diverse dataset is provided to lay a foundation for subsequent analysis, ensuring a wide coverage of data and being able to reflect the real market supply and demand situation.

[0100] S104.12.12. Discretize the data.

[0101] Specifically, the allocation system discretizes continuous data (such as salary level and years of work experience), converting it into categorical data. For example, the salary level is divided into "low (0 - 10K)", "medium (10K - 20K)", and "high (above 20K)". Equal-width binning, equal-frequency binning, or clustering methods are used for discretization. This helps simplify the data, facilitating subsequent classification and association rule analysis, and improving the interpretability and computational efficiency of the model.

[0102] S104.12.13. Adopt a method based on keywords or text similarity to evaluate the matching degree between the resume and the job description.

[0103] Specifically, the allocation system uses natural language processing (NLP) technology to extract keywords (such as skills and job titles) from the resume and the job description. Text similarity algorithms (such as cosine similarity and Jaccard similarity) are used to calculate the matching degree between the resume and the job description. Pretrained word vector models (such as Word2Vec and BERT) are used to capture semantic information, improving the matching accuracy. Furthermore, the matching degree between the resume and the job description is quantified, providing a basis for subsequent matching and improving the accuracy of matching and semantic understanding ability.

[0104] S104.12.14. Use the Apriori algorithm to find high-frequency patterns by calculating support and confidence, and then analyze the correlation between talent characteristics and job requirements.

[0105] Specifically, the allocation system cleans and organizes the original dataset to ensure that the data format is suitable for the Apriori algorithm, including converting text-described characteristics and requirements into quantifiable metrics or codes. According to the actual needs and the characteristics of the dataset, appropriate minimum support and minimum confidence thresholds are set, and these thresholds will be used to filter frequent item sets and strong association rules. The allocation system takes all items in the dataset as candidate item sets and counts the support of each candidate item set. In the analysis of talent characteristics and job requirements, candidate item sets include various combinations of characteristics and requirements. The allocation system filters out frequent item sets by comparing the support of candidate item sets with the minimum support threshold. These frequent item sets represent combinations of talent characteristics and job requirements that occur frequently. Based on the frequent item sets, association rules are generated, and the confidence of each rule is calculated. In the analysis of talent characteristics and job requirements, association rules can reveal the potential relationship between characteristics and requirements. By comparing the confidence of association rules with the minimum confidence threshold, strong association rules are filtered out. These strong association rules provide valuable information about the correlation between talent characteristics and job requirements.

[0106] Convert talent characteristics and job requirements into a transaction dataset, use the Apriori algorithm to mine frequent item sets, and calculate support and confidence. Among them, the frequent item set is a combination of frequently occurring talent characteristics and job requirements. Support represents the probability (frequency) of an item set appearing in a transaction, and can also be understood as the "support level" of a certain item set by customers. In the analysis of talent characteristics and job requirements, support can measure the frequency of occurrence of a certain characteristic or requirement combination. Confidence is used to measure the reliability of an association rule, indicating the probability of the consequent occurring when the antecedent occurs. In the analysis of talent characteristics and job requirements, confidence can reflect the possibility of another characteristic or requirement occurring when a certain characteristic appears. The association rule is in the form of an implication expression like X→Y, where X and Y are disjoint item sets. X can represent talent characteristics, and Y can represent job requirements. For example, the rule "Python & Data Analysis → Data Scientist position" indicates a strong association between talents with Python and data analysis skills and the data scientist position. By setting the minimum support and confidence thresholds, strong association rules are screened out. Discover the potential associations between talent characteristics and job requirements, reveal the supply and demand laws, and provide data support for the construction of the rule set of the matching model.

[0107] S104.12.15. Establish a rule set and optimize the matching model according to the rules.

[0108] Specifically, the allocation system constructs the rule set of the matching model based on the association rules generated by the Apriori algorithm, uses a rule engine (such as Drools) or a machine learning model (such as decision tree, random forest) to implement rule matching, and adjusts the rule weights or introduces new rules according to the matching results to optimize the model. Furthermore, improve the accuracy and interpretability of the matching model. Through rule optimization, the model can better adapt to different supply and demand scenarios.

[0109] S104.12.16. Further optimize the generalization ability of the model through cross-validation and confusion matrix.

[0110] Specifically, the allocation system uses K-fold cross-validation to evaluate the stability of the model to ensure that the model performs consistently on different datasets, uses indicators such as confusion matrix, accuracy, recall rate, and F1-score to evaluate the performance of the model, and adjusts the model parameters or feature weights according to the evaluation results. Furthermore, improve the generalization ability of the model, avoid overfitting or underfitting, and clarify the optimization direction of the model through quantitative evaluation.

[0111] S104.12.17. Analyze the talent demand trends in different industries and positions, and reveal the periodic fluctuations of talent supply and demand through cluster analysis and linear programming models.

[0112] Specifically, the allocation system uses time series analysis or clustering analysis (such as K-Means) to analyze the talent demand trends in different industries and positions, uses a linear programming model to optimize the talent supply and demand allocation, and reveals the periodic fluctuations in supply and demand. Thus, it provides talent demand forecasts for industries and positions, helps enterprises formulate recruitment plans, reveals the laws of supply and demand fluctuations, and provides a basis for policy formulation and market adjustment.

[0113] Adopt clustering algorithms such as K-means and hierarchical clustering to perform clustering analysis on talent data. According to the clustering results, talents are divided into different groups, revealing the characteristics and differences of each group in terms of supply and demand. Interpret the clustering results, clarify the characteristics of each group in terms of education, experience, skills, etc. According to the clustering results, formulate more accurate talent recruitment and training strategies for each industry and position.

[0114] Based on historical data, construct a linear programming model to predict the talent supply and demand in each industry and position for a period of time in the future. The planning model considers the impact of factors such as economic cycles, industrial structure adjustments, and policy changes on talent supply and demand. According to the actual situation, set the parameters of the model, such as the objective function, constraints, etc. Adopt optimization algorithms to adjust and optimize the model parameters to improve the prediction accuracy of the model. Use the linear programming model to predict the talent supply and demand in each industry and position for a period of time in the future, analyze the prediction results, and reveal the periodic fluctuation characteristics of talent supply and demand, providing a scientific basis for formulating talent policies and plans.

[0115] Combining the talent demand trend analysis, clustering analysis results, and linear programming model prediction results, comprehensively analyze the talent supply and demand in industries and positions, revealing the characteristics and differences of each industry and position in terms of talent supply and demand, as well as the laws of periodic fluctuations.

[0116] S104.12.18. Recommend suitable positions for job seekers and suitable talents for enterprises based on association rules and machine learning algorithms.

[0117] Specifically, the allocation system recommends suitable positions for job seekers and suitable talents for enterprises according to association rules and matching models, and uses recommendation algorithms (such as content-based recommendation, collaborative filtering) to improve the accuracy of recommendations. Furthermore, it improves the matching efficiency between job seekers and enterprises, reduces recruitment costs, and enhances the user experience through personalized recommendations.

[0118] Using association rule mining techniques such as the Apriori algorithm and the FP-Growth algorithm, extract the association rules between job seeker characteristics and job requirements from historical recruitment data. These rules involve the correlations between the professional skills, work experience, and educational background of job seekers and specific positions. For the resumes of new job seekers, match their characteristics in the association rule library to find the position with the highest degree of association for recommendation. Similarly, for the job requirements posted by enterprises, the job seeker characteristics that match them can be found in the association rule library to recommend suitable talents.

[0119] Build a job seeker and job matching model using supervised learning algorithms (such as logistic regression, support vector machines, random forests, etc.) or unsupervised learning algorithms (such as clustering algorithms). Use historical recruitment data as the training set to train and optimize the model to improve the accuracy and efficiency of matching. For new job seekers or job requirements, input their characteristics into the trained model, and the model will output a list of matching positions or job seekers. Based on the matching results, recommend suitable positions for job seekers or suitable talents for enterprises.

[0120] Combine the results of association rule mining with the prediction results of machine learning models to improve the accuracy and diversity of recommendations. Use association rules to reveal the potential connections between job seeker characteristics and job requirements, and provide valuable feature inputs for machine learning models.

[0121] S104.12.19. Combine the collaborative filtering algorithm to improve the accuracy and usability of the recommendation system.

[0122] Specifically, the allocation system uses the collaborative filtering algorithm (such as user-item collaborative filtering) to analyze user behavior data (such as the browsing records and application records of job seekers), and combines matrix factorization (such as SVD) or deep learning models (such as neural collaborative filtering) to improve the accuracy of recommendations. Improve the accuracy and personalization of the recommendation system, dynamically adjust the recommendation results through user behavior analysis, and enhance the usability of the system.

[0123] In another embodiment, the following steps are also included before S104.12.14:

[0124] S104.12.21. The Apriori algorithm mines association rules by constructing frequent item sets, and pre-determines the optimal thresholds for support and confidence.

[0125] Specifically, the Apriori algorithm mines association rules by constructing frequent item sets, and it is necessary to pre-determine the optimal thresholds of support and confidence to improve the accuracy of talent and job matching; according to business requirements and data distribution, the threshold ranges of support and confidence are initially set. In this embodiment, the support threshold range is 0.1% - 5%, and the confidence threshold range is 50% - 90%. By using empirical values or domain knowledge as the initial thresholds, initial parameters are provided for subsequent association rule mining, avoiding blind search, and initially screening out rules with a certain degree of universality and reliability.

[0126] S104.12.22. Use the weighted particle swarm optimization algorithm to dynamically adjust the support and confidence thresholds.

[0127] Specifically, the allocation system uses the weighted particle swarm optimization algorithm to dynamically adjust the support and confidence thresholds to obtain higher-quality rules; the support and confidence thresholds are used as optimization variables, and the objective function (such as the number of rules, rule quality) is defined. The weighted particle swarm optimization (WPSO) algorithm is used to search for the optimal combination within the threshold range. The particle swarm optimization (PSO) dynamically adjusts the thresholds by simulating the foraging behavior of bird flocks. The weighting mechanism is used to balance the importance of support and confidence, and iterative optimization is performed until the objective function converges or the maximum number of iterations is reached. Dynamically adjusting the thresholds avoids the cumbersome and subjective manual parameter tuning, improves the scientific nature and adaptability of the thresholds, and obtains higher-quality association rules.

[0128] S104.12.23. Continuously adjust the support and confidence thresholds through experiments.

[0129] Specifically, the allocation system continuously adjusts the support and confidence thresholds through experiments to find the best combination; the allocation system designs experiments, runs the Apriori algorithm under different threshold combinations, and generates association rules. Record the number of rules and rule quality (such as support, confidence, lift) of each experiment, analyze the experimental results, and select the threshold combination with a moderate number of rules and higher quality. Through experimental verification, find the optimal threshold combination, improve the practicality and reliability of the association rules, and avoid too few rules caused by too high thresholds or too many rules and decreased quality caused by too low thresholds.

[0130] S104.12.24. Conduct comparative analysis using different data sets and algorithms.

[0131] Specifically, the allocation system uses different data sets and algorithms for comparative analysis to verify the effectiveness of the selected threshold; runs the Apriori algorithm on multiple data sets to verify the generalization ability of the selected threshold, and compares it with other association rule mining algorithms (such as FP-Growth, Eclat) to analyze the effect of the threshold. Uses statistical methods (such as t-tests) to evaluate the significant differences of different threshold combinations. By verifying the universality and stability of the selected threshold, ensuring its effectiveness under different data sets and algorithms, and through comparative analysis, selects the threshold and algorithm most suitable for the current business scenario.

[0132] S104.12.25, Evaluate the quality of the generated association rules.

[0133] Specifically, the allocation system evaluates the quality of the generated association rules, and the quality of the association rules includes indicators such as support, confidence, and lift; in this embodiment, the following indicators are used to evaluate the quality of the association rules: Support is used to evaluate the frequency of the rule appearing in the data set; Confidence is used to evaluate the credibility of the rule; Lift is used to evaluate the relevance of the rule and measure whether the rule has practical significance. By setting a quality threshold, high-quality rules are screened out. For example, support > 1%, confidence > 70%, and lift > 1.5. Ensure that the generated rules have practical significance and usability, and through quantitative evaluation, screen out the rules most valuable for talent and position matching.

[0134] S104.12.26, Further optimize the threshold according to the evaluation results.

[0135] Specifically, the allocation system further optimizes the threshold according to the evaluation results to ensure that the generated rules are neither too many nor too few. According to the evaluation results of the rule quality, adjust the support and confidence thresholds. If there are too many rules and the quality is low, increase the threshold. If there are too few rules and important rules are missed, lower the threshold. By repeating the optimization process until the number and quality of the generated rules reach a balance, ensure that the generated rules are neither too many nor too few and meet the business requirements. Through continuous optimization, improve the accuracy and usability of the association rules.

[0136] In another embodiment, S104 further includes the following sub-steps:

[0137] S104.21, Obtain the employment quality evaluation target.

[0138] Specifically, the allocation system clarifies the evaluation objectives of employment quality based on policy guidance, industry demands, and enterprise goals. The evaluation objectives include "improving employment satisfaction", "optimizing salary levels", and "enhancing career development space". By communicating with the human resources department, industry experts, and job seekers, the specific connotations of the evaluation objectives are determined, providing a clear direction for the construction of the subsequent evaluation system, ensuring that the evaluation objectives are consistent with actual needs and have practical significance.

[0139] S104.22. Construct an evaluation system containing multiple evaluation indicators.

[0140] Specifically, the allocation system constructs an evaluation system containing multiple evaluation indicators according to the employment quality evaluation objectives. The multiple evaluation indicators are independent and operable, and can comprehensively reflect employment quality; according to the employment quality evaluation objectives, multiple independent and operable evaluation indicators are selected, which may specifically include:

[0141] Salary level: monthly salary, annual salary, etc.

[0142] Career development: promotion space, training opportunities, etc.

[0143] Working environment: work location, company culture, etc.

[0144] Employment stability: contract term, turnover rate, etc.

[0145] Ensure that the indicators are independent of each other, avoid redundancy, construct a comprehensive and scientific evaluation system, which can reflect employment quality from multiple dimensions, and provide a framework for subsequent data collection and analysis.

[0146] S104.23. Collect data samples related to employment quality and perform dimensionless processing on the data samples.

[0147] Specifically, the allocation system collects data samples related to employment quality and performs dimensionless processing on the data samples to eliminate the dimensional influence between different indicators; after collecting data related to employment quality from channels such as recruitment platforms, enterprise HR systems, and questionnaires, perform dimensionless processing on the data to eliminate the dimensional influence between different indicators: use standardization methods (such as Z-score standardization) or normalization methods (such as Min-Max normalization). For example, convert the salary level into a value between 0 and 1. Improve the comparability of the data, facilitate subsequent analysis and modeling, and eliminate the interference of dimensional differences on the model results.

[0148] S104.24. Select an employment quality status as the reference sequence.

[0149] Specifically, the allocation system selects an employment quality status as the reference sequence, and the employment quality status is the optimal value or the average value; according to business requirements, an employment quality status is selected as the reference sequence: specifically, the optimal value (the combination of the optimal values of each indicator) or the average value (the average value of each indicator) can be selected. For example, select the status with the highest salary level, the largest career development space, and the best working environment as the reference sequence. Provide a benchmark for grey relational analysis, facilitate quantifying the proximity of each indicator to the reference sequence, and ensure that the evaluation results have a clear reference standard.

[0150] S104.25. Calculate the grey relational coefficients between each evaluation indicator and the reference sequence.

[0151] Specifically, the larger the grey relational coefficient, the closer the indicator is to the reference sequence, that is, the greater the impact on employment quality; the allocation system uses the grey relational analysis method to calculate the grey relational coefficients between each evaluation indicator and the reference sequence.

[0152] Grey relational coefficient formula:

[0153]

[0154] Where x0 is the reference sequence, x0(k) is the value of the reference signal at time point k, x i is the evaluation indicator, x i (k) is the value of the signal to be compared at time point k, minmin∣x0(k)-x i (k)∣ is the minimum value of the absolute difference between the reference signal and the signal to be compared at all time points k, maxmax∣x0(k)-x i (k)∣ is the maximum value of the absolute difference between the reference signal and the signal to be compared at all time points k, and ρ is the resolution coefficient (usually taken as 0.5). The larger the grey relational coefficient, the closer the indicator is to the reference sequence, and by quantifying the impact degree of each indicator on employment quality, it provides a basis for weight allocation.

[0155] S104.26. Normalize the grey relational coefficients of each indicator.

[0156] Specifically, the allocation system normalizes the grey relational coefficients of each indicator to obtain the weight values of each indicator, and these weight values reflect the importance of each indicator in the employment quality evaluation; normalize the grey relational coefficients to calculate the weight values of each indicator:

[0157]

[0158] Among them, j is a summation index variable used to iteratively count each element in the set during the summation process. In each iteration, it represents a specific value in the sequence from 1 to n; n represents the total number of samples or data points considered.

[0159] The weight values reflect the importance of each indicator in the employment quality evaluation, provide a scientific weight allocation for the comprehensive evaluation, and ensure that the evaluation results can accurately reflect the actual contributions of each indicator.

[0160] S104.27. Define the optimization objective function.

[0161] Specifically, the allocation system defines the optimization objective function to measure the degree of balance between the supply and demand in the labor market; for example: minimize the supply-demand difference: where D i is the demand, and S i is the supply; maximize the employment quality: where Q i is the employment quality indicator. Use optimization algorithms (such as linear programming, genetic algorithms) to solve the objective function, quantify the degree of balance between the supply and demand in the labor market, and provide a basis for dynamically adjusting the supply and demand.

[0162] S104.28. Use the fuzzy clustering algorithm to calculate the distance between each sample point and the cluster center.

[0163] Specifically, the allocation system uses the fuzzy clustering algorithm to calculate the distance between each sample point and the cluster center, thereby determining the membership degree of each sample point; use the fuzzy clustering algorithm (such as FCM) to assign the sample points to different categories, calculate the distance between each sample point and the cluster center, and determine the membership degree of the sample points. For example, classify talents and positions into three categories: "high match", "medium match", and "low match" to achieve dynamic classification of the labor market, facilitate supply-demand matching, quantify the matching degree through the membership degree, and improve the flexibility of matching.

[0164] S104.29. Assign the sample points to different categories according to the membership degree.

[0165] Specifically, the allocation system assigns the sample points to different categories according to the membership degree, and adjusts the position of the cluster center through an optimization algorithm to achieve dynamic balance between the supply and demand in the labor market. According to the membership degree allocation result, use an optimization algorithm (such as gradient descent) to adjust the position of the cluster center. Repeat the iteration until the cluster center is stable or the maximum number of iterations is reached to achieve dynamic balance between the supply and demand in the labor market, and improve the accuracy and efficiency of matching by optimizing the cluster center.

[0166] This embodiment discloses an employment talent supply and demand allocation method based on data analysis, which specifically includes the following steps:

[0167] S201. Classify and analyze the basic information of the talents using data mining techniques.

[0168] Specifically, the allocation system uses data mining techniques (such as clustering analysis and classification algorithms) to classify the basic information of the talents. For example, the talents are classified into "technical", "management", "sales", etc. according to dimensions such as skills, industries, and experience. Analyze the classification results to identify the talent distribution characteristics and supply and demand trends, and then provide structured talent data for subsequent querying and matching, revealing the supply and demand characteristics of the talent market, and providing references for enterprises and job seekers.

[0169] S202. Receive the talent query requests sent by enterprise users through intelligent terminals.

[0170] Specifically, the allocation system receives the talent query requests sent by enterprise users through intelligent terminals. The talent query requests carry the talent query information for entering the talent query interface. The talent query information includes job requirements and talent characteristics. Among them, the job requirements include job title, skill requirements, and salary range. The talent characteristics include education background, work experience, and skills. Receive the query requests through the API interface or message queue. Provide a convenient query entry to improve the user experience and support real-time querying and dynamic matching.

[0171] S203. Design a multi-dimensional scoring standard according to the job requirements and talent characteristics.

[0172] Specifically, the allocation system designs a multi-dimensional scoring standard, including: Professional skills: The matching degree with the skills required for the job. Work experience: The matching degree with the required working years for the job. Personal qualities: Soft skills such as communication skills and teamwork. Development potential: Learning ability, career planning, etc. Define scoring rules and weights for each dimension, provide a comprehensive evaluation system, ensure that the scoring results are scientific and reasonable, and support flexible adjustment to meet the needs of different enterprises.

[0173] S204. Enterprise users adjust the specific scoring rules and weights of each dimension according to their own needs.

[0174] Specifically, the allocation system provides a visual interface in the system to support enterprise users to customize scoring rules and weights. For example, for technical positions, the weight of "professional skills" can be increased, and for management positions, the weight of "personal qualities" can be increased. Save the rules and weights customized by the user to the database for subsequent query use. By improving the flexibility and pertinence of scoring, meet the personalized needs of enterprises, and enhance the practicality and user stickiness of the system.

[0175] S205. Dynamically adjust the scoring system.

[0176] Specifically, the allocation system dynamically adjusts the scoring system to meet the needs of different enterprise users; it dynamically adjusts the scoring system according to the feedback from enterprise users and market demands. For example, new scoring dimensions (such as innovation ability) are added or the weight allocation is adjusted, and a version control mechanism is used to manage the changes in the scoring system to ensure that the scoring system keeps pace with the times, adapts to market changes, and improves the adaptability and scalability of the system.

[0177] S206. Calculate the matching degree score between the job seeker and the position using the collaborative filtering algorithm.

[0178] Specifically, the allocation system uses the collaborative filtering algorithm (such as user-item collaborative filtering) to calculate the matching degree between the job seeker and the position, and mines potential matching relationships based on historical data (such as the job application records of job seekers and the recruitment records of enterprises). Calculate the matching degree score, and the higher the score, the higher the matching degree, so as to improve the accuracy and personalization of the matching, and discover potential supply and demand relationships in a data-driven manner.

[0179] S207. Conduct a comprehensive evaluation and ranking of talents, form a talent recommendation list and push it to enterprise users.

[0180] Specifically, the allocation system calculates the comprehensive score of talents according to the multi-dimensional scoring criteria and the matching degree score, ranks the talents, generates a recommendation list, and pushes the recommendation list to the intelligent terminal of enterprise users. It is beneficial to provide intuitive recommendation results, facilitate enterprise users to quickly screen talents, improve the recruitment efficiency, and reduce the recruitment cost.

[0181] This embodiment discloses a method for allocating employment talents based on data analysis, which specifically includes the following steps:

[0182] S301. Receive the job query request sent by the individual user through the intelligent terminal.

[0183] Specifically, the allocation system receives the job query request sent by the individual user through the intelligent terminal. The job query request carries job query information for entering the job query interface. The job query information includes personal information and job expectation information; among them, personal information includes education background, work experience, skills, and salary expectation; job expectation information includes industry preference, job type, and work location; receive the query request through the API interface or message queue, provide a convenient query entrance, improve the user experience, and support real-time query and dynamic matching.

[0184] S302. Generate a career development plan and push it to the individual user.

[0185] Specifically, the allocation system queries information based on the said position, generates a career development plan and pushes it to individual users. The career development plan includes industry and region recommendations. Standardize the position query information, such as unifying skill names, industry classifications, etc., use natural language processing (NLP) technology to extract key features (such as skills, industry preferences), analyze the career development potential of individual users based on historical data and market trends, and use machine learning algorithms (such as decision trees, random forests) to predict suitable industries and regions.

[0186] According to the analysis results, generate a career development plan, including: Industry recommendation: Recommend suitable industries (such as IT, finance, education). Region recommendation: Recommend suitable work locations (such as first-tier cities, new first-tier cities). Skill improvement suggestions: Recommend skills that need to be improved (such as Python, project management). Push the career development plan to the intelligent terminal of individual users in a visual form (such as charts, text reports).

[0187] Based on the above method, an embodiment of this application also discloses an employment talent supply and demand allocation system based on data analysis. An employment talent supply and demand allocation system based on data analysis includes:

[0188] Basic information acquisition module, used to acquire talent basic information and position basic information;

[0189] Data processing module, used to clean and sort the talent basic information and position basic information, identify and delete duplicate data, and process missing values and outliers;

[0190] Feature extraction module, used to extract key features from the talent basic information and position basic information using deep learning technology, and weight or screen the key features;

[0191] Matching model establishment module, used to use machine learning algorithms to establish a matching model between the talent basic information and the position basic information;

[0192] Model optimization module, used to analyze the matching results, optimize the model according to the results, adjust parameters or feature weights, and improve the accuracy and efficiency of the matching;

[0193] Model evaluation module, used to evaluate the generalization ability and accuracy of the matching model through cross-validation and confusion matrix methods;

[0194] Continuous optimization module, used to continuously optimize the matching model, and adjust the matching model according to feedback and experience to improve the matching effect.

[0195] An embodiment of the present application also discloses an intelligent terminal, which includes a memory and a processor. Among them, a computer program capable of being loaded and executed by the processor, such as a method for allocating employment talents' supply and demand based on data analysis as described above, is stored on the memory.

[0196] An embodiment of the present application also discloses a computer-readable storage medium. A computer program capable of being loaded and executed by the processor, such as a method for allocating employment talents' supply and demand based on data analysis as described above, is stored in the computer-readable storage medium. The computer-readable storage medium includes, for example, various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs.

[0197] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than limiting the protection scope of the invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on these embodiments, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art can still, without conflict, make combinations, additions, deletions, or other adjustments to the features in the embodiments of the present invention according to the situation without creative efforts, so as to obtain different technical solutions that essentially do not deviate from the concept of the present invention. These technical solutions also belong to the scope of protection of the present invention.

Claims

1. A method for allocating supply and demand of employment talents based on data analysis, characterized in that: The following steps are involved: Obtain basic information of talents and positions; Clean and organize basic information of talents and positions, identify and delete duplicate data, and handle missing values ​​and outliers; Use deep learning technology to extract key features from basic information of talents and positions, and weight or screen the key features; Use machine learning algorithms to establish a matching model between basic information of talents and basic information of positions; Analyze the matching results, optimize the model based on the results, adjust parameters or feature weights, and improve the accuracy and efficiency of matching; The generalization ability and accuracy of the matching model were evaluated through cross-validation and confusion matrix methods; Continuously optimize the matching model and adjust the matching model based on feedback and experience to improve the matching effect.

2. The method for allocating supply and demand of employment talents based on data analysis according to claim 1, characterized in that: The step of using a machine learning algorithm to establish a matching model between the basic information of the talent and the basic information of the position specifically includes: Based on the basic information of talents and positions, talents and positions are classified to identify common characteristics of the supply and demand sides; The matching degree between talents and positions is analyzed through association rule algorithms to discover potential supply and demand relationships.

3. The method for allocating supply and demand of employment talents based on data analysis according to claim 2 is characterized in that: The step of analyzing the matching degree between talents and positions through association rule algorithms to discover potential supply and demand relationships specifically includes: Collect relevant data on talents and positions based on the basic information of the talents and positions, including educational background, work experience, skills, and salary levels; Discretize the data; Use keyword or text similarity-based methods to evaluate the match between resumes and job descriptions; Use the Apriori algorithm to find high-frequency patterns by calculating support and confidence, and then analyze the correlation between talent characteristics and job requirements; Establishing a rule set, and optimizing the matching model according to the rules; Further optimize the generalization ability of the model through cross-validation and confusion matrix; Analyze the talent demand trends in different industries and positions, and reveal the cyclical fluctuations of talent supply and demand through cluster analysis and linear programming models; Based on association rules and machine learning algorithms, we recommend suitable positions to job seekers and suitable talents to companies; Combined with collaborative filtering algorithm, the accuracy and usability of recommendation system can be improved.

4. The method for allocating supply and demand of employment talents based on data analysis according to claim 3 is characterized in that: Before the step of using the Apriori algorithm to calculate support and confidence to find high-frequency patterns and then analyze the correlation between talent characteristics and job requirements, the following steps are also included: The Apriori algorithm mines association rules by constructing frequent item sets, and it is necessary to predetermine the optimal thresholds of support and confidence to improve the accuracy of talent and job matching; Use weighted particle swarm optimization algorithm to dynamically adjust support and confidence thresholds to obtain higher quality rules; Through experiments, we continuously adjust the support and confidence thresholds to find the best combination; Comparative analysis was performed using different datasets and algorithms to verify the effectiveness of the selected thresholds; Evaluate the quality of the generated association rules, where the quality of the association rules includes indicators such as support, confidence, and lift; The thresholds are further optimized based on the evaluation results to ensure that neither too many nor too few rules are generated.

5. The method for allocating supply and demand of employment talents based on data analysis according to claim 1, characterized in that: The step of using a machine learning algorithm to establish a matching model between the basic information of the talent and the basic information of the position also includes: Obtaining employment quality evaluation targets; According to the employment quality evaluation objectives, an evaluation system including multiple evaluation indicators is constructed. The multiple evaluation indicators are independent and operable and can comprehensively reflect the employment quality. Collect data samples related to employment quality and perform dimensionless processing on the data samples to eliminate the dimensional impact between different indicators; Selecting an employment quality state as a reference sequence, wherein the employment quality state is an optimal value or an average value; Calculate the grey correlation coefficient between each evaluation index and the reference sequence. The larger the grey correlation coefficient is, the closer the index is to the reference sequence, that is, the greater the impact on employment quality. The grey correlation coefficient of each indicator is normalized to obtain the weight value of each indicator, which reflects the importance of each indicator in the employment quality evaluation. Define the optimization objective function to measure the degree of balance between supply and demand in the labor market; The fuzzy clustering algorithm is used to calculate the distance between each sample point and the cluster center, so as to determine the membership degree of each sample point; The sample points are assigned to different categories according to the degree of membership, and the positions of the cluster centers are adjusted through an optimization algorithm to achieve a dynamic balance between supply and demand in the labor market.

6. The method for allocating supply and demand of employment talents based on data analysis according to claim 1, characterized in that: Also includes: Using data mining technology to classify and analyze the basic information of the talents; Receiving a talent query request sent by an enterprise user through a smart terminal, the talent query request carrying talent query information for entering a talent query interface, the talent query information including job requirements and talent characteristics; Design a multi-dimensional scoring standard based on the job requirements and talent characteristics, including professional skills, work experience, personal qualities, and development potential; Enterprise users adjust the specific scoring rules and weights of each dimension according to their own needs; Dynamically adjust the scoring system to meet the needs of different corporate users; Use collaborative filtering algorithm to calculate the matching score between job seekers and positions. The higher the score, the better the matching degree. Talents are comprehensively scored and ranked to form a talent recommendation list and pushed to corporate users.

7. The method for allocating employment talent supply and demand based on data analysis according to claim 1, characterized in that: Also includes: Receiving a job query request sent by an individual user through a smart terminal, the job query request carrying job query information for entering a job query interface, the job query information including personal information and job expectation information; A career development plan is generated based on the job query information and pushed to individual users, wherein the career development plan includes industry and region recommendations.

8. A talent supply and demand allocation system based on data analysis, characterized in that: include: Basic information acquisition module, used to obtain basic information of talents and positions; Data processing module, used to clean and organize basic information of talents and positions, identify and delete duplicate data, and process missing values ​​and outliers; The feature extraction module is used to extract key features from basic information of talents and positions using deep learning technology, and to weight or screen the key features; A matching model building module is used to use a machine learning algorithm to build a matching model between basic information of talents and basic information of positions; Model optimization module, used to analyze matching results, optimize the model according to the results, adjust parameters or feature weights, and improve matching accuracy and efficiency; Model evaluation module, used to evaluate the generalization ability and accuracy of the matching model through cross-validation and confusion matrix methods; The continuous optimization module is used to continuously optimize the matching model and adjust the matching model based on feedback and experience to improve the matching effect.

Images