Dynamic soft skill quantification framework generation method and device based on artificial intelligence technology

Abstract

The invention relates to a dynamic soft skill quantification framework generation method and device based on an artificial intelligence technology. The method comprises the steps of obtaining current recruitment text data of a large number of users, and performing text preprocessing on the current recruitment text data to obtain each standard recruitment sample; in each standard recruitment sample, screening a candidate keyword set associated with a soft skill, and based on the candidate keyword set and each standard recruitment sample, identifying association information between candidate keywords; and collecting an initial soft skill framework, and generating a soft skill quantification framework through a soft skill quantification analysis strategy based on the initial soft skill framework and the association information among the candidate keywords. By adopting the method, the accuracy, comprehensiveness and practicability of the constructed soft skill quantification framework can be improved.

Description

Technical Field

[0001] The present application relates to the field of big data technology, and in particular to a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for generating a dynamic soft skill quantification framework based on artificial intelligence technology. Background Art

[0002] In today's knowledge-based economy and rapidly evolving technological landscape, accurately identifying, understanding, and quantifying the labor market's demand for various skills, particularly soft skills, as reflected in the vast amount of online job postings, is crucial for corporate talent strategies, individual career development, education and training reform, and even macroeconomic policymaking. Faced with the challenges of massive amounts of unstructured recruitment text data, as well as the complex and diverse concepts and expressions of soft skills, analyzing and quantifying the soft skills of the workforce is a current research priority.

[0003] Traditional soft skills analysis techniques often involve manually classifying soft skills and constructing a soft skills framework through subjective human quantification to analyze the soft skills information of different workforces. However, human analysis is subject to subjective factors and one-sidedness, which often leads to a one-sided framework for soft skills and poor quantitative accuracy of soft skills. Summary of the Invention

[0004] Based on this, it is necessary to provide a method, device, computer equipment, computer-readable storage medium and computer program product for generating a dynamic soft skill quantification framework based on artificial intelligence technology to address the above technical problems.

[0005] In a first aspect, the present application provides a method for generating a dynamic soft skills quantitative framework based on artificial intelligence technology, comprising:

[0006] Acquire current recruitment text data of a large number of users, and perform text preprocessing on the current recruitment text data to obtain various standardized recruitment samples;

[0007] In each of the standard recruitment samples, a candidate keyword set associated with soft skills is screened, and based on the candidate keyword set and each of the standard recruitment samples, association information between the candidate keywords is identified;

[0008] An initial soft skill framework is collected, and based on the initial soft skill framework and the association information between the candidate keywords, a soft skill quantitative analysis strategy is used to generate a soft skill quantitative framework.

[0009] Optionally, the text preprocessing of the current recruitment text data to obtain various standard recruitment samples includes:

[0010] Performing text normalization processing on each of the current recruitment text data according to a normalized text processing strategy to obtain each standardized current recruitment text data;

[0011] Using a Chinese word segmentation tool, perform Chinese word segmentation processing on the current recruitment text data of each standard to obtain each standard recruitment text sequence, and perform stop word filtering processing on each standard recruitment text sequence to obtain each initial recruitment sample data;

[0012] A soft skill dictionary is collected, and based on the soft skill dictionary, abnormal text data in each of the initial recruitment sample data is filtered through an abnormal text filtering strategy to obtain each of the standard recruitment samples.

[0013] Optionally, screening a set of candidate keywords associated with soft skills in each of the standard recruitment samples includes:

[0014] Based on the standard recruitment samples, high-frequency words are screened by using a high-frequency word extraction strategy to obtain a high-frequency word data set, and soft skill-related word data is screened in the high-frequency word data set;

[0015] Collecting candidate soft skill keywords from the enterprise side and candidate soft skill keywords from the job seeker side, and identifying market soft skill supply and demand information based on the candidate soft skill keywords from the enterprise side and the candidate soft skill keywords from the job seeker side;

[0016] Based on the soft skill related vocabulary data and the market soft skill supply and demand information, a candidate keyword set associated with soft skills is identified.

[0017] Optionally, identifying association information between the candidate keywords based on the candidate keyword set and the standard recruitment samples includes:

[0018] In the candidate keyword set, screening target candidate keywords corresponding to each standard recruitment sample;

[0019] For each standard recruitment sample, based on each target candidate keyword corresponding to the standard recruitment sample, a strategy is defined through a context window to identify contextual meaning information corresponding to each target candidate keyword, and based on the contextual meaning information corresponding to each target candidate keyword, each polysemous keyword is screened;

[0020] Using an ambiguous keyword processing strategy, all meaning information corresponding to each ambiguous keyword is identified, and based on all meaning information corresponding to each ambiguous keyword and contextual meaning information corresponding to each non-ambiguous keyword, each candidate keyword is divided into keyword categories;

[0021] The keyword category corresponding to each candidate keyword is used as the association information between the candidate keywords.

[0022] Optionally, the acquisition of an initial soft skills framework includes:

[0023] Collecting data related to each soft skill, and extracting soft skill theory keywords from each soft skill related data, as well as soft skill meaning information corresponding to each soft skill theory keyword;

[0024] Based on the soft skill meaning information corresponding to each soft skill theory keyword, the hierarchical relationship between each soft skill theory keyword is identified through the soft skill hierarchical model;

[0025] Based on the hierarchical relationship between each soft skill theory keyword, an association graph construction strategy is used to build an initial soft skill framework that includes all soft skill theory keywords.

[0026] Optionally, generating a soft skill quantification framework based on the initial soft skill framework and the association information between the candidate keywords through a soft skill quantification analysis strategy includes:

[0027] Based on the association information between the candidate keywords and the initial soft skill framework, a basic word frequency matrix is ​​generated; the basic word frequency matrix includes the target keyword from the candidate keywords;

[0028] Based on each of the standardized recruitment samples, the standardized word frequency of each of the target keywords is calculated, and based on the standardized word frequency of each of the target keywords, the basic word frequency matrix is ​​adjusted to obtain a standardized word frequency matrix;

[0029] Based on the standardized word frequency matrix, identifying the quantitative results of each target keyword through a relative importance algorithm, and based on the quantitative results of each target keyword and the standardized word frequency matrix, identifying the level quantitative results corresponding to each soft skill level;

[0030] Soft skill dynamic data of different analysis types are collected, and based on the soft skill dynamic data of each analysis type and the level quantification results corresponding to each soft skill level, the standardized word frequency matrix is ​​adjusted to obtain a soft skill quantification framework.

[0031] In a second aspect, the present application also provides a device for generating a dynamic soft skills quantification framework based on artificial intelligence technology, comprising:

[0032] An acquisition module is used to acquire current recruitment text data of a large number of users and perform text preprocessing on the current recruitment text data to obtain various standardized recruitment samples;

[0033] an identification module for screening a set of candidate keywords associated with soft skills in each of the standard recruitment samples, and identifying association information between each of the candidate keywords based on the set of candidate keywords and each of the standard recruitment samples;

[0034] The generation module is used to collect an initial soft skill framework and generate a soft skill quantitative framework based on the initial soft skill framework and the association information between the candidate keywords through a soft skill quantitative analysis strategy.

[0035] Optionally, the acquisition module is specifically configured to:

[0036] Performing text normalization processing on each of the current recruitment text data according to a normalized text processing strategy to obtain each standardized current recruitment text data;

[0037] Using a Chinese word segmentation tool, perform Chinese word segmentation processing on the current recruitment text data of each standard to obtain each standard recruitment text sequence, and perform stop word filtering processing on each standard recruitment text sequence to obtain each initial recruitment sample data;

[0038] A soft skill dictionary is collected, and based on the soft skill dictionary, abnormal text data in each of the initial recruitment sample data is filtered through an abnormal text filtering strategy to obtain each of the standard recruitment samples.

[0039] Optionally, the identification module is specifically configured to:

[0040] Based on the standard recruitment samples, high-frequency words are screened by using a high-frequency word extraction strategy to obtain a high-frequency word data set, and soft skill-related word data is screened in the high-frequency word data set;

[0041] Collecting candidate soft skill keywords from the enterprise side and candidate soft skill keywords from the job seeker side, and identifying market soft skill supply and demand information based on the candidate soft skill keywords from the enterprise side and the candidate soft skill keywords from the job seeker side;

[0042] Based on the soft skill related vocabulary data and the market soft skill supply and demand information, a candidate keyword set associated with soft skills is identified.

[0043] Optionally, the identification module is specifically configured to:

[0044] In the candidate keyword set, screening target candidate keywords corresponding to each standard recruitment sample;

[0045] For each standard recruitment sample, based on each target candidate keyword corresponding to the standard recruitment sample, a strategy is defined through a context window to identify contextual meaning information corresponding to each target candidate keyword, and based on the contextual meaning information corresponding to each target candidate keyword, each polysemous keyword is screened;

[0046] Using an ambiguous keyword processing strategy, all meaning information corresponding to each ambiguous keyword is identified, and based on all meaning information corresponding to each ambiguous keyword and contextual meaning information corresponding to each non-ambiguous keyword, each candidate keyword is divided into keyword categories;

[0047] The keyword category corresponding to each candidate keyword is used as the association information between the candidate keywords.

[0048] Optionally, the generating module is specifically configured to:

[0049] Collecting data related to each soft skill, and extracting soft skill theory keywords from each soft skill related data, as well as soft skill meaning information corresponding to each soft skill theory keyword;

[0050] Based on the soft skill meaning information corresponding to each soft skill theory keyword, the hierarchical relationship between each soft skill theory keyword is identified through the soft skill hierarchical model;

[0051] Based on the hierarchical relationship between each soft skill theory keyword, an association graph construction strategy is used to build an initial soft skill framework that includes all soft skill theory keywords.

[0052] Optionally, the generating module is specifically configured to:

[0053] Based on the association information between the candidate keywords and the initial soft skill framework, a basic word frequency matrix is ​​generated; the basic word frequency matrix includes the target keyword from the candidate keywords;

[0054] Based on each of the standardized recruitment samples, the standardized word frequency of each of the target keywords is calculated, and based on the standardized word frequency of each of the target keywords, the basic word frequency matrix is ​​adjusted to obtain a standardized word frequency matrix;

[0055] Based on the standardized word frequency matrix, identifying the quantitative results of each target keyword through a relative importance algorithm, and based on the quantitative results of each target keyword and the standardized word frequency matrix, identifying the level quantitative results corresponding to each soft skill level;

[0056] Soft skill dynamic data of different analysis types are collected, and based on the soft skill dynamic data of each analysis type and the level quantification results corresponding to each soft skill level, the standardized word frequency matrix is ​​adjusted to obtain a soft skill quantification framework.

[0057] In a third aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any one of the methods described in the first aspect when executing the computer program.

[0058] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of any one of the methods in the first aspect.

[0059] In a fifth aspect, the present application provides a computer program product, wherein the computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of any one of the methods in the first aspect are implemented.

[0060] The above-mentioned dynamic soft skill quantification framework generation method, device, computer equipment, computer-readable storage medium and computer program product based on artificial intelligence technology obtains current recruitment text data of a large number of users and performs text preprocessing on the current recruitment text data to obtain various standardized recruitment samples; in each of the standardized recruitment samples, a set of candidate keywords associated with soft skills is screened, and based on the candidate keyword set and each of the standardized recruitment samples, the correlation information between each of the candidate keywords is identified; an initial soft skill framework is collected, and based on the initial soft skill framework and the correlation information between each of the candidate keywords, a soft skill quantification framework is generated through a soft skill quantification analysis strategy. This solution combines a large amount of current recruitment text data to perform text preprocessing and keyword extraction, thereby identifying a set of candidate keywords associated with soft skills. Then, by combining the contextual relationship of the candidate keyword set in the standardized recruitment sample and the meaning information of each candidate keyword in the candidate keyword set, and based on the real-time demand information of the market dynamics, this solution collects the latest current recruitment text data from the market each time, and then combines the latest recruitment needs of the market recruiters and the latest resume information of the job seekers to construct an initial soft skills framework, and then performs soft skills quantitative analysis to generate a soft skills quantitative framework. This effectively avoids the problems of the existing technology of directly identifying soft skills information based on historical text information, which is single, fixed, and traditional, as well as the one-sidedness and poor quantitative accuracy of the artificially formulated soft skills framework, and effectively improves the accuracy, comprehensiveness, and practicality of the constructed soft skills quantitative framework. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments of the present application or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying any creative work.

[0062] Figure 1 A flowchart of a method for generating a dynamic soft skills quantification framework based on artificial intelligence technology in one embodiment is provided;

[0063] Figure 2 Schematic diagram of the framework structure of the initial soft skills framework in one embodiment;

[0064] Figure 3 Schematic diagram of the processing flow of a text preprocessing module in one embodiment;

[0065] Figure 4 Schematic diagram of the screening process of the keyword extraction and screening module in one embodiment;

[0066] Figure 5 Schematic diagram of the recognition process of polysemous word recognition in one embodiment;

[0067] Figure 6 Schematic diagram of the analysis process of the standardized quantification and analysis module in one embodiment;

[0068] Figure 7 A schematic diagram of a process for generating an example of a dynamic soft skills quantification framework based on artificial intelligence technology in one embodiment;

[0069] Figure 8 A structural block diagram of a device for generating a dynamic soft skill quantification framework based on artificial intelligence technology in one embodiment;

[0070] Figure 9 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0071] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0072] The embodiment of the present application provides a method for generating a dynamic soft skill quantitative framework based on artificial intelligence technology, which can be applied to a dynamic soft skill quantitative framework generation system based on artificial intelligence technology. The system can be applied to a terminal, which can be, but is not limited to, various personal computers, laptops, mid-range computers, etc. Among them, the terminal combines a large amount of current recruitment text data, performs text preprocessing and keyword extraction, thereby identifying a candidate keyword set associated with soft skills, and then combines the contextual relationship of the candidate keyword set in the standard recruitment sample and the meaning information of each candidate keyword in the candidate keyword set with the real-time demand information of the market dynamics. By collecting the latest current recruitment text data in the market each time, and then combining the latest recruitment needs of the market recruiter and the latest resume information of the job seeker, an initial soft skill framework is constructed, and then a soft skill quantitative analysis is performed to generate a soft skill quantitative framework, which effectively avoids the problems of the prior art of directly identifying soft skill information based on historical text information, which is single, fixed, and traditional, as well as the one-sidedness and poor quantitative accuracy of the artificially formulated soft skill framework, and effectively improves the accuracy, comprehensiveness, and practicality of the constructed soft skill quantitative framework.

[0073] In an exemplary embodiment, Figure 1 As shown, a method for generating a dynamic soft skill quantification framework based on artificial intelligence technology is provided. The method is applied to a terminal as an example for explanation, and includes the following steps S101 to S103.

[0074] in:

[0075] Step S101: obtaining current recruitment text data of a large number of users, and performing text preprocessing on the current recruitment text data to obtain various standard recruitment samples.

[0076] In this embodiment, the terminal, upon obtaining user authorization, obtains the latest recruitment information for job seekers from the enterprise side in real time, and also obtains the latest job search information from the client side in real time, thereby obtaining current recruitment text data corresponding to different users. This current recruitment text data is dynamic recruitment text data obtained by the inventors in real time. That is, each time this method is executed to construct a soft skills quantification framework, it is combined with the latest recruitment text data to ensure the dynamic, real-time, market adaptability, and accuracy of the soft skills quantification framework. The terminal then performs text preprocessing on the obtained recruitment text data using the text preprocessing module within the soft skills framework architecture designed by the inventors of this solution, obtaining standardized recruitment samples. The preprocessed standardized recruitment samples are saved in a standardized format, containing the original text, word segmentation results, character and word count information, and other information, providing structured input data for subsequent modules. The soft skills framework architecture, designed by the inventors of this solution, is used to construct a real-time, updated soft skills quantification framework after comprehensive analysis of the real-time recruitment information obtained from each enterprise and the job search information of each job seeker. The soft skills quantification framework, which is structured by dividing each soft skill into three levels, includes both classification information for different soft skills and quantitative analysis of their importance based on current market demand (both the needs of hiring companies and the needs of job seekers). The specific construction and quantification process will be explained in detail later. The soft skills framework architecture includes five main modules: text preprocessing, keyword extraction and screening, context analysis and classification, multi-level framework construction, and standardized quantification and analysis.

[0077] Step S102 : screening a set of candidate keywords associated with soft skills in each standard recruitment sample, and identifying association information between each candidate keyword based on the candidate keyword set and each standard recruitment sample.

[0078] In this embodiment, the terminal uses a keyword extraction and screening module to select candidate soft skill-related keywords from each standardized recruitment sample. This invention employs an iterative keyword optimization strategy, continuously optimizing the keyword set through subsequent context analysis and practical application verification, ultimately forming a final set of 384 carefully selected soft skill keywords. This iterative optimization mechanism significantly improves the accuracy and representativeness of keywords, providing high-quality foundational data for subsequent framework construction.

[0079] Then, based on the candidate keyword set and each standardized recruitment sample, the terminal identifies the association information between each candidate keyword through the context analysis and classification module. The specific screening and identification process will be described in detail later. Among them, the association information between each candidate keyword is the keyword category corresponding to each candidate keyword. Among them, with the continuous accumulation of new data and the emergence of new soft skill expressions, the present invention designs a dynamic update mechanism for classification rules. Regularly analyze the newly collected text data, identify new polysemous words and context patterns, and promptly update and optimize the classification rules to maintain their timeliness and applicability.

[0080] Step S103 : collecting an initial soft skill framework, and generating a soft skill quantitative framework based on the initial soft skill framework and the association information between candidate keywords through a soft skill quantitative analysis strategy.

[0081] In this embodiment, Figure 2 As shown, the terminal collects the initial soft skill framework, and based on the initial soft skill framework and the correlation information between each candidate keyword, generates a soft skill quantitative framework through a soft skill quantitative analysis strategy. Among them, the collected initial soft skill framework is for the staff to conduct data-driven framework verification and optimization, polysemy processing and framework adjustment based on the soft skill keywords after historical document analysis, and finally through the framework dynamic update mechanism, as well as framework verification and tuning, to build a soft skill framework. The specific construction process will be described in detail later. Among them, through multiple rounds of verification and tuning, the present invention finally formed a complete soft skill framework system including 2 first-level dimensions, 12 second-level dimensions, 34 third-level dimensions and 384 keywords. The framework has both a solid theoretical foundation and good data support, and can comprehensively and accurately reflect the soft skill composition and demand characteristics in the current labor market. The specific method of generating a soft skill quantitative framework will be described in detail later.

[0082] Based on the above scheme, by combining a large amount of current recruitment text data, text preprocessing and keyword extraction are performed to identify a set of candidate keywords associated with soft skills. Then, by combining the contextual relationship of the candidate keyword set in the standardized recruitment sample and the meaning information of each candidate keyword in the candidate keyword set, and based on the real-time demand information of the market dynamics, this scheme collects the latest current recruitment text data from the market each time, and then combines the latest recruitment needs of the market recruiters and the latest resume information of the job seekers to construct an initial soft skills framework, and then performs soft skills quantitative analysis to generate a soft skills quantitative framework. This effectively avoids the problems of the existing technology of directly identifying soft skills information based on historical text information, which is single, fixed, and traditional, as well as the one-sidedness and poor quantitative accuracy of the artificially formulated soft skills framework, and effectively improves the accuracy, comprehensiveness, and practicality of the constructed soft skills quantitative framework.

[0083] Optionally, the current recruitment text data is subjected to text preprocessing to obtain various standardized recruitment samples, including: performing text normalization processing on each current recruitment text data according to a normalized text processing strategy to obtain various standardized current recruitment text data; performing Chinese word segmentation processing on each standardized current recruitment text data through a Chinese word segmentation tool to obtain various standardized recruitment text sequences, and performing stop word filtering processing on each standardized recruitment text sequence to obtain various initial recruitment sample data; collecting a soft skills dictionary, and based on the soft skills dictionary, filtering out abnormal text data in each initial recruitment sample data through an abnormal text filtering strategy to obtain various standardized recruitment samples.

[0084] In this embodiment, the terminal performs text normalization processing on each current recruitment text data according to the normalized text processing strategy to obtain each standardized current recruitment text data. Figure 3 As shown, the specific content of the standardized text processing strategy is to standardize the recruitment text, including removing HTML tags, processing special characters, and unifying punctuation marks. This invention uses regular expression matching to identify and remove interfering elements such as HTML tags and consecutive hyphens, while replacing special characters such as underscores and equal signs to ensure the consistency and readability of the text format.

[0085] Then, the terminal uses the Chinese word segmentation tool to perform Chinese word segmentation on the current recruitment text data of each standard to obtain each standard recruitment text sequence, and performs stop word filtering on each standard recruitment text sequence to obtain each initial recruitment sample data. Figure 3 As shown in the figure, the Chinese word segmentation process is carried out by filtering with stop words: a predefined stop word list is loaded to remove common stop words and HTML-related tag words. In addition to conventional stop words, the stop word list of the present invention also specifically adds a large number of HTML-related tags and style words, such as div, span, font, style, etc., which effectively improves the text quality.

[0086] Finally, the terminal uses the soft skills dictionary (i.e., custom dictionary) constructed by the inventor of this solution and the abnormal text filtering strategy to filter the abnormal text data in each initial recruitment sample data to obtain each standard recruitment sample, specifically, Figure 3As shown, the custom dictionary is constructed by constructing a custom dictionary for professional terms and compound words in the field of soft skills and loading it into the word segmenter. The present invention particularly strengthens the recognition of soft skill professional terms, such as compound words such as "program design", "communication skills", and "problem-solving ability", which significantly improves the recognition accuracy of soft skill expressions. The abnormal text filtering strategy is to first load the word segmenter, then re-execute the optimized word segmentation step, calculate the number of words in the text and the number of times after word segmentation, and finally filter the abnormal text samples to obtain various standardized recruitment samples. Among them, before filtering the abnormal text data in each initial recruitment sample data, it is necessary to perform quality assessment and filtering on the word segmentation results. The specific assessment and filtering methods are:

[0087] 1. Word and phrase count: Calculate the original word count and word count after tokenization for each recruitment text for subsequent standardization processing and quality assessment.

[0088] 2. Abnormal sample filtering: Eliminate samples with text that is too short (no substantive content) or too long (may contain a lot of redundant information) to ensure the representativeness and reliability of the data.

[0089] 3. Ensure format consistency: unify the format and structure of texts from different sources to facilitate subsequent batch processing and comparative analysis.

[0090] Based on the above solution, preprocessed text data is saved in a standardized format, including the original text, word segmentation results, character and word counts, and other information, providing structured input data for subsequent modules. This improves the standardization and recognizability of text data.

[0091] Optionally, in each standardized recruitment sample, a candidate keyword set associated with soft skills is screened, including: based on each standardized recruitment sample, through a high-frequency vocabulary extraction strategy, screening each high-frequency vocabulary to obtain a high-frequency vocabulary data set, and screening soft skill-related vocabulary data in the high-frequency vocabulary data set; collecting candidate soft skill keywords from the enterprise side and candidate soft skill keywords from the job seeker side, and identifying market soft skill supply and demand information based on the candidate soft skill keywords from the enterprise side and the candidate soft skill keywords from the job seeker side; identifying a candidate keyword set associated with soft skills based on the soft skill-related vocabulary data and the market soft skill supply and demand information.

[0092] In this embodiment, Figure 4As shown, based on standardized recruitment samples, the terminal uses the high-frequency vocabulary extraction strategy in the keyword extraction and screening module to screen high-frequency vocabulary, obtaining a high-frequency vocabulary dataset. Within this high-frequency vocabulary dataset, it then filters soft-skill-related vocabulary data. Specifically, the terminal performs word frequency statistics on the segmented text to obtain a basic dataset of vocabulary and its frequency of occurrence. The present invention employs an efficient word frequency statistics algorithm to statistically analyze the segmentation results of all recruitment texts, forming an initial dataset containing vocabulary and its frequency. Compared to existing technologies, the present invention not only counts the frequency of individual vocabulary words but also specifically focuses on phrases and compound words related to soft skills, improving its ability to capture soft-skill expressions. Then, based on the Pareto principle (80 / 20 rule), the terminal selects high-frequency vocabulary words whose cumulative frequency reaches 80% of the total frequency as the basic dataset. For example, this solution screens approximately 2,250 high-frequency vocabulary words from an original dataset of approximately 400,000 individual vocabulary words, significantly reducing the computational complexity of subsequent processing while retaining most of the key information. This screening method is more scientific than the existing method of simply setting a frequency threshold, and can adaptively determine the screening range based on the actual data distribution. Finally, the terminal combines manual judgment and auxiliary analysis methods to screen candidate keywords related to soft skills from high-frequency vocabulary. For example, this solution accurately identified 441 candidate keywords related to soft skills from high-frequency vocabulary. The auxiliary analysis method is a semantic analysis method corresponding to a large language model based on natural language processing technology.

[0093] Then, the terminal collects candidate soft skill keywords from the enterprise side and candidate soft skill keywords from the job seeker side, and identifies market soft skill supply and demand information based on the candidate soft skill keywords from the enterprise side and the job seeker side. The terminal then merges and removes duplicate candidate keywords from the enterprise side and the job seeker side to form a unified set of candidate keywords. The present invention forms a preliminary set of 618 candidate words by comprehensively analyzing the soft skill expressions in the company recruitment texts and the job seeker self-assessment texts. This dual-perspective integration method is more comprehensive than the existing technology and can capture the labor market's expression and demand for soft skills in all aspects.

[0094] Finally, the terminal identifies a set of candidate keywords associated with soft skills based on soft skill-related vocabulary data and market soft skill supply and demand information.

[0095] Based on the above scheme, the candidate keyword set was continuously optimized through multiple rounds of iterative analysis and verification. This paper employs an iterative keyword optimization strategy, continuously optimizing the keyword set through subsequent contextual analysis and practical application verification, ultimately forming a final set of 384 carefully selected soft skill keywords. This iterative optimization mechanism significantly improved the accuracy and representativeness of keywords, providing high-quality foundational data for subsequent framework construction.

[0096] Optionally, based on the candidate keyword set and each standardized recruitment sample, the association information between each candidate keyword is identified, including: in the candidate keyword set, screening each target candidate keyword corresponding to each standardized recruitment sample; for each standardized recruitment sample, based on each target candidate keyword corresponding to the standardized recruitment sample, defining a strategy through a context window, identifying the contextual meaning information corresponding to each target candidate keyword, and screening each polysemous keyword based on the contextual meaning information corresponding to each target candidate keyword; through a polysemous keyword processing strategy, identifying all meaning information corresponding to each polysemous keyword, and dividing each candidate keyword into keyword categories based on all meaning information corresponding to each polysemous keyword and the contextual meaning information corresponding to each non-polysemous keyword; using the keyword category corresponding to each candidate keyword as the association information between each candidate keyword.

[0097] In this embodiment, the terminal screens the target candidate keywords corresponding to each standard recruitment sample in the candidate keyword set. Then for each standard recruitment sample, the terminal defines a strategy through a context window based on the target candidate keywords corresponding to the standard recruitment sample, and identifies the contextual meaning information corresponding to each target candidate keyword. Specifically, the terminal defines a context window for each soft skill keyword to capture its contextual information in the text. The present invention innovatively proposes a context window structure of "three words before and after", that is, extracting three participles before and after the target keyword to form a seven-word window (three words before and after + target word + three words after). This window design can not only capture sufficient contextual information, but also control the computational complexity, and achieves a good balance between practicality and accuracy.

[0098] The terminal then identifies polysemous soft skill keywords through a combination of word frequency distribution analysis and manual judgment. This invention specifically focuses on keywords that may belong to different soft skill categories in different contexts, such as "interaction," "active," and "friends." By analyzing the distribution characteristics and collocation patterns of these words in different contexts, their polysemous characteristics are identified, laying the foundation for subsequent accurate classification. This polysemous word identification method is a key innovation of this invention, resolving the accuracy issues caused by the single classification of keywords in the prior art.

[0099] The terminal screens each polysemous keyword based on the contextual meaning information corresponding to each target candidate keyword; specifically, the recognition process of the polysemous keyword (i.e., polysemous word) is as follows: Figure 5As shown in the figure, 1) all text segments containing "interaction" are extracted; 2) the context window of the three words before and after "interaction" is extracted for each segment; 3) feature words in the context are analyzed, such as "product", "user", "experience" or "front-end", "development", etc.; 4) classification rules are applied based on the feature words to determine the category of "interaction" in this context; finally, the terminal classifies "interaction" in this instance into the corresponding soft skill category based on the classification results.

[0100] Then, the terminal uses the polysemous keyword processing strategy to identify all the meaning information corresponding to each polysemous keyword, and divides each candidate keyword into keyword categories based on all the meaning information corresponding to each polysemous keyword and the contextual meaning information corresponding to each non-polysemous keyword. Specifically, the terminal performs pattern analysis on the context window of the polysemous word to extract feature words and patterns that can distinguish different meanings. The present invention adopts a pattern recognition method based on context feature words. For each polysemous word, the word distribution and co-occurrence pattern in its context window are analyzed to identify feature words that can clearly distinguish its different meanings. For example, for the word "interaction": when words such as "user", "product", "experience", and "demand" appear in the context, it reflects interpersonal communication skills; when words such as "front-end", "Ajax", and "development" appear in the context, it reflects technical implementation capabilities. This pattern analysis method based on context feature words can accurately capture the meaning differences of keywords in different contexts and provide a basis for accurate classification.

[0101] Then, the terminal formulates precise classification rules for each polysemous word based on the results of the context pattern analysis. The present invention has formulated a series of context-based classification rules for different polysemous words, which realizes the precision and automation of keyword classification. These rules are not simple one-size-fits-all divisions, but complex logical judgments based on contextual feature words, which can flexibly respond to various contextual situations. For example, for the word "active": when the context reflects personal traits (such as "active thinking"), it is classified as "thinking ability-critical thinking"; when the context reflects the team atmosphere (such as "team activity"), it is classified as "management ability-team management"; this rule-based precise classification method is an important innovation of the present invention, which can significantly improve the accuracy of soft skill keyword recognition.

[0102] Finally, the terminal uses the keyword category corresponding to each candidate keyword as the association information between the candidate keywords.

[0103] Based on the above solution, by combining the context window to identify and analyze polysemous words in soft skill related terms, this solves the accuracy problem caused by the single classification of keywords in the existing technology and improves the accuracy of soft skill classification.

[0104] Optionally, collecting an initial soft skill framework includes: collecting each soft skill association data, and extracting the soft skill theory keywords in each soft skill association data, as well as the soft skill meaning information corresponding to each soft skill theory keyword; based on the soft skill meaning information corresponding to each soft skill theory keyword, identifying the hierarchical relationship between each soft skill theory keyword through a soft skill hierarchical model; based on the hierarchical relationship between each soft skill theory keyword, constructing an initial soft skill framework containing all soft skill theory keywords through an association graph construction strategy.

[0105] In this embodiment, the terminal collects soft skill-related data and extracts the soft skill theory keywords and the corresponding soft skill meaning information from each soft skill-related data. The soft skill-related data is obtained by screening scientific literature and identifying related text related to soft skills. When collecting the soft skill-related data, research samples can be manually screened and input into the terminal to obtain the soft skill-related data. This manual screening method, for example, uses inclusive search keywords such as "Soft Skill" and "Non-cognitive Skill"; limits the screening range to 1980 to the latest year; selects high-quality literature based on journal standards such as UTD24, FT50, and ABS; and expands the research scope through both backward and forward tracing strategies. In the actual screening process, the inventors selected 38 core literature as a research sample. The terminal then extracted 384 soft skill theory keywords from the research sample, laying the theoretical foundation for framework construction.

[0106] Then, based on the soft skill meaning information corresponding to each soft skill theory keyword, the terminal identifies the hierarchical relationship between each soft skill theory keyword through the soft skill hierarchical model. Specifically, the terminal classifies 384 soft skill theory keywords based on the soft skills mentioned in the theoretical literature in a bottom-up manner to construct a preliminary multi-level soft skill framework. The present invention adopts the method of "three experts back-to-back classification combined with AI assistance" to jointly propose a four-layer framework structure of "keywords-third-level soft skills-second-level soft skills-first-level soft skills", which divides soft skills into two first-level dimensions: interpersonal soft skills and intrinsic soft skills, and has 13 second-level dimension skills. The specific implementation process includes:

[0107] First, three domain experts (i.e., staff) independently classified and summarized 384 theoretical keywords to obtain classification information between each soft skill. Then, based on the soft skill meaning information corresponding to each soft skill theoretical keyword, the terminal identified the hierarchical relationship between each soft skill theoretical keyword through a soft skill hierarchical model. The soft skill hierarchical model is a large language model based on natural language processing technology, that is, the terminal analyzes the semantic similarity and hierarchical relationship between each theoretical keyword through the large language model.

[0108] Finally, the terminal constructs an initial soft skill framework containing all soft skill theory keywords based on the semantic similarity and hierarchical relationship. Figure 2 Specifically, the terminal uses actual current recruitment text data to verify and optimize the preliminary framework, forming a more refined and practical soft skills framework through a deep integration of data and theory. This invention conducts text mining on job information and job seeker data from a large number of recruitment platforms, screening out high-frequency and meaningful soft skills keywords based on the Pareto principle, and comparing and integrating them with existing definitions of soft skills in the literature. This bidirectional alignment of theory and data is a unique innovation of the present invention and is more scientific and comprehensive than existing methods based solely on theory or data.

[0109] Based on the above solution, by combining the collected soft skill data, through manual classification and AI-assisted analysis technology, an initial soft skill framework containing all soft skill theory keywords is comprehensively constructed, which improves the comprehensiveness, accuracy, and scientificity of the initial soft skill framework.

[0110] Optionally, based on the initial soft skill framework and the correlation information between each candidate keyword, a soft skill quantitative analysis strategy is used to generate a soft skill quantitative framework, including: generating a basic word frequency matrix based on the correlation information between each candidate keyword and the initial soft skill framework; the basic word frequency matrix includes target keywords from each candidate keyword; based on each standardized recruitment sample, the standardized word frequency of each target keyword is calculated, and based on the standardized word frequency of each target keyword, the basic word frequency matrix is ​​adjusted to obtain a standardized word frequency matrix; based on the standardized word frequency matrix, the quantitative results of each target keyword are identified through a relative importance algorithm, and based on the quantitative results of each target keyword and the standardized word frequency matrix, the level quantitative results corresponding to each soft skill level are identified; soft skill dynamic data of different analysis types are collected, and based on the soft skill dynamic data of each analysis type and the level quantitative results corresponding to each soft skill level, the standardized word frequency matrix is ​​adjusted to obtain a soft skill quantitative framework.

[0111] In this embodiment, Figure 6As shown, the terminal generates a basic word frequency matrix based on the correlation information between each candidate keyword and the initial soft skill framework. Specifically, the terminal counts the frequency of occurrence of the identified soft skill keywords in the recruitment text. The present invention first calculates the original frequency of occurrence of each soft skill keyword in each recruitment text to form a basic word frequency matrix. Compared with the prior art, the present invention uses a more efficient sparse matrix storage method to process large-scale text data, and can effectively process the statistical calculation of hundreds of soft skill keywords in millions of recruitment texts, significantly improving computing efficiency and resource utilization.

[0112] Then, the terminal's basic word frequency matrix includes the target keyword from each candidate keyword; based on each standardized recruitment sample, the standardized word frequency of each target keyword is calculated, and based on the standardized word frequency of each target keyword, the basic word frequency matrix is ​​adjusted to obtain a standardized word frequency matrix. Specifically, this solution takes into account the length differences of different recruitment texts and uses a standardization processing method based on text length in the terminal to eliminate the deviation in the accuracy of the calculated importance information due to the difference in length of the target keyword. The specific implementation method is as follows:

[0113] The terminal calculates the number of words (number of characters) in each recruitment text. Then, the terminal divides the original word frequency of each soft skill keyword by the total number of words in the text to obtain the standardized word frequency. Finally, the terminal adds the standardized word frequency of each target keyword to the basic word frequency matrix to form a standardized word frequency matrix that takes into account the difference in text length. This standardization processing method solves the problem of deviation caused by the influence of text length on simple word frequency statistics in the existing technology, so that the intensity of soft skill requirements in texts of different lengths can be fairly compared. Compared with the existing technology, this standardization method is more scientific and reasonable, and can more accurately reflect the true intensity of soft skill requirements.

[0114] Then, based on the standardized word frequency matrix, the relative importance algorithm is used to identify the quantitative results of each target keyword, and based on the quantitative results of each target keyword and the standardized word frequency matrix, the hierarchical quantitative results corresponding to each soft skill level are identified. Specifically, the terminal calculates the sum of the standardized word frequencies of all soft skill keywords (i.e., target keywords) in each recruitment text. Then, the terminal divides the standardized word frequency of each soft skill keyword by the sum of the standardized word frequencies to obtain the relative word frequency ratio. Finally, the terminal uses each relative word frequency ratio as the quantitative result of each target keyword. Then, based on the constructed multi-level soft skill framework, the terminal hierarchically aggregates the quantitative results at the keyword level to obtain the hierarchical quantitative results corresponding to each soft skill level. For example, (1) Level 3 soft skill quantification: sum up the quantitative results of keywords belonging to the same level 3 soft skill; (2) Level 2 soft skill quantification: sum up the quantitative results of level 3 soft skills belonging to the same level 2 soft skill; (3) Level 1 soft skill quantification: sum up the quantitative results of level 2 soft skills belonging to the same level 1 soft skill.

[0115] The terminal then collects dynamic soft skill data for different analysis types. This includes, but is not limited to, changes in soft skill demand between consecutive quarters, trends in soft skill demand between different years, differences in soft skill demand between different industries, and structural differences in soft skill demand between different positions. Finally, based on the dynamic soft skill data for each analysis type, the terminal adjusts the hierarchical quantification results corresponding to each soft skill level in real time, thereby generating new hierarchical quantification results for each soft skill level and updating the standardized word frequency matrix to obtain a dynamic soft skill quantification framework. This dynamic comparative analysis method can capture the evolution of soft skill demand and market trends, providing forward-looking guidance for businesses, job seekers, and educational institutions. Compared to static analysis methods in existing technologies, this dynamic analysis method is more timely and predictive, helping stakeholders anticipate market development trends and make more informed decisions.

[0116] The terminal can present the soft skills quantification framework to staff through various display methods, thereby improving the practicality and operability of the analysis results. Specific visualization methods and application interfaces include, but are not limited to, soft skills radar charts: intuitively displaying the demand intensity of soft skills at different levels; trend change charts: displaying the temporal trend of soft skills demand; comparative analysis charts: displaying the differences in soft skills demand across different industries and positions; and application programming interfaces: providing standardized data interfaces to support integration with other systems.

[0117] Based on the above solution, the dynamic comparative analysis method can capture the evolution of soft skill demand and market trends, provide forward-looking guidance for enterprises, job seekers and educational institutions, and improve the real-time, dynamic accuracy, and timeliness of the constructed soft skill quantitative framework.

[0118] The application also provides an example of generating a dynamic soft skills quantitative framework based on artificial intelligence technology, such as Figure 7 As shown, the specific processing process includes the following steps:

[0119] Step S701: Acquire current recruitment text data of a large number of users.

[0120] In step S702 , each current recruitment text data is subjected to text normalization processing according to a normalized text processing strategy to obtain each normalized current recruitment text data.

[0121] In step S703, a Chinese word segmentation tool is used to perform Chinese word segmentation processing on each standard current recruitment text data to obtain each standard recruitment text sequence, and stop word filtering processing is performed on each standard recruitment text sequence to obtain each initial recruitment sample data.

[0122] Step S704 , collecting a soft skill dictionary, and filtering abnormal text data in each initial recruitment sample data based on the soft skill dictionary through an abnormal text filtering strategy to obtain each standard recruitment sample.

[0123] In step S705 , based on each standard recruitment sample, high-frequency words are screened by using a high-frequency word extraction strategy to obtain a high-frequency word data set, and soft skill related word data is screened in the high-frequency word data set.

[0124] Step S706 : collecting candidate soft skill keywords from the enterprise side and candidate soft skill keywords from the job seeker side, and identifying market soft skill supply and demand information based on the candidate soft skill keywords from the enterprise side and the job seeker side.

[0125] Step S707 : identifying a set of candidate keywords associated with soft skills based on the soft skill related vocabulary data and the market soft skill supply and demand information.

[0126] Step S708: Filter target candidate keywords corresponding to each standard recruitment sample in the candidate keyword set.

[0127] Step S709, for each standard recruitment sample, based on the target candidate keywords corresponding to the standard recruitment sample, a strategy is defined through a context window to identify the contextual meaning information corresponding to each target candidate keyword, and based on the contextual meaning information corresponding to each target candidate keyword, each polysemous keyword is screened.

[0128] In step S710, all meaning information corresponding to each polysemous keyword is identified through the polysemous keyword processing strategy, and each candidate keyword is divided into keyword categories based on all meaning information corresponding to each polysemous keyword and contextual meaning information corresponding to each non-polysemous keyword.

[0129] In step S711 , the keyword category corresponding to each candidate keyword is used as association information between the candidate keywords.

[0130] Step S712: collect the soft skill related data, and extract the soft skill theory keywords in each soft skill related data, and the soft skill meaning information corresponding to each soft skill theory keyword.

[0131] Step S713 : Based on the soft skill meaning information corresponding to each soft skill theory keyword, the hierarchical relationship between each soft skill theory keyword is identified through the soft skill hierarchical model.

[0132] Step S714: Based on the hierarchical relationship between each soft skill theory keyword, an initial soft skill framework including all soft skill theory keywords is constructed through an association graph construction strategy.

[0133] Step S715 : generating a basic word frequency matrix based on the association information between the candidate keywords and the initial soft skill framework.

[0134] Step S716: Calculate the normalized word frequency of each target keyword based on each standard recruitment sample, and adjust the basic word frequency matrix based on the normalized word frequency of each target keyword to obtain a normalized word frequency matrix.

[0135] Step S717 : Based on the normalized word frequency matrix, the quantitative results of each target keyword are identified by a relative importance algorithm, and based on the quantitative results of each target keyword and the normalized word frequency matrix, the level quantitative results corresponding to each soft skill level are identified.

[0136] Step S718: collect dynamic data of soft skills of different analysis types, and adjust the standardized word frequency matrix based on the dynamic data of soft skills of each analysis type and the level quantification results corresponding to each soft skill level to obtain a soft skill quantification framework.

[0137] It should be understood that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence according to the instructions of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above can include multiple steps or multiple stages, and these steps or stages are not necessarily executed and completed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0138] Based on the same inventive concept, the embodiments of the present application also provide a dynamic soft skill quantification framework generation device based on artificial intelligence technology for implementing the above-mentioned dynamic soft skill quantification framework generation method based on artificial intelligence technology. The implementation solution provided by this device is similar to the implementation solution described in the above-mentioned method. Therefore, the specific limitations of the one or more embodiments of the dynamic soft skill quantification framework generation device based on artificial intelligence technology provided below can be referred to the limitations of the dynamic soft skill quantification framework generation method based on artificial intelligence technology above, and will not be repeated here.

[0139] In an exemplary embodiment, Figure 8 As shown, a dynamic soft skill quantification framework generation device based on artificial intelligence technology is provided, including: an acquisition module 810, an identification module 820 and a generation module 830, wherein:

[0140] An acquisition module 810 is used to acquire current recruitment text data of a large number of users and perform text preprocessing on the current recruitment text data to obtain various standardized recruitment samples;

[0141] An identification module 820 is configured to screen a set of candidate keywords associated with soft skills in each of the standard recruitment samples, and identify association information between each of the candidate keywords based on the set of candidate keywords and each of the standard recruitment samples;

[0142] The generation module 830 is configured to collect an initial soft skill framework and generate a soft skill quantitative framework based on the initial soft skill framework and the association information between the candidate keywords through a soft skill quantitative analysis strategy.

[0143] Optionally, the acquisition module 810 is specifically configured to:

[0144] Performing text normalization processing on each of the current recruitment text data according to a normalized text processing strategy to obtain each standardized current recruitment text data;

[0145] Using a Chinese word segmentation tool, perform Chinese word segmentation processing on the current recruitment text data of each standard to obtain each standard recruitment text sequence, and perform stop word filtering processing on each standard recruitment text sequence to obtain each initial recruitment sample data;

[0146] A soft skill dictionary is collected, and based on the soft skill dictionary, abnormal text data in each of the initial recruitment sample data is filtered through an abnormal text filtering strategy to obtain each of the standard recruitment samples.

[0147] Optionally, the identification module 820 is specifically configured to:

[0148] Based on the standard recruitment samples, high-frequency words are screened by using a high-frequency word extraction strategy to obtain a high-frequency word data set, and soft skill-related word data is screened in the high-frequency word data set;

[0149] Collecting candidate soft skill keywords from the enterprise side and candidate soft skill keywords from the job seeker side, and identifying market soft skill supply and demand information based on the candidate soft skill keywords from the enterprise side and the candidate soft skill keywords from the job seeker side;

[0150] Based on the soft skill related vocabulary data and the market soft skill supply and demand information, a candidate keyword set associated with soft skills is identified.

[0151] Optionally, the identification module 820 is specifically configured to:

[0152] In the candidate keyword set, screening target candidate keywords corresponding to each standard recruitment sample;

[0153] For each standard recruitment sample, based on each target candidate keyword corresponding to the standard recruitment sample, a strategy is defined through a context window to identify contextual meaning information corresponding to each target candidate keyword, and based on the contextual meaning information corresponding to each target candidate keyword, each polysemous keyword is screened;

[0154] Using an ambiguous keyword processing strategy, all meaning information corresponding to each ambiguous keyword is identified, and based on all meaning information corresponding to each ambiguous keyword and contextual meaning information corresponding to each non-ambiguous keyword, each candidate keyword is divided into keyword categories;

[0155] The keyword category corresponding to each candidate keyword is used as the association information between the candidate keywords.

[0156] Optionally, the generating module 830 is specifically configured to:

[0157] Collecting data related to each soft skill, and extracting soft skill theory keywords from each soft skill related data, as well as soft skill meaning information corresponding to each soft skill theory keyword;

[0158] Based on the soft skill meaning information corresponding to each soft skill theory keyword, the hierarchical relationship between each soft skill theory keyword is identified through the soft skill hierarchical model;

[0159] Based on the hierarchical relationship between each soft skill theory keyword, an association graph construction strategy is used to build an initial soft skill framework that includes all soft skill theory keywords.

[0160] Optionally, the generating module 830 is specifically configured to:

[0161] Based on the association information between the candidate keywords and the initial soft skill framework, a basic word frequency matrix is ​​generated; the basic word frequency matrix includes the target keyword from the candidate keywords;

[0162] Based on each of the standardized recruitment samples, the standardized word frequency of each of the target keywords is calculated, and based on the standardized word frequency of each of the target keywords, the basic word frequency matrix is ​​adjusted to obtain a standardized word frequency matrix;

[0163] Based on the standardized word frequency matrix, identifying the quantitative results of each target keyword through a relative importance algorithm, and based on the quantitative results of each target keyword and the standardized word frequency matrix, identifying the level quantitative results corresponding to each soft skill level;

[0164] Soft skill dynamic data of different analysis types are collected, and based on the soft skill dynamic data of each analysis type and the level quantification results corresponding to each soft skill level, the standardized word frequency matrix is ​​adjusted to obtain a soft skill quantification framework.

[0165] Each module in the aforementioned AI-based dynamic soft skills quantification framework generation device can be implemented in whole or in part through software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in hardware form, or stored in a computer device memory in software form, so that the processor can call and execute the corresponding operations of each module.

[0166] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 9As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected via a system bus, and the communication interface, the display unit and the input device are connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, a mobile cellular network, near field communication (NFC) or other technologies. When the computer program is executed by the processor, a dynamic soft skill quantification framework generation method based on artificial intelligence technology is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device casing, or an external keyboard, touchpad or mouse.

[0167] Those skilled in the art will understand that Figure 9 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0168] In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps corresponding to the method for generating a dynamic soft skill quantification framework based on artificial intelligence technology.

[0169] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the computer program implements the steps corresponding to the method for generating a dynamic soft skill quantification framework based on artificial intelligence technology.

[0170] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements steps corresponding to a method for generating a dynamic soft skill quantification framework based on artificial intelligence technology.

[0171] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0172] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile memory and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processor involved in the various embodiments provided herein may be, but are not limited to, a general-purpose processor, a central processing unit (CPU), a graphics processing unit (GPU), a digital signal processor (DSP), a programmable logic unit (PLC), a data processing logic unit based on quantum computing, an artificial intelligence (AI) processor, and the like.

[0173] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0174] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A method for generating a dynamic soft skills quantitative framework based on artificial intelligence technology, characterized in that: The method comprises: Acquire current recruitment text data of a large number of users, and perform text preprocessing on the current recruitment text data to obtain standardized recruitment samples; the users include enterprises and individuals; In each of the standard recruitment samples, a candidate keyword set associated with soft skills is screened, and based on the candidate keyword set and each of the standard recruitment samples, association information between the candidate keywords is identified; An initial soft skill framework is collected, and based on the initial soft skill framework and the association information between the candidate keywords, a soft skill quantitative analysis strategy is used to generate a soft skill quantitative framework.

2. The method according to claim 1, characterized in that The text preprocessing of the current recruitment text data to obtain various standard recruitment samples includes: Performing text normalization processing on each of the current recruitment text data according to a normalized text processing strategy to obtain each standardized current recruitment text data; Using a Chinese word segmentation tool, perform Chinese word segmentation processing on the current recruitment text data of each standard to obtain each standard recruitment text sequence, and perform stop word filtering processing on each standard recruitment text sequence to obtain each initial recruitment sample data; A soft skill dictionary is collected, and based on the soft skill dictionary, abnormal text data in each of the initial recruitment sample data is filtered through an abnormal text filtering strategy to obtain each of the standard recruitment samples.

3. The method according to claim 1, characterized in that In each of the standard recruitment samples, a candidate keyword set associated with soft skills is screened, including: Based on the standard recruitment samples, high-frequency words are screened by using a high-frequency word extraction strategy to obtain a high-frequency word data set, and soft skill-related word data is screened in the high-frequency word data set; Collecting candidate soft skill keywords from the enterprise side and candidate soft skill keywords from the job seeker side, and identifying market soft skill supply and demand information based on the candidate soft skill keywords from the enterprise side and the candidate soft skill keywords from the job seeker side; Based on the soft skill related vocabulary data and the market soft skill supply and demand information, a candidate keyword set associated with soft skills is identified.

4. The method according to claim 1, wherein The step of identifying association information between the candidate keywords based on the candidate keyword set and the standard recruitment samples includes: In the candidate keyword set, screening target candidate keywords corresponding to each standard recruitment sample; For each standard recruitment sample, based on each target candidate keyword corresponding to the standard recruitment sample, a strategy is defined through a context window to identify contextual meaning information corresponding to each target candidate keyword, and based on the contextual meaning information corresponding to each target candidate keyword, each polysemous keyword is screened; Using an ambiguous keyword processing strategy, all meaning information corresponding to each ambiguous keyword is identified, and based on all meaning information corresponding to each ambiguous keyword and contextual meaning information corresponding to each non-ambiguous keyword, each candidate keyword is divided into keyword categories; The keyword category corresponding to each candidate keyword is used as the association information between the candidate keywords.

5. The method according to claim 1, wherein The initial soft skills framework includes: Collecting data related to each soft skill, and extracting soft skill theory keywords from each soft skill related data, as well as soft skill meaning information corresponding to each soft skill theory keyword; Based on the soft skill meaning information corresponding to each soft skill theory keyword, the hierarchical relationship between each soft skill theory keyword is identified through the soft skill hierarchical model; Based on the hierarchical relationship between each soft skill theory keyword, an association graph construction strategy is used to build an initial soft skill framework that includes all soft skill theory keywords.

6. The method according to claim 1, characterized in that The step of generating a soft skill quantification framework based on the initial soft skill framework and the association information between the candidate keywords through a soft skill quantification analysis strategy includes: Based on the association information between the candidate keywords and the initial soft skill framework, a basic word frequency matrix is ​​generated; the basic word frequency matrix includes the target keyword from the candidate keywords; Based on each of the standardized recruitment samples, the standardized word frequency of each of the target keywords is calculated, and based on the standardized word frequency of each of the target keywords, the basic word frequency matrix is ​​adjusted to obtain a standardized word frequency matrix; Based on the standardized word frequency matrix, identifying the quantitative results of each target keyword through a relative importance algorithm, and based on the quantitative results of each target keyword and the standardized word frequency matrix, identifying the level quantitative results corresponding to each soft skill level; Soft skill dynamic data of different analysis types are collected, and based on the soft skill dynamic data of each analysis type and the level quantification results corresponding to each soft skill level, the standardized word frequency matrix is ​​adjusted to obtain a soft skill quantification framework.

7. A dynamic soft skills quantitative framework generation device based on artificial intelligence technology, characterized by: The device comprises: An acquisition module is used to acquire current recruitment text data of a large number of users and perform text preprocessing on the current recruitment text data to obtain various standardized recruitment samples; an identification module for screening a set of candidate keywords associated with soft skills in each of the standard recruitment samples, and identifying association information between each of the candidate keywords based on the set of candidate keywords and each of the standard recruitment samples; The generation module is used to collect an initial soft skill framework and generate a soft skill quantitative framework based on the initial soft skill framework and the association information between the candidate keywords through a soft skill quantitative analysis strategy.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images