Intelligent employment planning method and system based on fusion data analysis

Through the integration of data analysis technology, including keyword extraction, synonym replacement and semantic encoding, the semantic similarity between employment information and job seeker information is solved, and the problem of insufficient information coverage in the existing technology is achieved, and the precise matching of employment information is achieved and personalized push.

CN119066265AInactive Publication Date: 2024-12-03GUANGXI NORMAL UNIV +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411080669.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-08
Publication Date
2024-12-03
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing employment information push method only uses basic keyword extraction technology, which may miss other similar expressions related to the position, lack of in-depth semantic analysis, resulting in insufficient information coverage and inability to achieve accurate matching, resulting in job seekers not receiving it. Related job recommendations.

Method used

Using a smart employment planning method based on fusion data analysis, the semantic similarity between employment information and job seeker information is calculated through keyword extractors, synonyms replacement, semantic coding and feature aggregation networks, and the semantic similarity between employment information and job seeker information is automatically judged whether to push employment information to ensure that Relevance and personalization of information.

Benefits of technology

Improve the matching accuracy between job seekers, ensuring that the employment information pushed to job seekers is more relevant and personalized, and meets the specific needs of job seekers.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119066265A_ABST
    Figure CN119066265A_ABST
Patent Text Reader

Abstract

The invention relates to the field of intelligent data analysis, in particular to an intelligent employment planning method and system based on fusion data analysis. The method comprises the following steps: extracting first employment information from an employment information database, obtaining job seeker information, performing keyword extraction synonym replacement and semantic understanding on the employment information by adopting an artificial intelligence-based information processing and analysis technology, performing semantic coding on the job seeker information, and performing semantic coding on the job seeker information; therefore, whether the employment information is pushed to the job seeker or not is automatically judged according to the comparison between the similarity between the semantic features of the two and a preset threshold value. Through the mode, key synonym replacement and deep semantic understanding can be performed on the employment information so as to more accurately grasp the core requirements of the employment information, and the matching precision between the position and the job seeker is improved through a mode of performing similarity measurement and matching with the semantic information of the job seeker. And the employment information pushed to the job seeker is ensured to be more relevant and personalized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of intelligent data analysis, and specifically to a smart employment planning method and system based on fusion data analysis. Background Art

[0002] With the continuous advancement of information technology, the informatization of the employment market is also moving forward, giving rise to many online employment information platforms. These platforms have accumulated a large amount of employer recruitment information. In order to help job seekers quickly locate their ideal positions, the platforms provide search functions based on keywords such as cities and job types to match information and display results. However, the existing search results are not optimized and sorted, and the specific needs of users are not fully considered, resulting in job seekers being unable to quickly identify the most relevant job information. As a result, users have to browse through a long list of positions one by one, which greatly reduces the speed and efficiency of finding a satisfactory job.

[0003] In response to the above technical problems, China's public patent CN114254198A proposes a method for pushing employment information, which analyzes the behavioral data of job seekers, sorts the keywords by importance, and prioritizes the keywords with the highest weight to screen employment information. In addition, it also includes querying the keyword index table based on employment information to improve search efficiency, and also considers the job seeker's status information, such as credit points, job qualifications and education level, to push qualified jobs and improve job search efficiency.

[0004] However, the above-mentioned employment information push method only uses basic keyword extraction technology for matching, which may miss other similar expressions related to the position, which limits the coverage of the information. In addition, there is a lack of in-depth semantic analysis of employment information and job seeker information. The keyword screening method easily ignores the different meanings of keywords in different contexts, thereby affecting the accurate understanding of employment information and job search information, making it impossible to achieve accurate matching, and causing job seekers to receive irrelevant job recommendations.

[0005] Therefore, an intelligent employment planning method based on fusion data analysis is desired. Summary of the invention

[0006] This application is made in consideration of the above problems. One purpose of this application is to provide a smart employment planning method and system based on fusion data analysis.

[0007] The embodiment of the present application provides a smart employment planning method based on fusion data analysis, which includes:

[0008] Extracting first employment information from an employment information database;

[0009] Inputting the first employment information into a keyword extractor to obtain a keyword extraction result;

[0010] Based on the keyword extraction result, performing keyword synonym replacement on the first employment information to obtain a set of synonymous expressions of the first employment information;

[0011] Performing semantic coding on the set of synonymous expressions of the first employment information and the first employment information to obtain a set of semantic coding feature vectors of synonymous expressions of the first employment information and a semantic coding feature vector of the first employment information;

[0012] Inputting the set of semantic encoding feature vectors of the synonymous expression of the first employment information and the semantic encoding feature vector of the first employment information into a feature aggregation network based on feature local energy distribution gating to obtain a first semantic aggregation representation vector of the employment information;

[0013] Obtain job applicant information;

[0014] Performing semantic encoding on the job seeker information to obtain a semantic encoding feature vector of the job seeker information;

[0015] Calculating the cosine similarity between the semantic aggregation representation vector of the first employment information and the semantic encoding feature vector of the job seeker information;

[0016] Based on the comparison between the cosine similarity and a preset threshold, it is determined whether to push the first employment information to the job seeker.

[0017] For example, according to the intelligent employment planning method based on fusion data analysis of an embodiment of the present application, wherein the first employment information is input into a keyword extractor to obtain a keyword extraction result, including: inputting the first employment information into a keyword extractor based on an AIGC model to obtain the keyword extraction result.

[0018] For example, according to the intelligent employment planning method based on fusion data analysis of an embodiment of the present application, the set of semantic encoding feature vectors of the synonymous expression of the first employment information and the first employment information semantic encoding feature vector are input into a feature aggregation network based on feature local energy distribution gating to obtain the first employment information semantic aggregation representation vector, including:

[0019] Determine, based on the maximum value, minimum value, mean value and variance of each first employment information synonymous expression semantic coding feature vector in the set of the first employment information synonymous expression semantic coding feature vectors, the characteristic energy level coefficients of each first employment information synonymous expression semantic coding feature vector to obtain a set of first employment information synonymous expression semantic coding feature energy level coefficients;

[0020] Determine the scale of the local neighborhood, calculate the local neighborhood significance descriptor of each first employment information proximity expression semantic coding feature energy level coefficient in the set of the first employment information proximity expression semantic coding feature energy level coefficients to obtain a set of first employment information proximity expression semantic coding feature energy level significance descriptors;

[0021] Inputting the set of energy level significance descriptors of semantic coding features of synonymous expressions of the first employment information into a mask module based on a gating function to obtain a set of mask-gated probabilistic energy level significance descriptors of semantic coding features of synonymous expressions of the first employment information;

[0022] Taking the set of the mask-gated probabilistic semantic encoding feature energy level significance descriptors of the first employment information proximate expression as a weighted sequence, calculating the position-weighted sum of the set of the first employment information proximate expression semantic encoding feature vectors to obtain the first employment information proximate expression semantic feature energy level guided aggregation representation vector;

[0023] The first employment information synonymous expression semantic feature energy level guided aggregation representation vector and the first employment information semantic encoding feature vector are cascaded to obtain the first employment information semantic aggregation representation vector.

[0024] For example, according to the intelligent employment planning method based on fusion data analysis of an embodiment of the present application, based on the maximum value, minimum value, mean value and variance of each first employment information synonymous expression semantic coding feature vector in the set of the first employment information synonymous expression semantic coding feature vector, the characteristic energy level coefficient of each first employment information synonymous expression semantic coding feature vector is determined to obtain a set of first employment information synonymous expression semantic coding feature energy level coefficients, including:

[0025] Respectively extracting the maximum value and the minimum value of each position in the semantic coding feature vector of the synonymous expression of the first employment information to obtain the maximum value of the semantic coding feature of the synonymous expression of the first employment information and the minimum value of the semantic coding feature of the synonymous expression of the first employment information;

[0026] Respectively calculating the mean and variance of the semantic coding feature vector of the synonymous expression of the first employment information to obtain the mean of the semantic coding feature of the synonymous expression of the first employment information and the variance of the semantic coding feature of the synonymous expression of the first employment information;

[0027] Adding the variance of the semantic coding feature of the synonymous expression of the first employment information and a preset hyperparameter to obtain a first energy level coefficient of the semantic coding feature of the synonymous expression of the first employment information;

[0028] Calculate the square of the difference between the maximum value of the semantic coding feature of the synonymous expression of the first employment information and the mean value of the semantic coding feature of the synonymous expression of the first employment information to obtain the maximum difference value of the semantic coding of the synonymous expression of the first employment information;

[0029] Calculate the square of the difference between the minimum value of the semantic coding feature of the synonymous expression of the first employment information and the mean value of the semantic coding feature of the synonymous expression of the first employment information to obtain the minimum difference value of the semantic coding of the synonymous expression of the first employment information;

[0030] The maximum difference value of the semantic coding of the synonymous expression of the first employment information is added to the minimum difference value of the semantic coding of the synonymous expression of the first employment information and the preset hyperparameter to obtain the second energy level coefficient of the semantic coding feature of the synonymous expression of the first employment information;

[0031] The first energy level coefficient of the semantic coding feature of the synonymous expression of the first employment information is divided by the second energy level coefficient of the semantic coding feature of the synonymous expression of the first employment information to obtain the energy level coefficient of the semantic coding feature of the synonymous expression of the first employment information.

[0032] For example, according to the intelligent employment planning method based on fusion data analysis of an embodiment of the present application, the set of semantic encoding feature energy level significance descriptors of the synonymous expression of the first employment information is input into a mask module based on a gating function to obtain a mask gated probabilistic set of semantic encoding feature energy level significance descriptors of the synonymous expression of the first employment information, including:

[0033] Normalizing each first employment information synonymous expression semantic coding feature energy level significance descriptor in the set of the first employment information synonymous expression semantic coding feature energy level significance descriptors to obtain a set of normalized first employment information synonymous expression semantic coding feature energy level significance descriptors;

[0034] Each normalized first employment information proximate expression semantic coding feature energy level significance descriptor in the set of normalized first employment information proximate expression semantic coding feature energy level significance descriptors is input into a gating function for masking processing to obtain the set of mask-gated probabilistic first employment information proximate expression semantic coding feature energy level significance descriptors.

[0035] For example, according to the intelligent employment planning method based on fusion data analysis of the embodiment of the present application, each normalized first employment information synonymous expression semantic coding feature energy level significance descriptor in the set of normalized first employment information synonymous expression semantic coding feature energy level significance descriptors is input into the gating function for masking processing to obtain the set of masked gated probabilistic first employment information synonymous expression semantic coding feature energy level significance descriptors, including: in response to the normalized first employment information synonymous expression semantic coding feature energy level significance descriptors in the set of normalized first employment information synonymous expression semantic coding feature energy level significance descriptors being greater than a predetermined threshold, the normalized first employment information synonymous expression semantic coding feature energy level significance descriptors greater than the predetermined threshold are set to the original value, otherwise they are set to zero to obtain the set of masked gated probabilistic first employment information synonymous expression semantic coding feature energy level significance descriptors.

[0036] For example, according to the intelligent employment planning method based on fusion data analysis of the embodiment of the present application, it also includes a training step: used to train the keyword extractor based on the AIGC model and the feature aggregation network based on feature local energy distribution gating.

[0037] For example, according to the intelligent employment planning method based on fusion data analysis of an embodiment of the present application, the training step includes:

[0038] Acquire training data, wherein the training data includes first training employment information, training job seeker information, and true cosine similarity extracted from an employment information database;

[0039] Inputting the first training employment information into the keyword extractor based on the AIGC model to obtain a training keyword extraction result;

[0040] Based on the training keyword extraction result, performing keyword synonym replacement on the first training employment information to obtain a set of synonymous expressions of the first training employment information;

[0041] Performing semantic coding on the set of synonymous expressions of the first training employment information and the first training employment information to obtain a set of semantic coding feature vectors of synonymous expressions of the first training employment information and a semantic coding feature vector of the first training employment information;

[0042] Inputting the set of semantic encoding feature vectors of the first training employment information synonymous expression and the first training employment information semantic encoding feature vector into the feature aggregation network based on feature local energy distribution gating to obtain the first training employment information semantic aggregation representation vector;

[0043] Performing semantic encoding on the training job seeker information to obtain a semantic encoding feature vector of the training job seeker information;

[0044] Calculating the training cosine similarity between the first training employment information semantic aggregation representation vector and the training job seeker information semantic encoding feature vector;

[0045] Calculating the difference between the training cosine similarity and the true cosine similarity as a difference loss function value;

[0046] Calculating the employment and job-seeking information constraint loss function value between the first training employment information semantic aggregation representation vector and the training job seeker information semantic encoding feature vector;

[0047] The keyword extractor based on the AIGC model and the feature aggregation network based on feature local energy distribution gating are trained through gradient back propagation based on the weighted sum of the difference loss function value and the job search information constraint loss function value as the loss function value.

[0048] The embodiment of the present application also provides a smart employment planning system based on fusion data analysis, which includes:

[0049] An employment information extraction module, used to extract first employment information from an employment information database;

[0050] A keyword extraction module, used for inputting the first employment information into a keyword extractor to obtain a keyword extraction result;

[0051] A keyword synonym replacement module, used for performing keyword synonym replacement on the first employment information based on the keyword extraction result to obtain a set of synonymous expressions of the first employment information;

[0052] An employment information semantic coding module, configured to semantically code the set of synonymous expressions of the first employment information and the first employment information to obtain a set of semantic coding feature vectors of synonymous expressions of the first employment information and a semantic coding feature vector of the first employment information;

[0053] A feature aggregation module, used for inputting the set of semantic encoding feature vectors of the synonymous expression of the first employment information and the first semantic encoding feature vector of the employment information into a feature aggregation network based on feature local energy distribution gating to obtain a first semantic aggregation representation vector of the employment information;

[0054] A job search information acquisition module is used to obtain job seeker information;

[0055] A job-seeking information semantic encoding module, used for semantically encoding the job-seeker information to obtain a semantic encoding feature vector of the job-seeker information;

[0056] A similarity calculation module, used to calculate the cosine similarity between the first employment information semantic aggregation representation vector and the job seeker information semantic encoding feature vector;

[0057] A push module is used to determine whether to push the first employment information to the job seeker based on the comparison between the cosine similarity and a preset threshold.

[0058] For example, according to the intelligent employment planning system based on fusion data analysis of the embodiment of the present application, the keyword extraction module is used to: input the first employment information into a keyword extractor based on the AIGC model to obtain the keyword extraction result.

[0059] According to the intelligent employment planning method and system based on fusion data analysis of the embodiments of the present application, it is possible to replace key synonyms and perform deep semantic understanding of employment information to more accurately grasp the core requirements of employment information, and improve the matching accuracy between positions and job seekers by measuring and matching similarity with job seekers' semantic information, thereby ensuring that the employment information pushed to job seekers is more relevant and personalized. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings of the embodiments of the present application are briefly introduced below. Obviously, the drawings described below only relate to some embodiments of the present application, and are not intended to limit the present application.

[0061] Figure 1 A schematic diagram of the application architecture of the intelligent employment planning method based on fusion data analysis in an embodiment of the present application is shown;

[0062] Figure 2 A flow chart of a smart employment planning method based on fusion data analysis in an embodiment of the present application is shown;

[0063] Figure 3 A flowchart of sub-step S550 of the intelligent employment planning method based on fusion data analysis in an embodiment of the present application is shown;

[0064] Figure 4 A schematic diagram of the structure of an intelligent employment planning system based on fusion data analysis in an embodiment of the present application is shown; and

[0065] Figure 5 An application scenario diagram of the intelligent employment planning method based on fusion data analysis in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0066] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present application.

[0067] The terms used in this specification are those common terms currently widely used in the art in consideration of the functions of the present application, but these terms may vary according to the intentions of those skilled in the art, precedents, or new technologies in the art. In addition, specific terms may be selected, and in this case, their detailed meanings will be described in the detailed description of the present application. Therefore, the terms used in the specification should not be understood as simple names, but rather as a general description based on the meaning of the terms and the present application.

[0068] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0069] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. At the same time, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0070] Figure 1 A schematic diagram of the application architecture of an intelligent employment planning method based on fusion data analysis in an embodiment of the present application is shown, including a server 100 and a terminal device 200.

[0071] The terminal device 200 and the server 100 can be connected via the Internet to achieve mutual communication. Optionally, the above-mentioned Internet uses standard communication technology and / or protocol. The Internet is usually the Internet, but it can also be any network, including but not limited to any combination of a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a mobile, wired or wireless network, a dedicated network or a virtual private network. In some embodiments, the data exchanged through the network is represented by technologies and / or formats including Hyper Text Markup Language (HTML), Extensible Markup Language (XML), etc. In addition, conventional encryption technologies such as Secure Socket Layer (SSL), Transport Layer Security (TLS), Virtual Private Network (VPN), Internet Protocol Security (IPsec) can also be used to encrypt all or some links. In other embodiments, customized and / or dedicated data communication technologies can also be used to replace or supplement the above-mentioned data communication technologies.

[0072] The server 100 can provide various network services for the terminal device 200, wherein the server 100 can be a single server, a server cluster consisting of several servers, or a cloud computing center. Specifically, the server 100 may include a processor 110 (Center Processing Unit, CPU), a memory 120, an input device 130, and an output device 140, etc. The input device 130 may include a keyboard, a mouse, a touch screen, etc., and the output device 140 may include a display device, such as a liquid crystal display (Liquid Crystal Display, LCD), a cathode ray tube (Cathode Ray Tube, CRT), etc.

[0073] The memory 120 may include a read-only memory (ROM) and a random access memory (RAM), and provides the processor 110 with program instructions and data stored in the memory 120. In an embodiment of the present application, the memory 120 may be used to store the program of the intelligent employment planning method based on fusion data analysis in an embodiment of the present application.

[0074] The processor 110 calls the program instructions stored in the memory 120, and the processor 110 is used to execute the steps of any one of the intelligent employment planning methods based on fusion data analysis in the embodiments of the present application according to the obtained program instructions.

[0075] In addition, the application architecture diagram in the embodiment of the present application is intended to more clearly illustrate the technical solution in the embodiment of the present application, and does not constitute a limitation on the technical solution provided in the embodiment of the present application. Of course, for other application architectures and business applications, the technical solution provided in the embodiment of the present application is also applicable to similar problems.

[0076] The following is a non-restrictive explanation of the intelligent employment planning method based on fusion data analysis provided according to at least one embodiment of the present application through several examples or embodiments. As described below, different features in these specific examples or embodiments can be combined with each other without conflicting with each other to obtain new examples or embodiments, and these new examples or embodiments also fall within the scope of protection of the present application.

[0077] In response to the above technical problems, the technical concept of this application is to extract the first employment information from the employment information database, obtain the job seeker information, and use information processing and analysis technology based on artificial intelligence to perform keyword extraction, synonym replacement and semantic understanding on the employment information, and semantic encoding on the job seeker information, so as to automatically determine whether to push the employment information to the job seeker based on the comparison between the similarity between the semantic features of the two and the preset threshold. In this way, it is possible to replace key synonyms and deeply understand the semantics of the employment information to more accurately grasp the core requirements of the employment information, and improve the matching accuracy between the position and the job seeker by measuring and matching the similarity with the job seeker's semantic information, ensuring that the employment information pushed to the job seeker is more relevant and personalized.

[0078] Figure 2 The flowchart of the intelligent employment planning method based on fusion data analysis in the embodiment of the present application is shown. For example, the intelligent employment planning method based on fusion data analysis can be executed by a server, which can be Figure 1 The server 100 shown in FIG. Figure 2As shown, according to the embodiment of the present application, the intelligent employment planning method based on fusion data analysis includes the following steps: S510, extracting first employment information from an employment information database; S520, inputting the first employment information into a keyword extractor to obtain a keyword extraction result; S530, based on the keyword extraction result, performing keyword synonym replacement on the first employment information to obtain a set of synonymous expressions of the first employment information; S540, semantically encoding the set of synonymous expressions of the first employment information and the first employment information to obtain a set of semantic encoding feature vectors of the first employment information synonymous expressions and a semantic encoding feature vector of the first employment information; S550 , input the set of semantic coding feature vectors of synonymous expressions of the first employment information and the semantic coding feature vector of the first employment information into a feature aggregation network based on feature local energy distribution gating to obtain a first semantic aggregation representation vector of employment information; S560, obtain job seeker information; S570, semantically encode the job seeker information to obtain a semantic coding feature vector of job seeker information; S580, calculate the cosine similarity between the first semantic aggregation representation vector of employment information and the semantic coding feature vector of job seeker information; S590, determine whether to push the first employment information to the job seeker based on a comparison between the cosine similarity and a preset threshold.

[0079] Specifically, in the technical solution of the present application, first, the first employment information is extracted from the employment information database. Next, considering that the first employment information may contain keyword information such as job description, job requirements, skill requirements, work experience, educational background, salary range, etc., these key information are crucial for job seekers. Based on this, in the technical solution of the present application, the first employment information is input into a keyword extractor based on the AIGC model to obtain a keyword extraction result. It is worth mentioning that the use of the keyword extractor based on the AIGC model can accurately identify and extract key words related to the employment information from the first employment information, providing support for subsequent intelligent recommendations.

[0080] Accordingly, in step S520, the first employment information is input into a keyword extractor to obtain a keyword extraction result, including: inputting the first employment information into a keyword extractor based on an AIGC model to obtain the keyword extraction result.

[0081] Then, it is considered that different job seekers or employment information may use different terms or expressions to describe the same concept or skill. Therefore, in order to enhance the system's ability to understand different language expressions and the possible differences in the use of terms between job seeker information and employment information, in the technical solution of the present application, based on the keyword extraction results, the first employment information is subjected to keyword synonym replacement to obtain a set of synonymous expressions of the first employment information. In other words, by using synonyms, the scope of search and matching of employment information can be expanded, and it can be not only limited to the original keywords, but also include other words with similar meanings to the keywords, avoiding the omission of employment information that uses different but related terms to describe similar positions or skills, thereby ensuring that they can search for relevant position information even if they use different words, so as to have a more comprehensive understanding of employment information and more accurate position matching.

[0082] Next, considering that the set of synonymous expressions of the first employment information and each keyword and employment information in the first employment information have contextual semantic association influence relationships regarding the employment information, based on this, in order to more clearly understand the contextual semantic association information of the synonymous expressions of the first employment information and the first employment information, in the technical solution of the present application, the set of synonymous expressions of the first employment information and the first employment information are semantically encoded to respectively capture and extract the core semantic content of each information and the semantic association between the context to obtain the set of semantic encoding feature vectors of the synonymous expressions of the first employment information and the semantic encoding feature vector of the first employment information.

[0083] Then, considering that the set of semantic encoding feature vectors of the synonymous expression of the first employment information contains shallow statistical feature information and semantic information of the synonymous expression under the deep employment information. The semantic encoding feature vector of the first employment information expresses the semantic information contained in the original text. Therefore, in order to provide a comprehensive, multi-dimensional semantic feature representation of the first employment information to more comprehensively understand the semantic content of the employment information, in the technical solution of the present application, further, the set of semantic encoding feature vectors of the synonymous expression of the first employment information and the first semantic encoding feature vector of the employment information are input into a feature aggregation network based on feature local energy distribution gating to obtain the first employment information semantic aggregation representation vector. It is worth mentioning that the feature aggregation network based on feature local energy distribution gating first analyzes the maximum value, minimum value, mean and variance of each vector in the set of semantic encoding feature vectors of the synonymous expression of the first employment information. These statistics are used to calculate the feature energy level coefficient, quantify the energy distribution of the feature and evaluate the energy level of the feature. Then, the scale of the local neighborhood is determined, which is to consider the position and importance of each feature vector in the context of the set of semantic encoding feature vectors of the synonymous expression of the first employment information. Next, the local neighborhood significance descriptor of each feature energy level coefficient is calculated, which helps to capture the energy distribution characteristics of the semantic coding features of the synonymous expression of the first employment information in the local neighborhood, so as to identify the semantic features that are significant in the local range. After that, the significance descriptor is input into the mask module based on the gating function, and the dynamic adjustment of the significance of the semantic features of the synonymous expression of the first employment information is realized through nonlinear probabilistic transformation and adaptive feature screening. That is, through the gating function and the masking function, the probabilistic significance descriptors are obtained, and these descriptors are used as weights to reflect the importance of each semantic feature of the synonymous expression of the first employment information in the aggregation process. Next, the weighted sum of the semantic feature vector set of the synonymous expression of the first employment information is calculated using the probabilistic significance descriptor as the weight to strengthen the significant features of the local synonymous semantic expression and suppress unimportant features. Finally, the energy level guided aggregation representation vector that integrates the shallow statistical features and the deep synonymous expression semantic features is cascaded with the first employment information semantic coding feature vector to obtain a more comprehensive and rich first employment information semantic aggregation representation vector.

[0084] Accordingly, in step S550, if Figure 3As shown, the set of semantic coding feature vectors of the first employment information synonymous expression and the first employment information semantic coding feature vector are input into a feature aggregation network based on feature local energy distribution gating to obtain a first employment information semantic aggregation representation vector, including: S551, based on the maximum value, minimum value, mean value and variance of each first employment information synonymous expression semantic coding feature vector in the set of the first employment information synonymous expression semantic coding feature vector, determining the feature energy level coefficient of each first employment information synonymous expression semantic coding feature vector to obtain a set of first employment information synonymous expression semantic coding feature energy level coefficients; S552, determining the scale of the local neighborhood, calculating the local neighborhood significance descriptor of each first employment information synonymous expression semantic coding feature energy level coefficient in the set of the first employment information synonymous expression semantic coding feature energy level coefficients to obtain the first employment information synonymous expression semantic coding feature energy level coefficients. S553, inputting the set of semantic coding feature energy level significance descriptors of the synonymous expression of the first employment information into a mask module based on a gating function to obtain a mask-gated probabilistic set of semantic coding feature energy level significance descriptors of the first employment information synonymous expression; S554, using the mask-gated probabilistic set of semantic coding feature energy level significance descriptors of the first employment information synonymous expression as a sequence of weights, calculating the position-weighted sum of the set of semantic coding feature vectors of the first employment information synonymous expression to obtain the first employment information synonymous expression semantic feature energy level guided aggregation representation vector; S555, cascading the first employment information synonymous expression semantic feature energy level guided aggregation representation vector and the first employment information semantic coding feature vector to obtain the first employment information semantic aggregation representation vector.

[0085] Among them, in step S551, based on the maximum value, minimum value, mean value and variance of each first employment information synonymous expression semantic coding feature vector in the set of the first employment information synonymous expression semantic coding feature vector, the characteristic energy level coefficients of each first employment information synonymous expression semantic coding feature vector are determined to obtain a set of first employment information synonymous expression semantic coding feature energy level coefficients, including: respectively extracting the maximum value and the minimum value of each position in the first employment information synonymous expression semantic coding feature vector to obtain the maximum value of the first employment information synonymous expression semantic coding feature and the minimum value of the first employment information synonymous expression semantic coding feature; respectively calculating the mean value and the variance of the first employment information synonymous expression semantic coding feature vector to obtain the mean value of the first employment information synonymous expression semantic coding feature and the variance of the first employment information synonymous expression semantic coding feature; adding the variance of the first employment information synonymous expression semantic coding feature and the preset hyperparameter to obtain the first employment information synonymous expression semantic coding feature. The first energy level coefficient of the semantic coding feature of the synonymous expression of the employment information; calculating the square of the difference between the maximum value of the semantic coding feature of the first employment information synonymous expression and the mean value of the semantic coding feature of the first employment information synonymous expression to obtain the maximum difference value of the semantic coding of the first employment information synonymous expression; calculating the square of the difference between the minimum value of the semantic coding feature of the first employment information synonymous expression and the mean value of the semantic coding feature of the first employment information synonymous expression to obtain the minimum difference value of the semantic coding of the first employment information synonymous expression; adding the maximum difference value of the semantic coding of the first employment information synonymous expression to the minimum difference value of the semantic coding of the first employment information synonymous expression and the preset hyperparameter to obtain the second energy level coefficient of the semantic coding feature of the first employment information synonymous expression; dividing the first energy level coefficient of the semantic coding feature of the first employment information synonymous expression by the second energy level coefficient of the semantic coding feature of the first employment information synonymous expression to obtain the energy level coefficient of the semantic coding feature of the first employment information synonymous expression.

[0086] Among them, in step S553, the set of the first employment information synonymous expression semantic coding feature energy level significance descriptors is input into a mask module based on a gating function to obtain a set of mask-gated probabilistic first employment information synonymous expression semantic coding feature energy level significance descriptors, including: normalizing each first employment information synonymous expression semantic coding feature energy level significance descriptor in the set of the first employment information synonymous expression semantic coding feature energy level significance descriptors to obtain a set of normalized first employment information synonymous expression semantic coding feature energy level significance descriptors; inputting each normalized first employment information synonymous expression semantic coding feature energy level significance descriptor in the set of normalized first employment information synonymous expression semantic coding feature energy level significance descriptors into a gating function for masking processing to obtain the mask-gated probabilistic first employment information synonymous expression semantic coding feature energy level significance descriptors set.

[0087] Specifically, each normalized first employment information synonymous expression semantic coding feature energy level significance descriptor in the set of normalized first employment information synonymous expression semantic coding feature energy level significance descriptors is input into a gating function for masking processing to obtain the set of mask-gated probabilistic first employment information synonymous expression semantic coding feature energy level significance descriptors, including: in response to a normalized first employment information synonymous expression semantic coding feature energy level significance descriptor in the set of normalized first employment information synonymous expression semantic coding feature energy level significance descriptors being greater than a predetermined threshold, setting the normalized first employment information synonymous expression semantic coding feature energy level significance descriptor greater than the predetermined threshold to the original value, otherwise setting it to zero to obtain the set of mask-gated probabilistic first employment information synonymous expression semantic coding feature energy level significance descriptors.

[0088] In one example, the set of semantically encoded feature vectors of synonymous expressions of the first employment information and the first semantically encoded feature vector of the first employment information are input into a feature aggregation network based on feature local energy distribution gating to obtain a first semantically aggregated representation vector of the employment information, including: inputting the set of semantically encoded feature vectors of synonymous expressions of the first employment information and the first semantically encoded feature vector of the first employment information into the feature aggregation network based on feature local energy distribution gating, and processing them with the following feature aggregation formula to obtain the first semantically aggregated representation vector of the employment information; wherein the feature aggregation formula is:

[0089] X={x1,x2,…,x k , …, x n}

[0090]

[0091] w si =mask(w i )

[0092]

[0093] z=concat(h c ,h s )

[0094] Wherein, X is the set of semantically encoded feature vectors of the synonymous expression of the first employment information, n represents the number of feature vectors in the set of semantically encoded feature vectors of the synonymous expression of the first employment information, and μ i and σ 2 i are the mean and variance of the semantic encoding feature vector of the i-th synonymous expression of the first employment information, max(x i ) and min(x i) are the maximum and minimum values ​​of each position in the semantic encoding feature vector of the i-th first employment information synonymous expression, ∈ is the preset hyperparameter, e i is the characteristic energy level coefficient of the semantic encoding feature vector of the i-th synonymous expression of the first employment information, e q is the characteristic energy level coefficient of the qth semantic encoding feature vector of the synonymous expression of the first employment information, Num r Indicates that the i The number of characteristic energy coefficients in the local neighborhood centered at ci is the energy level significance descriptor of the i-th semantic encoding feature vector of the synonymous expression of the first employment information, w i is the ith normalized first employment information proximate expression semantic coding feature energy level significant descriptor in the set of normalized first employment information proximate expression semantic coding feature energy level significant descriptors, mask(·) is the masking process, w si is the ith mask-gated probabilistic first employment information synonymous expression semantic coding feature energy level significance descriptor in the set of the mask-gated probabilistic first employment information synonymous expression semantic coding feature energy level significance descriptors, θ is a predetermined threshold, Y is a set of weighted feature vectors of the semantic coding of the first employment information synonymous expression, y i is the weighted feature vector of semantic encoding of the synonymous expression of the i-th first employment information, h c is the semantic feature energy level guided aggregation representation vector of the synonymous expression of the first employment information, h s is the semantic encoding feature vector of the first employment information, concat(·,·) is the cascade processing, and z is the semantic aggregation representation vector of the first employment information.

[0095] Next, obtain job seeker information.

[0096] Then, considering that the job seeker information contains basic information about the job seeker's job search and semantic content of the job seeker's needs, in the technical solution of the present application, the job seeker information is semantically encoded to obtain a semantic encoding feature vector of the job seeker information. This enables a deep semantic understanding of the text information in the job seeker's resume, including skills, experience, educational background, etc., thereby achieving more efficient information processing and more accurate job matching.

[0097] Next, considering that the first employment information semantic aggregation representation vector integrates all the key semantic features of employment information, including semantic information directly related to the position, such as job description, job requirements, salary range, etc., it highlights the most core and important parts of the job description. The semantic encoding feature vector of the job applicant information reflects the semantic information of key professional characteristics such as the job applicant's professional skills, relevant experience, qualification certification, personal advantages, etc. Based on this, in order to be able to more accurately evaluate the match between job applicants and positions, in the technical solution of the present application, the cosine similarity between the first employment information semantic aggregation representation vector and the job applicant information semantic encoding feature vector is calculated. That is, through measurement methods such as cosine similarity, the similarity or consistency between the first employment information semantic aggregation representation vector and the job applicant information semantic encoding feature vector at the semantic level can be accurately quantified, thereby helping job applicants find positions that match their abilities and experience.

[0098] Then, based on the comparison between the cosine similarity and the preset threshold, it is determined whether to push the first employment information to the job seeker. That is, the comparison between the cosine similarity between the semantic aggregation representation vector of the first employment information and the semantic encoding feature vector of the job seeker information and the preset threshold is used to automatically determine whether to push the first employment information to the job seeker. In this way, key synonym replacement and deep semantic understanding can be performed on the first employment information and job seeker information to more accurately grasp the core requirements of the employment information, as well as the qualifications and expectations of the job seeker, thereby improving the matching accuracy between positions and job seekers, and ensuring that the employment information pushed to job seekers is more relevant and personalized.

[0099] Furthermore, in the technical solution of the present application, the intelligent employment planning method based on fusion data analysis also includes a training step: for training the keyword extractor based on the AIGC model and the feature aggregation network based on feature local energy distribution gating.

[0100] The training step includes: obtaining training data, wherein the training data includes first training employment information, training job seeker information and true cosine similarity extracted from an employment information database; inputting the first training employment information into the keyword extractor based on the AIGC model to obtain a training keyword extraction result; based on the training keyword extraction result, performing keyword synonym replacement on the first training employment information to obtain a set of synonymous expressions of the first training employment information; performing semantic encoding on the set of synonymous expressions of the first training employment information and the first training employment information to obtain a set of semantic encoding feature vectors of the synonymous expressions of the first training employment information and a semantic encoding feature vector of the first training employment information; inputting the set of semantic encoding feature vectors of the synonymous expressions of the first training employment information and the semantic encoding feature vector of the first training employment information into the feature local energy based method. A distribution-gated feature aggregation network is used to obtain a first training employment information semantic aggregation representation vector; the training job seeker information is semantically encoded to obtain a training job seeker information semantic encoding feature vector; the training cosine similarity between the first training employment information semantic aggregation representation vector and the training job seeker information semantic encoding feature vector is calculated; the difference between the training cosine similarity and the true cosine similarity is calculated as the difference loss function value; the employment and job search information constraint loss function value between the first training employment information semantic aggregation representation vector and the training job seeker information semantic encoding feature vector is calculated; the keyword extractor based on the AIGC model and the feature aggregation network based on feature local energy distribution gating are trained through gradient back propagation based on the weighted sum of the difference loss function value and the employment and job search information constraint loss function value as the loss function value.

[0101] Preferably, considering that the first training employment information semantic aggregation representation vector and the training job applicant information semantic encoding feature vector respectively represent the text semantic aggregation features based on feature local energy distribution gating of the synonymous expressions of the first training employment information and the first training employment information and the text semantic features of the training job applicant information, the first training employment information semantic aggregation representation vector and the training job applicant information semantic encoding feature vector may also cause an imbalance in the correspondence ratio of the text semantic fine-grained distribution due to the difference in the sample density of the text semantic feature distribution, resulting in a mapping offset relative to the common cosine similarity domain. To this end, the present application hopes to improve the common domain mapping offset of the first training employment information semantic aggregation representation vector and the training job applicant information semantic encoding feature vector under the imbalance in the correspondence ratio of the text semantic fine-grained distribution.

[0102] Therefore, in a preferred embodiment of the present application, in the training of the keyword extractor based on the AIGC model and the feature aggregation network based on the feature local energy distribution gating based on the difference loss function value and through back propagation of gradient descent, a labor and job search information constraint loss function value other than the difference loss function value is further introduced, that is, the weighted sum of the difference loss function value and the labor and job search information constraint loss function value is used as the loss function value to train the keyword extractor based on the AIGC model and the feature aggregation network based on the feature local energy distribution gating through back propagation of gradient, including the following steps: calculating the L1 norm and L2 norm of the mean feature vector of the first training employment information semantic aggregation representation vector and the training job seeker information semantic encoding feature vector, wherein the L1 norm of the mean feature vector is the sum of each eigenvalue of the mean feature vector, and the L2 norm of the mean feature vector is the square root of the sum of the squares of each eigenvalue of the mean feature vector; calculating the square root of the L2 norm The inverse of and the weighted sum of the L1 norm are taken, and the dot product is performed with the dot difference result of the first training employment information semantic aggregation representation vector and the training job seeker information semantic encoding feature vector to obtain the dot product result feature vector, and all eigenvalues ​​of the dot product result feature vector are summed to obtain the first loss term; the dot division feature vector of the first training employment information semantic aggregation representation vector and the training job seeker information semantic encoding feature vector is dotted with the square root of the scale of the first training employment information semantic aggregation representation vector to obtain the dot addition feature vector, and all eigenvalues ​​of the dot addition feature vector are summed to obtain the second loss term; the difference between the first loss term and the second loss term multiplied by the weight as a hyperparameter is calculated to obtain the employment and job seeker information constraint loss function value; the keyword extractor based on the AIGC model and the feature aggregation network based on feature local energy distribution gating are trained based on the weighted sum of the difference loss function value and the employment and job seeker information constraint loss function value as the loss function value and the back propagation of gradient descent.

[0103] The calculation process of the employment and job search information constraint loss function value is expressed as follows:

[0104]

[0105] Wherein, V1 and V2 are respectively the semantic aggregation representation vector of the first training employment information and the semantic encoding feature vector of the training job seeker information, V μ is the mean feature vector of the first training employment information semantic aggregation representation vector and the training job seeker information semantic encoding feature vector, ||·|| L1 and ||·|| L2They represent the L1 norm and L2 norm of the eigenvector respectively, ⊙ is the point multiplication by position, is the point difference by position, S is the dot product result feature vector, f1 is the feature value of each position in the dot product result feature vector, L is the scale of the first training employment information semantic aggregation representation vector, C is the point-added feature vector, f2 is the feature value of each position in the point-added feature vector, and α and β are the weighted sum weights as hyperparameters, and Loss is the employment and job search information constraint loss function value.

[0106] Therefore, based on the constrained representation of the structured foreground and background distinction of the supermanifold of the first training employment information semantic aggregation representation vector and the training job applicant information semantic encoding feature vector based on the mean feature vector norm between the first training employment information semantic aggregation representation vector and the training job applicant information semantic encoding feature vector, the feature-level key differential correspondence between the feature vectors of the first training employment information semantic aggregation representation vector and the training job applicant information semantic encoding feature vector is modeled, and the global differential association relationship between the corresponding features is adjusted, so as to make positive fine-grained differential correspondence suggestions through the unbalanced ratio control between the corresponding feature values ​​of the first training employment information semantic aggregation representation vector and the training job applicant information semantic encoding feature vector, so as to avoid the imbalance of the feature vector-level text semantic distribution correspondence ratio of the first training employment information semantic aggregation representation vector and the training job applicant information semantic encoding feature vector by focusing on the focus, improve its mapping offset relative to the common cosine similarity domain, and thus improve the accuracy of the obtained training cosine similarity. In this way, key synonym replacement and deep semantic understanding can be performed on the first training employment information and training job seeker information to more accurately grasp the core requirements of the employment information, as well as the qualifications and expectations of job seekers, thereby improving the matching accuracy between positions and job seekers and ensuring that the employment information pushed to job seekers is more relevant and personalized.

[0107] Based on the above embodiments, see Figure 4As shown, it is a structural diagram of a smart employment planning system 800 based on fusion data analysis in an embodiment of the present application. The smart employment planning system 800 based on fusion data analysis includes: an employment information extraction module 810, which is used to extract the first employment information from the employment information database; a keyword extraction module 820, which is used to input the first employment information into a keyword extractor to obtain a keyword extraction result; a keyword synonym replacement module 830, which is used to perform keyword synonym replacement on the first employment information based on the keyword extraction result to obtain a set of synonymous expressions of the first employment information; an employment information semantic encoding module 840, which is used to perform semantic encoding on the set of synonymous expressions of the first employment information and the first employment information to obtain a set of semantic encoding feature vectors of the first employment information synonymous expressions and a first employment information semantic encoding feature vector; a feature aggregation module 850, A module 860 is configured to input the set of semantically encoded feature vectors of synonymous expressions of the first employment information and the semantically encoded feature vector of the first employment information into a feature aggregation network based on feature local energy distribution gating to obtain a semantically aggregated representation vector of the first employment information; a job search information acquisition module 860 is configured to obtain job seeker information; a job search information semantic encoding module 870 is configured to semantically encode the job seeker information to obtain a semantically encoded feature vector of the job seeker information; a similarity calculation module 880 is configured to calculate the cosine similarity between the semantically aggregated representation vector of the first employment information and the semantically encoded feature vector of the job seeker information; and a push module 890 is configured to determine whether to push the first employment information to the job seeker based on a comparison between the cosine similarity and a preset threshold.

[0108] In one example, in the above-mentioned intelligent employment planning system 800 based on fusion data analysis, the keyword extraction module 820 is used to: input the first employment information into a keyword extractor based on the AIGC model to obtain the keyword extraction result.

[0109] Here, those skilled in the art can understand that the specific functions and operations of each module in the above-mentioned intelligent employment planning system 800 based on fusion data analysis have been referred to above. Figure 2 to Figure 3 It has been introduced in detail in the description of the intelligent employment planning method based on fusion data analysis, and therefore, its repeated description will be omitted.

[0110] Figure 5 FIG. 1 is an application scenario diagram of the intelligent employment planning method based on fusion data analysis according to an embodiment of the present application. Figure 5 As shown, in this application scenario, first, the first employment information (for example, Figure 5 D1 as shown in the figure) and obtaining information about job seekers (e.g. Figure 5Then, the first employment information and the job seeker information are input into a server (for example, Figure 5 In S) shown in , the server is able to use the intelligent employment planning algorithm based on fusion data analysis to process the first employment information and the job seeker information to determine whether to push the first employment information to the job seeker.

[0111] Based on the above embodiments, another exemplary embodiment of an electronic device is also provided in the embodiments of the present application. In some possible implementations, the electronic device in the embodiments of the present application may include a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor can implement the steps of the intelligent employment planning method based on fusion data analysis in the above embodiments when executing the program.

[0112] For example, in the case of electronic equipment Figure 1 Taking the server 100 in the example as an example, the processor in the electronic device is the processor 110 in the server 100, and the memory in the electronic device is the memory 120 in the server 100.

[0113] The embodiment of the present application also provides a computer-readable storage medium having computer executable instructions stored thereon. When the computer executable instructions are executed by a processor, the intelligent employment planning method based on fusion data analysis according to the embodiment of the present application described with reference to the above figures can be executed. The computer-readable storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. The volatile memory may, for example, include a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may, for example, include a read-only memory (ROM), a hard disk, a flash memory, etc.

[0114] The embodiment of the present application also provides a computer program product or a computer program, which includes computer executable instructions, and the computer executable instructions are stored in a computer readable storage medium. The processor of the computer device reads the computer executable instructions from the computer readable storage medium, and the processor executes the computer executable instructions, so that the computer device executes the intelligent employment planning method based on fusion data analysis according to the embodiment of the present application.

[0115] Those skilled in the art will appreciate that the contents disclosed in this application may be subject to various variations and improvements. For example, the various devices or components described above may be implemented by hardware, or by software, firmware, or a combination of some or all of the three.

[0116] In addition, although the present application makes various references to certain units in the system according to embodiments of the present application, any number of different units can be used and run on the client and / or server. The units are only illustrative, and different aspects of the system and method can use different units.

[0117] Those skilled in the art will appreciate that all or part of the steps in the above method can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, such as a read-only memory, a disk or an optical disk. Optionally, all or part of the steps in the above embodiment can also be implemented using one or more integrated circuits. Accordingly, each module / unit in the above embodiment can be implemented in the form of hardware or in the form of software function modules. The present application is not limited to any particular form of combination of hardware and software.

[0118] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art to which this application belongs. It should also be understood that terms such as those defined in common dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an idealized or extremely formal sense, unless explicitly defined as such herein.

[0119] The above is an explanation of the present application and should not be considered as limiting thereof. Although several exemplary embodiments of the present application have been described, those skilled in the art will readily appreciate that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present application.

Claims

1. A smart employment planning method based on fusion data analysis, characterized in that: include: Extracting first employment information from an employment information database; Inputting the first employment information into a keyword extractor to obtain a keyword extraction result; Based on the keyword extraction result, performing keyword synonym replacement on the first employment information to obtain a set of synonymous expressions of the first employment information; Performing semantic coding on the set of synonymous expressions of the first employment information and the first employment information to obtain a set of semantic coding feature vectors of synonymous expressions of the first employment information and a semantic coding feature vector of the first employment information; Inputting the set of semantic encoding feature vectors of the synonymous expression of the first employment information and the semantic encoding feature vector of the first employment information into a feature aggregation network based on feature local energy distribution gating to obtain a first semantic aggregation representation vector of the employment information; Obtain job applicant information; Performing semantic encoding on the job seeker information to obtain a semantic encoding feature vector of the job seeker information; Calculating the cosine similarity between the semantic aggregation representation vector of the first employment information and the semantic encoding feature vector of the job seeker information; Based on the comparison between the cosine similarity and a preset threshold, it is determined whether to push the first employment information to the job seeker.

2. The intelligent employment planning method based on fusion data analysis according to claim 1 is characterized in that: Inputting the first employment information into a keyword extractor to obtain a keyword extraction result includes: inputting the first employment information into a keyword extractor based on an AIGC model to obtain the keyword extraction result.

3. The intelligent employment planning method based on fusion data analysis according to claim 2 is characterized in that: Inputting the set of semantically encoded feature vectors of the synonymous expression of the first employment information and the semantically encoded feature vector of the first employment information into a feature aggregation network based on feature local energy distribution gating to obtain a first semantically aggregated representation vector of the employment information, including: Determine, based on the maximum value, minimum value, mean value and variance of each first employment information synonymous expression semantic coding feature vector in the set of the first employment information synonymous expression semantic coding feature vectors, the characteristic energy level coefficients of each first employment information synonymous expression semantic coding feature vector to obtain a set of first employment information synonymous expression semantic coding feature energy level coefficients; Determine the scale of the local neighborhood, calculate the local neighborhood significance descriptor of each first employment information proximity expression semantic coding feature energy level coefficient in the set of the first employment information proximity expression semantic coding feature energy level coefficients to obtain a set of first employment information proximity expression semantic coding feature energy level significance descriptors; Inputting the set of energy level significance descriptors of semantic coding features of synonymous expressions of the first employment information into a mask module based on a gating function to obtain a set of mask-gated probabilistic energy level significance descriptors of semantic coding features of synonymous expressions of the first employment information; Taking the set of the mask-gated probabilistic semantic encoding feature energy level significance descriptors of the first employment information proximate expression as a weighted sequence, calculating the position-weighted sum of the set of the first employment information proximate expression semantic encoding feature vectors to obtain the first employment information proximate expression semantic feature energy level guided aggregation representation vector; The first employment information synonymous expression semantic feature energy level guided aggregation representation vector and the first employment information semantic encoding feature vector are cascaded to obtain the first employment information semantic aggregation representation vector.

4. The intelligent employment planning method based on fusion data analysis according to claim 3 is characterized in that: Based on the maximum value, minimum value, mean value and variance of each first employment information synonymous expression semantic coding feature vector in the set of the first employment information synonymous expression semantic coding feature vectors, determining the characteristic energy level coefficients of each first employment information synonymous expression semantic coding feature vector to obtain a set of first employment information synonymous expression semantic coding feature energy level coefficients, including: Respectively extracting the maximum value and the minimum value of each position in the semantic coding feature vector of the synonymous expression of the first employment information to obtain the maximum value of the semantic coding feature of the synonymous expression of the first employment information and the minimum value of the semantic coding feature of the synonymous expression of the first employment information; Respectively calculating the mean and variance of the semantic coding feature vector of the synonymous expression of the first employment information to obtain the mean of the semantic coding feature of the synonymous expression of the first employment information and the variance of the semantic coding feature of the synonymous expression of the first employment information; Adding the variance of the semantic coding feature of the synonymous expression of the first employment information and a preset hyperparameter to obtain a first energy level coefficient of the semantic coding feature of the synonymous expression of the first employment information; Calculate the square of the difference between the maximum value of the semantic coding feature of the synonymous expression of the first employment information and the mean value of the semantic coding feature of the synonymous expression of the first employment information to obtain the maximum difference value of the semantic coding of the synonymous expression of the first employment information; Calculate the square of the difference between the minimum value of the semantic coding feature of the synonymous expression of the first employment information and the mean value of the semantic coding feature of the synonymous expression of the first employment information to obtain the minimum difference value of the semantic coding of the synonymous expression of the first employment information; The maximum difference value of the semantic coding of the synonymous expression of the first employment information is added to the minimum difference value of the semantic coding of the synonymous expression of the first employment information and the preset hyperparameter to obtain the second energy level coefficient of the semantic coding feature of the synonymous expression of the first employment information; The first energy level coefficient of the semantic coding feature of the synonymous expression of the first employment information is divided by the second energy level coefficient of the semantic coding feature of the synonymous expression of the first employment information to obtain the energy level coefficient of the semantic coding feature of the synonymous expression of the first employment information.

5. The intelligent employment planning method based on fusion data analysis according to claim 4 is characterized in that: Inputting the set of energy level significance descriptors of semantic coding features of synonymous expressions of the first employment information into a mask module based on a gating function to obtain a set of mask-gated probabilistic energy level significance descriptors of semantic coding features of synonymous expressions of the first employment information, including: Normalizing each first employment information synonymous expression semantic coding feature energy level significance descriptor in the set of the first employment information synonymous expression semantic coding feature energy level significance descriptors to obtain a set of normalized first employment information synonymous expression semantic coding feature energy level significance descriptors; Each normalized first employment information proximate expression semantic coding feature energy level significance descriptor in the set of normalized first employment information proximate expression semantic coding feature energy level significance descriptors is input into a gating function for masking processing to obtain the set of mask-gated probabilistic first employment information proximate expression semantic coding feature energy level significance descriptors.

6. The intelligent employment planning method based on fusion data analysis according to claim 5 is characterized in that: Each normalized first employment information proximal expression semantic coding feature energy level significance descriptor in the set of normalized first employment information proximal expression semantic coding feature energy level significance descriptors is input into a gating function for masking processing to obtain the set of mask-gated probabilistic first employment information proximal expression semantic coding feature energy level significance descriptors, including: in response to a normalized first employment information proximal expression semantic coding feature energy level significance descriptor in the set of normalized first employment information proximal expression semantic coding feature energy level significance descriptors being greater than a predetermined threshold, setting the normalized first employment information proximal expression semantic coding feature energy level significance descriptor greater than the predetermined threshold to the original value, otherwise setting it to zero to obtain the set of mask-gated probabilistic first employment information proximal expression semantic coding feature energy level significance descriptors.

7. The intelligent employment planning method based on fusion data analysis according to claim 6 is characterized in that: It also includes a training step: used to train the keyword extractor based on the AIGC model and the feature aggregation network based on feature local energy distribution gating.

8. The intelligent employment planning method based on fusion data analysis according to claim 7 is characterized in that: The training step comprises: Acquire training data, wherein the training data includes first training employment information, training job seeker information, and true cosine similarity extracted from an employment information database; Inputting the first training employment information into the keyword extractor based on the AIGC model to obtain a training keyword extraction result; Based on the training keyword extraction result, performing keyword synonym replacement on the first training employment information to obtain a set of synonymous expressions of the first training employment information; Performing semantic coding on the set of synonymous expressions of the first training employment information and the first training employment information to obtain a set of semantic coding feature vectors of synonymous expressions of the first training employment information and a semantic coding feature vector of the first training employment information; Inputting the set of semantic encoding feature vectors of the first training employment information synonymous expression and the first training employment information semantic encoding feature vector into the feature aggregation network based on feature local energy distribution gating to obtain the first training employment information semantic aggregation representation vector; Performing semantic encoding on the training job seeker information to obtain a semantic encoding feature vector of the training job seeker information; Calculating the training cosine similarity between the first training employment information semantic aggregation representation vector and the training job seeker information semantic encoding feature vector; Calculating the difference between the training cosine similarity and the true cosine similarity as a difference loss function value; Calculating the employment and job-seeking information constraint loss function value between the first training employment information semantic aggregation representation vector and the training job seeker information semantic encoding feature vector; The keyword extractor based on the AIGC model and the feature aggregation network based on feature local energy distribution gating are trained through gradient back propagation based on the weighted sum of the difference loss function value and the job search information constraint loss function value as the loss function value.

9. An intelligent employment planning system based on fusion data analysis, characterized in that: include: An employment information extraction module, used to extract first employment information from an employment information database; A keyword extraction module, used for inputting the first employment information into a keyword extractor to obtain a keyword extraction result; A keyword synonym replacement module, configured to perform keyword synonym replacement on the first employment information based on the keyword extraction result to obtain a set of synonymous expressions of the first employment information; An employment information semantic coding module, configured to semantically code the set of synonymous expressions of the first employment information and the first employment information to obtain a set of semantic coding feature vectors of synonymous expressions of the first employment information and a semantic coding feature vector of the first employment information; A feature aggregation module, used for inputting the set of semantic encoding feature vectors of the synonymous expression of the first employment information and the first semantic encoding feature vector of the employment information into a feature aggregation network based on feature local energy distribution gating to obtain a first semantic aggregation representation vector of the employment information; A job search information acquisition module is used to obtain job seeker information; A job-seeking information semantic encoding module, used for semantically encoding the job-seeker information to obtain a semantic encoding feature vector of the job-seeker information; A similarity calculation module, used to calculate the cosine similarity between the first employment information semantic aggregation representation vector and the job seeker information semantic encoding feature vector; A push module is used to determine whether to push the first employment information to the job seeker based on the comparison between the cosine similarity and a preset threshold.

10. The intelligent employment planning system based on fusion data analysis according to claim 9 is characterized in that: The keyword extraction module is used to: input the first employment information into a keyword extractor based on the AIGC model to obtain the keyword extraction result.