Intelligent multimodal evaluation implementation method and metaverse intelligent evaluation system
Through the metacosmic intelligent evaluation system and multimodal intelligent evaluation platform, the problems of low recruitment efficiency and unstable evaluation in the traditional recruitment model are solved, intelligent evaluation and automated matching of applicants are achieved, and recruitment efficiency and evaluation reliability are improved. , suitable for the selection and appointment of precise talents for small and medium-sized enterprises.
Patent Information
- Application Number
- CN202410019082.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2023-10-31
- Filing Date
- 2024-01-04
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2044-01-04
AI Technical Summary
Under the traditional model of the existing Internet recruitment platform, the low recruitment efficiency and unstable evaluation have led to challenges in talent selection and recruitment, especially the recruitment quality and efficiency problems caused by inexperienced recruitment of assessors.
Using the metacosmic intelligent evaluation system, a multimodal intelligent evaluation platform combined with deep learning capabilities, VR, AR, and MR devices are used to train deep learning models through multimodal data sets to realize intelligent multimodal evaluation for candidates and improve Evaluate efficiency and stability, automatically match candidate recruitment candidates, and conduct remote meta-universe immersive simulation work tests.
It improves the recruitment efficiency and reliability of evaluation, reduces human intervention, ensures the stability and objectivity of evaluation results, is suitable for the selection and recruitment of small and medium-sized enterprises, and improves the flexibility and efficiency of recruitment.
Smart Images

Figure CN117875918B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of metaverse technology, and in particular to an intelligent multimodal evaluation implementation method and a metaverse intelligent evaluation system. Background Art
[0002] The existing human resource recruitment method is usually that the employer releases recruitment information, and the applicants submit their resumes to the employer's target position through the recruitment information, and then the human resources department conducts a face-to-face assessment of the applicant.
[0003] Since the emergence of Monster Internet recruitment platform in the United States in 1999, the global Internet recruitment platform has developed rapidly. However, the current Internet recruitment platforms only set a dozen simple matching fields for recruitment positions (such as education, gender, age, etc.). Under this traditional Internet recruitment model, an average of more than 200 people submit resumes for each position. Recruiters need to manually screen whether the applicants need to come to the on-site interview. In addition, it takes a lot of time for both the applicants and the employer's recruiters to communicate face to face. There is a problem that the applicants cannot arrive at the company in time for evaluation, or when multiple applicants need to be evaluated, the evaluators in the human resources department of the employer are inexperienced, which leads to low recruitment efficiency for the employer's target positions and low application efficiency for the applicants. In addition, different evaluators may have different evaluation criteria for the same problem, and the lack of stability in the evaluation scoring criteria will affect the overall effect of the recruiter, as well as the reliability and efficiency of the recruitment quality. Finally, 98% of the total number of enterprises in my country are small and medium-sized enterprises with less than 50 employees; 90% of these small and medium-sized enterprises lack experienced recruitment and assessment personnel. How to use artificial intelligence technology to help tens of millions of small and medium-sized enterprises select talents more accurately is an innovation with great social value. It can be seen that it is particularly important to provide a new assessment method to improve the efficiency of recruitment assessment and improve the reliability and flexibility of recruitment. Summary of the invention
[0004] The technical problem to be solved by the present invention is to provide an intelligent multimodal evaluation implementation method and a metaverse intelligent evaluation system, which can help improve the evaluation efficiency, as well as improve the stability and reliability of the evaluation of applicants, and further help improve the reliability and efficiency of recruitment.
[0005] In order to solve the above technical problems, the first aspect of the present invention discloses a method for realizing intelligent multimodal evaluation, which is applied to a Metaverse intelligent evaluation system, wherein the Metaverse intelligent evaluation system includes a multimodal intelligent evaluation platform with deep learning capabilities, a target device, and a talent selection element set, wherein the target device includes one or more of a VR device, an AR device, and an MR device, and the method is implemented by the multimodal intelligent evaluation platform and / or the target device included in the Metaverse intelligent evaluation system; the method includes:
[0006] The Metaverse intelligent assessment system inputs the multimodal data set into a predetermined deep learning model for training based on the multimodal data set collected from several users when using the multimodal intelligent assessment platform for recruitment and assessment, so as to obtain a target deep learning model trained to convergence; wherein, the multimodal assessment data set includes a multimodal assessment data pair corresponding to each of the multimodal assessments in several multimodal assessments acquired in real time from the Metaverse intelligent assessment system, and the assessment data pair includes assessment multimodal data and assessment result data corresponding to the assessment multimodal data, wherein the assessment multimodal data includes real-time multimodal data of assessors and real-time multimodal data of assessment objects, and the assessment result data includes an assessment result output by at least one of the assessors for at least one of the assessment objects, and the Metaverse intelligent assessment system includes the target deep learning model;
[0007] The Metaverse intelligent evaluation system collects real-time evaluation multimodal data of a number of target objects to be evaluated, analyzes the real-time evaluation multimodal data of each target object, obtains a multimodal analysis result of the target object, determines at least one assessment factor corresponding to the target object based on the multimodal analysis result of the target object, and outputs all the assessment factors determined for the target object to the multimodal evaluation terminal corresponding to the target object, so that the multimodal evaluation terminal outputs all the assessment factors for the target object to solve;
[0008] For each of the target objects, the Metaverse intelligent assessment system obtains the response processing result information of the target object for each of the assessment factors, inputs all the response processing result information corresponding to the target object into the target deep learning model, and obtains the response processing result analysis result; wherein, the response processing result information includes image information obtained by image acquisition of the target object, voice information obtained by voice acquisition of the target object, and a variety of information contained in the data stream formed by the image information and voice information of the target object within a specific unit time;
[0009] For each of the target objects, the Metaverse Intelligent Assessment System determines an assessment result of the target object based on an analysis result of the response processing result of the target object;
[0010] The Metaverse intelligent assessment system performs an iterative training operation on the target deep learning model in the Metaverse intelligent assessment system based on the collected assessment result data corresponding to multiple assessed objects that have undergone multimodal assessment within a certain time period and the real-time assessment multimodal data corresponding to each of the assessed objects, so as to update the target deep learning model, wherein all the assessed objects include at least one assessed object that has undergone an autonomous multimodal assessment through the Metaverse intelligent assessment system within the certain time period and / or at least one assessed object that has undergone a non-autonomous multimodal assessment through the Metaverse intelligent assessment system within the certain time period; the autonomous multimodal assessment includes a multimodal assessment without the participation of assessors, and the non-autonomous multimodal assessment includes a multimodal assessment with the participation of corresponding assessors.
[0011] As an optional implementation, in the first aspect of the present invention, for each of the target objects, after the Metaverse intelligent evaluation system determines the evaluation result of the target object according to the analysis result of the response processing result of the target object, the method further includes:
[0012] For each of the target objects, the Metaverse intelligent assessment system obtains the recruitment demand information of the target position of the target employer corresponding to the target object, and inputs the recruitment demand information of the target position of the target employer corresponding to the target object into a predetermined demand analysis model to obtain the demand analysis result of the target position of the target employer corresponding to the target object;
[0013] For each of the target objects, the Metaverse intelligent assessment system calculates the matching degree between the assessment result of the target object and the demand analysis result of the target position of the target employer corresponding to the target object through a predetermined matching algorithm, obtains the person-position matching degree between the target object and the target employer corresponding to the target object, and generates a position matching result and report between the target object and the target employer corresponding to the target object based on the person-position matching degree;
[0014] And, the method further comprises:
[0015] For each of the target objects, the Metaverse intelligent assessment system determines whether the job matching degree between the target object and the target employer corresponding to the target object meets the preset job matching conditions based on the job matching degree result between the target object and the target employer corresponding to the target object;
[0016] For each of the target objects, when the Metaverse intelligent assessment system determines that the job matching degree between the target object and the target employer corresponding to the target object meets the preset job matching condition, the target object is determined as a candidate for recruitment of the target position of the target employer corresponding to the target object;
[0017] For each of the target employers, the Metaverse intelligent assessment system generates assessment recommendation information based on the assessment results of each candidate recruitment object corresponding to the target employer, and pushes the assessment recommendation information to the terminal corresponding to each candidate recruitment object and the terminal corresponding to the target employer.
[0018] As an optional implementation manner, in the first aspect of the present invention, for each of the target objects, the target object is determined as a candidate for recruitment of a target position of a target employer corresponding to the target object, and the method further includes:
[0019] For each associated object that matches the target object, the Metaverse Intelligent Evaluation Platform determines the value parameters of the evaluation information corresponding to the associated object based on the selection factor reply information of the associated object, and determines the remuneration information of the intelligent selection optimization contribution that matches the associated object based on the information value parameters and a predetermined incentive fee accounting algorithm, and pays the associated object the remuneration fee that matches the associated object based on the remuneration information.
[0020] As an optional implementation manner, in the first aspect of the present invention, after determining, for each target object, the target object as a candidate for recruitment of a target position of a target employer corresponding to the target object, the method further comprises:
[0021] For each candidate for recruitment of a target position of the target employer, the Metaverse intelligent assessment system performs a remote Metaverse immersive simulated work test on the candidate based on the assessment result of the candidate, obtains the simulated work performance information of the candidate, and performs quantitative scoring on the simulated work performance information of the candidate according to a preset work performance quantification algorithm, obtains the Metaverse immersive remote simulated work performance score of the candidate, and determines whether the remote simulated work performance score of the candidate is greater than or equal to a preset work performance threshold;
[0022] For each candidate recruitment object for the target position of the target employer, when the Metaverse intelligent assessment system determines that the remote simulation work performance score of the candidate recruitment object is greater than or equal to the preset work performance threshold, the Metaverse intelligent assessment system generates a work simulation test report and recommendation information based on the work simulation test results of each candidate recruitment object corresponding to the target employer, and pushes the work simulation test report and recommendation information to the terminal corresponding to each candidate recruitment object and the terminal corresponding to the target employer.
[0023] As an optional implementation, in the first aspect of the present invention, for each of the target objects, the Metaverse intelligent assessment system obtains the response processing result information of the target object for each of the assessment elements, inputs all the response processing result information corresponding to the target object into the target deep learning model, and before obtaining the response processing result analysis result, the method further includes:
[0024] For each of the target objects, the Metaverse intelligent assessment system determines the facial image information of the target object based on the image information corresponding to the target object, and determines the target voice information of the target object based on the voice information corresponding to the target object;
[0025] Wherein, the inputting of all the corresponding processing result information corresponding to the target object into the target deep learning model to obtain the corresponding processing result analysis result includes:
[0026] The facial image information of the target object and the voice information corresponding to the target object are input into the target deep learning model to obtain the corresponding processing result analysis result.
[0027] As an optional implementation, in the first aspect of the present invention, for each of the target objects, the Metaverse intelligent assessment system determines the facial image information of the target object according to the image information corresponding to the target object, and determines the target voice information of the target object according to the voice information corresponding to the target object, and the method further includes:
[0028] The Metaverse intelligent assessment system collects a micro-expression and micro-movement training data set and a voice intonation and / or tone data training set, and performs a training operation on the micro-expression and micro-movement training data set and the voice intonation and / or tone data training set through the predetermined deep learning algorithm to construct a multimodal personality assessment model;
[0029] The Metaverse intelligent assessment system calculates the model loss parameter of the multimodal personality assessment model, and determines whether the model loss parameter meets the preset model generation condition;
[0030] When the Metaverse intelligent assessment system determines that the model loss parameter satisfies the preset model generation condition, for each target object, the facial image information, behavior information and target voice information of the target object are input into the multimodal personality assessment model to obtain a comprehensive assessment result of the target object;
[0031] For each of the target objects, the Metaverse Intelligent Assessment System determines the assessment result of the target object according to the analysis result of the response processing result of the target object, including:
[0032] For each of the target objects, the Metaverse intelligent assessment system determines the assessment result of the target object based on the analysis result of the response processing result of the target object and the multimodal personality assessment result of the target object.
[0033] As an optional embodiment, in the first aspect of the present invention, the method further comprises:
[0034] The Metaverse intelligent assessment system determines the employment demand information of the target position of at least one employer, and generates a set of alternative assessment questions and inspection tasks based on the employment demand information of the target position of each employer, wherein the set of alternative assessment questions and inspection tasks includes a plurality of alternative assessment questions and inspection tasks;
[0035] The generation of the candidate evaluation questions and the inspection task set based on the employment demand information of the target positions of each employer includes:
[0036] The Metaverse Intelligent Assessment System obtains an assessment question and inspection task training data set, wherein the assessment question and inspection task training data set includes a plurality of assessment training questions and a training answer corresponding to each of the assessment training questions, and includes a plurality of inspection task training data and a training inspection processing result corresponding to each of the inspection task training data;
[0037] The Metaverse Intelligent Assessment System performs model training on a predetermined candidate question generation model through a predetermined deep learning algorithm and the assessment question and inspection task training data set to obtain a question generation model;
[0038] The Metaverse Intelligent Assessment System inputs the employment demand information of the target positions of all the employers into the question generation model to obtain the model output results, and based on the model output results, generates a set of alternative assessment questions and inspection tasks, wherein the model output results include at least one alternative assessment question and inspection task;
[0039] And, the determining, based on the multimodal analysis result of the target object, at least one investigation factor corresponding to the target object includes:
[0040] The Metaverse intelligent assessment system determines the assessment elements of the target object from the set of alternative assessment questions and assessment tasks based on the multimodal analysis results of the target object, wherein the assessment elements include assessment questions and / or assessment tasks.
[0041] As an optional implementation, in the first aspect of the present invention, the Metaverse intelligent evaluation system determines the evaluation factors of the target object from the set of candidate evaluation questions and evaluation tasks based on the multimodal analysis results of the target object, including:
[0042] The Metaverse intelligent evaluation system determines at least one evaluation scoring factor corresponding to the target object based on the multimodal analysis result;
[0043] The Metaverse Intelligent Assessment System performs a weight allocation operation on each of the assessment scoring factors according to all the assessment scoring factors corresponding to the target object, and obtains a weight value corresponding to each of the assessment scoring factors;
[0044] The Metaverse Intelligent Assessment System determines a set of assessment factors for the target object from the set of candidate assessment questions and assessment tasks according to the weight value corresponding to each of the assessment scoring factors;
[0045] And, for each of the target objects, the Metaverse intelligent assessment system obtains the response processing result information of the target object for each of the assessment elements, inputs all the response processing result information corresponding to the target object into the target deep learning model, and before obtaining the response processing result analysis result, the method further includes:
[0046] The Metaverse intelligent evaluation system determines the amount of training data of all the evaluation data pairs included in the evaluation data set, and determines whether the amount of training data is greater than or equal to a preset quantity threshold;
[0047] When the Metaverse Intelligent Assessment System determines that the amount of training data is greater than or equal to the preset quantity threshold, it triggers the execution of the operation in which, for each of the target objects, the Metaverse Intelligent Assessment System obtains the response processing result information of the target object for each of the assessment factors, inputs all the response processing result information corresponding to the target object into the target deep learning model, and obtains the response processing result analysis result.
[0048] As an optional implementation, in the first aspect of the present invention, before the Metaverse intelligent assessment system determines the employment demand information of the target position of at least one employer, the method further includes:
[0049] The Metaverse intelligent assessment system collects recruitment information of target positions of at least one employer, and for each of the employers, determines the recruitment position of the employer's target position based on the recruitment information of the employer's target position, and obtains the target unit information of the employer's target position, and determines the position requirement information of the recruitment position based on the target unit information of the employer's target position; wherein the target unit information includes one or more of the work scenario information of the employer's target position, the position requirement information of the employer's target position, and the skill requirement information of the employer's target position;
[0050] For each of the employers, the Metaverse intelligent assessment system generates a job analysis result of the employer's target position based on the job requirement information of the employer's target position;
[0051] The Metaverse intelligent assessment system determines the employment demand information of the target position of at least one employer, including, for each of the employers, the Metaverse intelligent assessment system determines the employment demand information of the target position of the employer according to the job analysis result of the target position of the employer;
[0052] The employment demand information includes, but is not limited to, one or more of work task information, job qualification information, salary range information, employment time information, and work location information.
[0053] A second aspect of the present invention discloses a metaverse intelligent assessment system, the metaverse intelligent assessment system comprising a multimodal intelligent assessment platform with deep learning capabilities, a target device and a talent selection element set, the target device comprising one or more of a VR device, an AR device, and an MR device, the metaverse intelligent assessment system comprising the multimodal intelligent assessment platform and / or the target device,
[0054] The Metaverse Intelligent Assessment System comprises:
[0055] A training module, for inputting the multimodal data set into a predetermined deep learning model for training based on a multimodal data set collected from several users when using the multimodal intelligent evaluation platform for recruitment and evaluation, to obtain a target deep learning model trained to convergence; wherein the multimodal evaluation data set includes a multimodal evaluation data pair corresponding to each of the multimodal evaluations in several multimodal evaluations acquired in real time from the Metaverse Intelligent Evaluation System, the evaluation data pair includes evaluation multimodal data and evaluation result data corresponding to the evaluation multimodal data, wherein the evaluation multimodal data includes real-time multimodal data of an evaluator and real-time multimodal data of an evaluation object, the evaluation result data includes an evaluation result output by at least one of the evaluators for at least one of the evaluation objects, and the Metaverse Intelligent Evaluation System includes the target deep learning model;
[0056] An acquisition module, used to acquire real-time evaluation multimodal data of a number of target objects to be evaluated;
[0057] An analysis module, configured to analyze the real-time evaluation multimodal data of each target object to obtain a multimodal analysis result of the target object;
[0058] A determination module, configured to determine at least one investigation element corresponding to the target object based on the multimodal analysis result of the target object;
[0059] An output module, used for outputting all the examination factors determined for the target object to a multimodal evaluation terminal corresponding to the target object, so that the multimodal evaluation terminal outputs all the examination factors for the target object to solve;
[0060] An acquisition module, used for acquiring, for each target object, information on the processing result of the target object for each of the inspection elements;
[0061] An input module, used to input all the response processing result information corresponding to the target object into the target deep learning model to obtain the response processing result analysis result; wherein the response processing result information includes image information obtained by image acquisition of the target object, voice information obtained by voice acquisition of the target object, and multiple information contained in the data stream formed by the image information and voice information of the target object in a specific unit time;
[0062] The determination module is further configured to determine, for each target object, an evaluation result of the target object according to an analysis result of a response processing result of the target object;
[0063] The training module is also used to perform iterative training operations on the target deep learning model in the Metaverse intelligent evaluation system based on the collected evaluation result data corresponding to multiple evaluated objects that have undergone multimodal evaluation within a certain time period and the real-time evaluation multimodal data corresponding to each of the evaluated objects, so as to update the target deep learning model, wherein all the evaluated objects include at least one evaluated object that has undergone autonomous multimodal evaluation through the Metaverse intelligent evaluation system within the certain time period and / or at least one evaluated object that has undergone non-autonomous multimodal evaluation through the Metaverse intelligent evaluation system within the certain time period; the autonomous multimodal evaluation includes a multimodal evaluation without the participation of evaluators, and the non-autonomous multimodal evaluation includes a multimodal evaluation with the participation of corresponding evaluators.
[0064] As an optional implementation, in the second aspect of the present invention, the acquisition module is further used for obtaining, for each target object, recruitment demand information of a target position of a target employer corresponding to the target object after the Metaverse intelligent evaluation system determines the evaluation result of the target object according to the analysis result of the response processing result of the target object in the determination module;
[0065] The input module is further used to input the recruitment demand information of the target position of the target employer corresponding to the target object into a predetermined demand analysis model to obtain the demand analysis result of the target position of the target employer corresponding to the target object;
[0066] The Metaverse Intelligent Assessment System further includes:
[0067] A calculation module is used to calculate, for each target object, a matching degree between an evaluation result of the target object and a demand analysis result of a target position of a target employer corresponding to the target object by using a predetermined matching algorithm, so as to obtain a person-position matching degree between the target object and the target employer corresponding to the target object;
[0068] A generation module, used to generate a job matching result and report between the target object and the target employer corresponding to the target object based on the person-job matching;
[0069] A judgment module is used to judge, for each of the target objects, whether the job matching degree between the target object and the target employer corresponding to the target object meets a preset job matching condition according to the job matching result between the target object and the target employer corresponding to the target object;
[0070] The determination module is further configured to, for each of the target objects, determine the target object as a candidate for recruitment of the target position of the target employer corresponding to the target object when the determination module determines that the job matching degree between the target object and the target employer corresponding to the target object meets the preset job matching condition;
[0071] The generating module is further configured to generate evaluation recommendation information for each target employer according to the evaluation result of each candidate recruitment object corresponding to the target employer;
[0072] The push module is used to push the evaluation recommendation information to the terminal corresponding to each of the candidate recruitment objects and the terminal corresponding to the target employer.
[0073] As an optional embodiment, in the second aspect of the present invention, the determination module is also used to, after each target object is determined as a candidate for recruitment of the target position of the target employer corresponding to the target object, for each associated object matched by the target object, the Metaverse intelligent evaluation platform determines the value parameters of the evaluation information corresponding to the associated object based on the selection factor reply information of the associated object, and determines the remuneration information of the intelligent selection and optimization contribution that matches the associated object based on the information value parameters and a predetermined incentive cost accounting algorithm, and pays the remuneration fee that matches the associated object to the associated object based on the remuneration information.
[0074] As an optional implementation, in the second aspect of the present invention, the Metaverse Intelligent Assessment System further includes:
[0075] A simulation module, which is used for, after the determination module determines, for each target object, the target object as a candidate for recruitment of the target position of the target employer corresponding to the target object, for each candidate for recruitment of the target position of the target employer, the Metaverse intelligent evaluation system performs a remote Metaverse immersive simulated work test on the candidate recruitment object according to the evaluation result of the candidate recruitment object, obtains the simulated work result information of the candidate recruitment object, and performs quantitative scoring on the simulated work result information of the candidate recruitment object according to a preset work result quantification algorithm, to obtain the Metaverse immersive remote simulated work result score of the candidate recruitment object;
[0076] The judgment module is further used to judge whether the remote simulation work achievement score of the candidate recruitment object is greater than or equal to a preset work achievement threshold;
[0077] The generation module is also used to generate a work simulation test report and recommendation information based on the work simulation test results of each candidate recruitment object corresponding to the target employer when the judgment module determines that the remote simulation work performance score of the candidate recruitment object is greater than or equal to the preset work performance threshold, and push the work simulation test report and recommendation information to the terminal corresponding to each candidate recruitment object and the terminal corresponding to the target employer.
[0078] As an optional implementation, in the second aspect of the present invention, the determination module is also used for, in the acquisition module, for each of the target objects, the Metaverse intelligent assessment system obtains the response processing result information of the target object for each of the assessment elements, inputs all the response processing result information corresponding to the target object into the target deep learning model, and before obtaining the response processing result analysis result, for each of the target objects, determines the facial image information of the target object according to the image information corresponding to the target object, and determines the target voice information of the target object according to the voice information corresponding to the target object;
[0079] The input module inputs all the corresponding processing result information corresponding to the target object into the target deep learning model, and the specific method of obtaining the corresponding processing result analysis result includes:
[0080] The facial image information of the target object and the voice information corresponding to the target object are input into the target deep learning model to obtain the corresponding processing result analysis result.
[0081] As an optional implementation, in the second aspect of the present invention, the training module is further used for collecting a micro-expression and micro-movement training data set and a tone and / or tone data training set after the determination module determines the facial image information of the target object according to the image information corresponding to the target object, and determines the target voice information of the target object according to the voice information corresponding to the target object for each of the target objects, and performs training operations on the micro-expression and micro-movement training data set and the tone and / or tone data training set through the predetermined deep learning algorithm to construct a multimodal personality assessment model;
[0082] The calculation module is further used to calculate the model loss parameter of the multimodal personality assessment model;
[0083] The judging module is further used to judge whether the model loss parameter meets the preset model generation condition;
[0084] The input module is further configured to input facial image information, behavior information and target voice information of each target object into the multimodal personality assessment model to obtain a comprehensive assessment result of the target object when the judgment module determines that the model loss parameter satisfies the preset model generation condition;
[0085] The specific manner in which the determination module determines the evaluation result of each target object according to the analysis result of the response processing result of the target object includes:
[0086] For each of the target objects, an assessment result of the target object is determined according to the analysis result of the coping treatment result of the target object and the personality assessment result of the target object.
[0087] As an optional implementation, in the second aspect of the present invention, the determination module is further used to determine the employment demand information of the target position of at least one employer;
[0088] The generating module is further used to generate a set of candidate evaluation questions and inspection tasks based on the employment demand information of the target position of each employer, wherein the set of candidate evaluation questions and inspection tasks includes a plurality of candidate evaluation questions and inspection tasks;
[0089] The specific method of generating the candidate evaluation questions and the inspection task set based on the employment demand information of the target position of each employer includes:
[0090] Acquire an assessment question and inspection task training data set, wherein the assessment question and inspection task training data set includes a plurality of assessment training questions and a training answer corresponding to each of the assessment training questions, and includes a plurality of inspection task training data and a training inspection processing result corresponding to each of the inspection task training data;
[0091] The predetermined candidate question generation model is trained by using a predetermined deep learning algorithm and the evaluation question and inspection task training data set to obtain a question generation model;
[0092] Inputting the employment demand information of the target positions of all the employers into the question generation model to obtain a model output result, and generating a set of alternative evaluation questions and inspection tasks based on the model output result, wherein the model output result includes at least one alternative evaluation question and inspection task;
[0093] Furthermore, the specific manner in which the determination module determines at least one investigation element corresponding to the target object based on the multimodal analysis result of the target object includes:
[0094] Based on the multimodal analysis result of the target object, the inspection elements of the target object are determined from the set of alternative evaluation questions and inspection tasks, wherein the inspection elements include evaluation questions and / or inspection tasks.
[0095] As an optional implementation, in the second aspect of the present invention, the specific manner in which the determination module determines the inspection elements of the target object from the set of alternative evaluation questions and inspection tasks based on the multimodal analysis results of the target object includes:
[0096] Based on the multimodal analysis results, determining at least one evaluation scoring factor corresponding to the target object;
[0097] According to all the evaluation and scoring factors corresponding to the target object, a weight allocation operation is performed on each of the evaluation and scoring factors to obtain a weight value corresponding to each of the evaluation and scoring factors;
[0098] According to the weight value corresponding to each of the evaluation scoring factors, a set of examination elements for the target object is determined from the set of candidate evaluation questions and examination tasks;
[0099] Furthermore, the determination module is further used to obtain, in the input module, for each target object, the response processing result information of the target object for each of the inspection elements, input all the response processing result information corresponding to the target object into the target deep learning model, and determine the number of training data of all the evaluation data pairs included in the evaluation data set before obtaining the response processing result analysis result;
[0100] The judgment module is also used to judge whether the amount of training data is greater than or equal to a preset quantity threshold; when it is judged that the amount of training data is greater than or equal to the preset quantity threshold, the acquisition module is triggered to execute the operation of obtaining the response processing result information of the target object for each of the examination elements for each of the target objects, and the input module is triggered to execute the operation of inputting all the response processing result information corresponding to the target object into the target deep learning model to obtain the response processing result analysis result.
[0101] As an optional implementation, in the second aspect of the present invention, the determination module is further used to collect recruitment information of the target position of at least one employer before determining the employment demand information of the target position of at least one employer, and for each of the employers, determine the recruitment position of the target position of the employer according to the recruitment information of the target position of the employer;
[0102] The acquisition module is further used to acquire target unit information of the target position of the employer;
[0103] The determination module is further configured to determine the job requirement information of the recruitment position according to the target unit information of the target position of the employer; wherein the target unit information includes one or more of the work scenario information of the target position of the employer, the job requirement information of the target position of the employer, and the skill requirement information of the target position of the employer;
[0104] The generating module is further configured to generate, for each of the employers, a job analysis result of the employer's target position according to the job requirement information of the employer's target position;
[0105] The specific manner in which the determination module determines the employment demand information of the target position of at least one employer includes:
[0106] For each of the employers, determining the employment demand information of the target position of the employer according to the job analysis results of the target position of the employer;
[0107] The employment demand information includes one or more of work task information, job qualification information, salary range information, employment time information, and work location information.
[0108] The third aspect of the present invention discloses another metaverse intelligent evaluation system, the system comprising:
[0109] A memory storing executable program code;
[0110] a processor coupled to the memory;
[0111] The processor calls the executable program code stored in the memory to execute the intelligent multimodal evaluation implementation method disclosed in the first aspect of the present invention.
[0112] The fourth aspect of the present invention discloses a computer storage medium, wherein the computer storage medium stores computer instructions, and when the computer instructions are called, they are used to execute the intelligent multimodal evaluation implementation method disclosed in the first aspect of the present invention.
[0113] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:
[0114] In an embodiment of the present invention, a target deep learning model trained to convergence is obtained based on a predetermined evaluation data set and input into a deep learning model, real-time evaluation multimodal data of several target objects to be evaluated are collected and analyzed to obtain multimodal analysis results to determine at least one examination element that the target object needs to answer and output to the terminal corresponding to the target object so that the target object can solve it, the response processing result information of the target object is obtained and the response processing result analysis result is obtained, and based on the collected evaluation result data corresponding to multiple evaluated objects that have been multimodally evaluated within a certain time period and the real-time evaluation multimodal data, the target deep learning model is iteratively trained to update the target deep learning model. It can be seen that the implementation of the present invention can be beneficial to improving evaluation efficiency, as well as improving the stability and reliability of applicant scoring, and further improving the reliability and flexibility of recruitment. BRIEF DESCRIPTION OF THE DRAWINGS
[0115] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0116] Figure 1 It is a flowchart of a method for realizing intelligent multi-modal evaluation disclosed in an embodiment of the present invention;
[0117] Figure 2 It is a flowchart of another intelligent multi-modal evaluation implementation method disclosed in an embodiment of the present invention;
[0118] Figure 3 It is a structural schematic diagram of a metaverse intelligent evaluation system disclosed in an embodiment of the present invention;
[0119] Figure 4 It is a structural schematic diagram of another metaverse intelligent evaluation system disclosed in an embodiment of the present invention;
[0120] Figure 5 It is a structural diagram of another metaverse intelligent evaluation system disclosed in an embodiment of the present invention. DETAILED DESCRIPTION
[0121] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0122] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or terminal including a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units inherent to these processes, methods, products or terminals.
[0123] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.
[0124] The present invention discloses an intelligent multimodal evaluation implementation method and a metaverse intelligent evaluation system, which can help improve evaluation efficiency, as well as improve the stability and reliability of applicant scoring, and further help improve the reliability and flexibility of recruitment. The following are detailed descriptions.
[0125] Embodiment 1
[0126] See also Figure 1 , Figure 1 : is a flow chart of a method for implementing intelligent multi-modal evaluation disclosed in an embodiment of the present invention. Figure 1 The described intelligent multimodal evaluation implementation method can be applied to the Metaverse intelligent evaluation system, can also be applied to the device for implementing intelligent multimodal evaluation, can also be applied to the cloud server or local server for implementing intelligent multimodal evaluation, and the embodiments of the present invention are not limited thereto. Figure 1 As shown, the intelligent multimodal evaluation implementation method may include the following operations:
[0127] 101. The Metaverse Intelligent Assessment System is based on a multimodal data set collected from several users when they use the multimodal intelligent assessment platform for recruitment and assessment. The multimodal data set is input into a predetermined deep learning model for training to obtain a target deep learning model that is trained to convergence.
[0128] In an embodiment of the present invention, an evaluation data set includes an evaluation data pair corresponding to each of several multimodal evaluations obtained in real time from a Metaverse intelligent evaluation system, the evaluation data pair includes evaluation multimodal data and evaluation result data corresponding to the evaluation multimodal data, wherein the evaluation multimodal data includes real-time multimodal data of evaluators and real-time multimodal data of evaluation objects, the evaluation result data includes evaluation results output by at least one evaluator for at least one evaluation object, and the Metaverse intelligent evaluation system includes a target deep learning model.
[0129] In the embodiment of the present invention, optionally, one evaluator may correspond to multiple evaluation objects, and one evaluation object may correspond to multiple evaluators, which is not specifically limited in the embodiment of the present invention. For example, one evaluator may evaluate multiple evaluation objects, and the evaluator outputs the evaluation result for each evaluation object; multiple evaluators may evaluate one evaluation object, and each evaluator outputs the evaluation result for the evaluation object; multiple evaluators may evaluate multiple objects, and each evaluator outputs the evaluation result for each evaluation object.
[0130] In an embodiment of the present invention, optionally, the multimodal data set includes one or more of video data, image data, audio data, text data, and interactive multimedia data (such as tactile data, user operation data, user action data), which is not specifically limited in the embodiment of the present invention.
[0131] In an embodiment of the present invention, optionally, the target deep learning model is applied to a Metaverse intelligent evaluation system, and the Metaverse intelligent evaluation system includes an intelligent evaluation platform and a target device, and the intelligent evaluation platform can realize multi-person online multimodal evaluation.
[0132] In the embodiment of the present invention, it is optional to simulate and fit the multimodal talent selection elements such as the work scenario, work tasks, job requirements and job qualifications that the interviewed candidates will be engaged in through the Metaverse Intelligent Assessment and Interview System; further, it can be realized through a number of data automatically collected through the market-oriented self-motivation model. The method is applicable to any talent allocation decision that needs to examine whether the candidate is qualified for the new position, including but not limited to work and learning occasions such as recruitment, promotion, talent inventory, training development and team building.
[0133] In an embodiment of the present invention, optionally, the Metaverse intelligent assessment system uses multimodal hardware and software devices and a market-oriented self-motivated platform to collect multimodal data sets of interview input multimodal data and assessment results based on multiple positions from several users who use the platform for interviews or assessments.
[0134] In the embodiment of the present invention, further optionally, before inputting the evaluation data set into the predetermined deep learning model for training, the method further includes:
[0135] Determine the number of data pairs of all evaluation data pairs included in the evaluation data set, and determine whether the number of data pairs is greater than or equal to a preset data number threshold;
[0136] When it is determined that the number of data pairs is greater than or equal to a preset data number threshold, an operation of inputting the evaluation data set into a predetermined deep learning model for training is triggered;
[0137] When it is determined that the number of data pairs is less than the preset data number threshold, the evaluation data set is reacquired, and the operation of determining the number of data pairs of all evaluation data pairs included in the evaluation data set and determining whether the number of data pairs is greater than or equal to the preset data number threshold is triggered again.
[0138] 102. The Metaverse intelligent assessment system collects real-time assessment multimodal data of several target objects to be assessed, and for each target object, analyzes the real-time assessment multimodal data of the target object to obtain a multimodal analysis result of the target object, and based on the multimodal analysis result of the target object, determines at least one assessment factor that needs to correspond to the target object, and outputs all the assessment factors determined for the target object to the multimodal assessment terminal corresponding to the target object, so that the multimodal assessment terminal outputs all the assessment factors for the target object to solve.
[0139] In the embodiment of the present invention, optionally, the number of target objects to be evaluated may be one or more, which is not specifically limited in the embodiment of the present invention. Further optionally, the target object to be evaluated is a target object using the Metaverse Intelligent Evaluation System.
[0140] In an embodiment of the present invention, optionally, the real-time evaluation multimodal data of the target object may be real-time evaluation multimodal data in an evaluation scenario, or may be real-time evaluation multimodal data in an assessment scenario, wherein the assessment scenario may be a work assessment scenario.
[0141] In the embodiment of the present invention, optionally, the examination elements include examination elements corresponding to the target questions that the target object needs to answer or the examination tasks that need to be processed.
[0142] In an embodiment of the present invention, optionally, when the target object needs to deal with multiple target questions or examination elements for the processing results, the output order of each target question or examination element can be output in sequence to the multimodal evaluation terminal or account corresponding to the target object, or can be output all at once to the multimodal evaluation terminal or account corresponding to the target object, and the embodiment of the present invention does not make specific limitations.
[0143] In an embodiment of the present invention, optionally, the multimodal evaluation terminal or account corresponding to the target object includes one or more of a PC, a smart phone, a smart tablet, a smart computer, etc.
[0144] 103. For each target object, the Metaverse Intelligent Assessment System obtains the response processing result information of the target object for each assessment factor, inputs all the response processing result information corresponding to the target object into the target deep learning model, and obtains the response processing result analysis result.
[0145] In an embodiment of the present invention, the processing result analysis result includes image information obtained by capturing the image of the target object, voice information obtained by capturing the voice of the target object, and various information contained in the data stream formed by the image information and voice information of the target object within a specific unit time.
[0146] In an embodiment of the present invention, optionally, the image information of each target object includes one or more of the target object's facial expression information, facial emotion information, and facial eye information; the voice information of each target object includes one or more of the target object's voice semantic information, voice pitch information, voice tone information, and voice emotion information.
[0147] In an embodiment of the present invention, optionally, the multiple information contained in the data stream formed by the image information and the voice information of the target object within a specific unit time includes one or more of the image annotation information, voice annotation information, and time annotation information of the target object during the process of intelligent evaluation through the Metaverse intelligent evaluation system; for example, the multiple information contained in the data stream formed by the image information and the voice information of the target object within a specific unit time includes 16 frames of image information included in 1 second of the target object during the evaluation process.
[0148] In the embodiment of the present invention, optionally, the response processing result analysis result includes the response processing result of all response processing result information corresponding to the target object.
[0149] 104. For each target object, the Metaverse Intelligent Assessment System determines the assessment result of the target object based on the analysis result of the response processing result of the target object.
[0150] In an embodiment of the present invention, optionally, the evaluation result of each target object includes the evaluation result output by the Metaverse Intelligent Evaluation System for the target object.
[0151] In the embodiment of the present invention, optionally, each target object has an evaluation result corresponding to a threshold.
[0152] 105. The Metaverse Intelligent Assessment System performs iterative training operations on the target deep learning model in the Metaverse Intelligent Assessment System based on the collected assessment result data corresponding to multiple assessed objects that have undergone multimodal assessment within a certain time period and the real-time assessment multimodal data corresponding to each assessed object, so as to update the parameters of the target deep learning model.
[0153] In an embodiment of the present invention, all evaluated objects include at least one evaluated object that has undergone autonomous multimodal evaluation through the Metaverse intelligent evaluation system within the certain time period and / or at least one evaluated object that has undergone non-autonomous multimodal evaluation through the Metaverse intelligent evaluation system within the certain time period; autonomous multimodal evaluation includes multimodal evaluation without the participation of evaluators, and non-autonomous multimodal evaluation includes multimodal evaluation with the participation of corresponding evaluators.
[0154] In an embodiment of the present invention, optionally, during the use of the target deep model, training is performed through the evaluation output results of the evaluation object by multiple evaluators using the Metaverse intelligent evaluation system, and through the evaluation results of each target object that is autonomously evaluated using the Metaverse intelligent evaluation system.
[0155] In the embodiment of the present invention, the evaluator may be an evaluator or an assessor, which is not specifically limited in the embodiment of the present invention.
[0156] In an embodiment of the present invention, optionally, the process of performing an iterative training operation on a target deep learning model based on the evaluation result data of multiple evaluated objects and the real-time multimodal data of each evaluated object to update the target deep model can be updated in real time.
[0157] It can be seen that the implementation Figure 1The described intelligent multimodal evaluation implementation method can be trained by the Metaverse intelligent evaluation system based on a predetermined evaluation data set and input into a predetermined deep learning model to obtain a target deep learning model trained to convergence, collect real-time evaluation multimodal data of multiple target objects to be evaluated, analyze the real-time evaluation multimodal data of each target object to obtain a multimodal analysis result of each target object, and determine the target problem or inspection element that requires the target object to respond to the processing result based on the multimodal analysis result, and output all target problems or inspection elements to the multimodal evaluation terminal or account corresponding to the target object, and obtain the target object's response processing result information for each target problem or inspection element, and generate a response processing result analysis result based on the response processing result information to generate an evaluation The results are obtained by continuously iterating training based on the evaluation result data and real-time evaluation multimodal data to update the target deep learning model. Online intelligent evaluation can be realized based on the Metaverse intelligent evaluation system containing the target deep learning model, and multi-person online real-time evaluation can be realized, which is conducive to improving the efficiency and accuracy of the evaluation. In addition, real-time evaluation can be performed by training the target deep learning model to convergence, which is conducive to improving the objectivity and equality of multi-person evaluation, and preventing the phenomenon that the evaluation results of the target object to be evaluated are unequal due to subjective reasons of the evaluators, which is conducive to improving the fairness of the evaluation while improving the efficiency and accuracy of the evaluation, and thus is conducive to achieving the efficiency and accuracy of multiple people's evaluation through the Metaverse intelligent evaluation system and the fairness of multi-person evaluation.
[0158] Embodiment 2
[0159] See also Figure 2 , Figure 2 is a flow chart of another intelligent multi-modal evaluation implementation method disclosed in an embodiment of the present invention. Figure 2 The described intelligent multimodal evaluation implementation method can be applied to a metaverse intelligent evaluation system, to an apparatus for implementing intelligent multimodal evaluation, or to a cloud server or local server for implementing intelligent multimodal evaluation, and the embodiments of the present invention are not limited thereto.
[0160] like Figure 2 As shown, the intelligent multimodal evaluation implementation method may include the following operations:
[0161] 201. The Metaverse Intelligent Evaluation System inputs the evaluation data set into a predetermined deep learning model for training based on a predetermined evaluation data set, thereby obtaining a target deep learning model that is trained to convergence.
[0162] 202. The Metaverse intelligent assessment system collects real-time assessment multimodal data of several target objects to be assessed, and for each target object, analyzes the real-time assessment multimodal data of the target object to obtain a multimodal analysis result of the target object, and based on the multimodal analysis result of the target object, determines at least one assessment factor required for the target object, and outputs all the assessment factors determined for the target object to the multimodal assessment terminal corresponding to the target object, so that the multimodal assessment terminal outputs all the assessment factors for the target object to solve.
[0163] 203. For each target object, the Metaverse Intelligent Assessment System obtains the response processing result information of the target object for each assessment factor, inputs all the response processing result information corresponding to the target object into the target deep learning model, and obtains the response processing result analysis result.
[0164] 204. For each target object, the Metaverse Intelligent Assessment System determines the assessment result of the target object based on the analysis result of the response processing result of the target object.
[0165] 205. For each target object, the Metaverse Intelligent Assessment System obtains the recruitment demand information of the target position of the target employer corresponding to the target object, and inputs the recruitment demand information of the target position of the target employer corresponding to the target object into a predetermined demand analysis model to obtain the demand analysis result of the target position of the target employer corresponding to the target object.
[0166] In the embodiment of the present invention, optionally, the number of target positions of the target employer corresponding to the target object can be one or more, which is not limited in the embodiment of the present invention. Further, the target employer corresponding to the target object is the employer to which the target object submits his resume.
[0167] In an embodiment of the present invention, optionally, the recruitment demand information of the target position of each target employer includes the recruitment position information of the target position of the target employer, the recruitment quantity information corresponding to the recruitment position of the target position of the target employer, and the work content information of the target position of the target employer.
[0168] 206. For each target object, the Metaverse Intelligent Assessment System calculates the matching degree between the assessment result of the target object and the demand analysis result of the target position of the target employer corresponding to the target object through a predetermined matching algorithm, obtains the person-position matching degree between the target object and the target employer corresponding to the target object, and generates the position matching result and report between the target object and the target employer corresponding to the target object based on the person-position matching degree.
[0169] In the embodiment of the present invention, optionally, the matching degree between the evaluation result of the target object and the demand analysis result of the target position of the target employer corresponding to the target object is used to indicate the matching degree between the target object and the target employer corresponding to the target object. The matching degree includes one or more of the matching degree of professional background, matching degree of education, matching degree of work experience, matching degree of work location, matching degree of work hours, matching degree of working time, and matching degree of work environment.
[0170] In the embodiment of the present invention, optionally, the position matching result between the target object and the target employer corresponding to the target object includes a matching value between the target object and the target employer corresponding to the target object.
[0171] 207. For each target object, the Metaverse Intelligent Assessment System determines whether the job matching degree between the target object and the target employer corresponding to the target object meets the preset job matching conditions based on the job matching result between the target object and the target employer corresponding to the target object.
[0172] In the embodiment of the present invention, the job matching result between the target object and the target employer corresponding to the target object includes the matching degree between the target object and the target employer corresponding to the target object.
[0173] In the embodiment of the present invention, optionally, for each target object, determining whether the job matching degree between the target object and the target employer corresponding to the target object meets a preset job matching condition includes:
[0174] According to the preset job matching conditions, determine the preset job matching threshold;
[0175] Determine whether the matching degree between the target object and the target employer corresponding to the target object is greater than or equal to a preset job matching threshold;
[0176] When it is determined that the matching degree between the target object and the target employer corresponding to the target object is greater than or equal to a preset job matching threshold, it is determined that the job matching degree between the target object and the target employer corresponding to the target object meets the preset job matching condition;
[0177] When it is determined that the match degree between the target object and the target employer corresponding to the target object is less than a preset job matching threshold, it is determined that the job matching degree between the target object and the target employer corresponding to the target object does not meet the preset job matching condition.
[0178] 208. For each target object, when the Metaverse Intelligent Assessment System determines that the job matching degree between the target object and the target employer corresponding to the target object meets the preset job matching conditions, the target object is determined as a candidate for recruitment of the target position of the target employer corresponding to the target object.
[0179] In the embodiment of the present invention, further optionally, when it is determined that the position matching degree between the target object and the target employer corresponding to the target object does not satisfy a preset position matching condition, the target object is determined as not matching the target object and the target employer corresponding to the target object.
[0180] In the embodiment of the present invention, optionally, the number of candidate recruitment objects corresponding to the target employer may be one or more, which is not specifically limited in the embodiment of the present invention.
[0181] 209. For each target employer, the Metaverse Intelligent Assessment System generates assessment recommendation information based on the assessment results of each candidate recruitment object corresponding to the target employer, and pushes the assessment recommendation information to the terminal corresponding to each candidate recruitment object and the terminal corresponding to the target employer.
[0182] In an embodiment of the present invention, optionally, the evaluation recommendation information includes a matching result between the target employer and each target object of the target position of the target employer, a position matching degree between the target employer and each target object, and one or more of the recommended positions for each target object in the target employer.
[0183] 210. The Metaverse Intelligent Assessment System performs iterative training operations on the target deep learning model in the Metaverse Intelligent Assessment System based on the collected assessment result data corresponding to multiple assessed objects that have undergone multimodal assessment within a certain time period and the real-time assessment multimodal data corresponding to each assessed object to update the target deep learning model.
[0184] In the embodiment of the present invention, for the detailed description of steps 201 to 204 and step 210, please refer to other descriptions of steps 101 to 105 in the first embodiment, and the embodiment of the present invention will not be repeated.
[0185] It can be seen that the implementation Figure 2The described intelligent multimodal evaluation implementation method can obtain the recruitment demand information of the target position of the target employer corresponding to each target object, and obtain the demand analysis result of the target position of the target employer through a predetermined demand analysis model, calculate the matching degree between the evaluation result of the target object and the demand analysis result of the target position of the target employer corresponding to the target object through a predetermined matching algorithm, and generate the position matching degree based on the position matching degree, and if it is judged that the position matching degree meets the position matching condition, determine the target object as the candidate recruitment object for the target position of the target employer corresponding to the target object, and generate evaluation recommendation information based on the evaluation result of each candidate recruitment object, and push the evaluation recommendation information to the terminal or account corresponding to each candidate recruitment object and the terminal or account corresponding to the target employer, so as to match the target position of the target employer with the target object in combination with the demand analysis model. Analyzing the recruitment needs of target positions is conducive to improving the accuracy and reliability of the needs analysis results, as well as improving the intelligence and efficiency of the needs analysis results; by combining a predetermined matching algorithm, it is conducive to improving the accuracy and reliability of calculating the matching degree between the target object and the target employer, as well as improving the intelligence and efficiency of obtaining the matching degree; and then by generating evaluation recommendation information based on the evaluation results of each candidate recruitment object, it can improve the accuracy and reliability of generating evaluation recommendation information, as well as improve the intelligence and efficiency of generating evaluation recommendation information, and then it is also conducive to improving the experience and comfort of each target object and each target employer in using the Metaverse intelligent evaluation system, and it is also conducive to improving the reliability, flexibility and convenience of each target object and each target employer in evaluating using the Metaverse intelligent evaluation system.
[0186] In an optional embodiment, for each target object, after determining the target object as a candidate for recruitment of a target position of a target employer corresponding to the target object, the method further includes:
[0187] For each associated object that matches each target object, the Metaverse Intelligent Evaluation Platform determines the value parameters of the evaluation information corresponding to the associated object based on the selection factor reply information of the associated object, and determines the remuneration information of the intelligent selection optimization contribution that matches the associated object based on the information value parameters and the predetermined incentive fee accounting algorithm, and pays the associated object the remuneration fee that matches the associated object based on the remuneration information.
[0188] In this optional embodiment, optionally, there is an association relationship between each associated object and the target object. For example, each associated object may be a former colleague of the target object, or each associated object may be a current colleague of the target object, which is not specifically limited in the embodiment of the present invention.
[0189] In this optional embodiment, optionally, the value parameter of the evaluation information corresponding to each associated object is used to represent the value parameter of the reply information output by the associated object.
[0190] In this optional embodiment, optionally, the remuneration information includes a remuneration fee value corresponding to the associated object.
[0191] It can be seen that the implementation of this optional embodiment can determine the corresponding value parameters according to the selection factor response information of each associated object, and determine the matching remuneration information in combination with a predetermined incentive fee calculation algorithm, and pay the matching remuneration fee to the associated object based on the remuneration information. It can accumulate and integrate the valuable expertise of users through the platform, and can motivate the associated objects corresponding to each target object to participate in the selection factor response process through self-motivation. It can improve the accuracy and reliability of the evaluation results of each target object, and is conducive to improving the objectivity and effectiveness of the evaluation results of each target object.
[0192] In another optional embodiment, for each target object, after determining the target object as a candidate for recruitment of a target position of a target employer corresponding to the target object, the method further includes:
[0193] For each candidate for recruitment of a target position of a target employer, the Metaverse Intelligent Assessment System performs a remote Metaverse Immersive Simulated Work Test on the candidate based on the assessment result of the candidate, obtains the simulated work performance information of the candidate, and performs quantitative scoring on the simulated work performance information of the candidate according to a preset work performance quantification algorithm, obtains the Metaverse Immersive Remote Simulated Work Performance Score of the candidate, and determines whether the remote simulated work performance score of the candidate is greater than or equal to the preset work performance threshold;
[0194] For each candidate recruitment object for the target position of each target employer, when the Metaverse Intelligent Assessment System determines that the remote simulation work performance score of the candidate recruitment object is greater than or equal to the preset work performance threshold, the Metaverse Intelligent Assessment System generates a work simulation test report and recommendation information based on the work simulation test results of each candidate recruitment object corresponding to the target employer, and pushes the work simulation test report and recommendation information to the terminal corresponding to each candidate recruitment object and the terminal corresponding to the target employer.
[0195] In this optional embodiment, optionally, the simulated work results information of each candidate recruitment object includes the candidate recruitment object's simulated work time information, simulated work output information, simulated work efficiency information, and simulated work quantity information.
[0196] In this optional embodiment, further optionally, when it is determined that the remote simulation work performance score of the candidate recruitment object is less than a preset work performance threshold, the candidate recruitment object is eliminated.
[0197] In this optional embodiment, optionally, the preset work result quantification algorithm may include a deep learning algorithm.
[0198] In this optional embodiment, optionally, performing a remote computer simulation work test on each candidate for recruitment can be achieved through one or more of the AR devices, VR devices or MR devices included in the Metaverse intelligent assessment system.
[0199] It can be seen that the implementation of this optional embodiment can perform remote computer simulation work tests on candidate recruitment objects according to the evaluation results of each candidate recruitment object to obtain simulated work results information, and quantify the candidate recruitment objects according to the preset work results quantification algorithm to obtain remote simulation work results scores. If it is judged that the remote simulation work results score of the candidate recruitment object is greater than or equal to the work results threshold, work simulation test recommendation information is generated according to the work simulation test results of each candidate recruitment object corresponding to the target employer, and the work simulation test recommendation information is pushed to the terminal or account corresponding to each candidate recruitment object and the terminal or account corresponding to the target employer. It can be combined with one or more devices among AR devices, VR devices or MR devices included in the Metaverse intelligent evaluation system to perform simulation of the candidate recruitment objects and perform remote computer simulation work tests. Determining the simulated work results information of each candidate for recruitment can realize work simulation for each candidate for recruitment, which is conducive to improving the intelligence and convenience of predicting the work results of the candidate for recruitment, and is conducive to improving the accuracy and reliability of obtaining the predicted work results of the candidate for recruitment, and by quantitatively scoring the simulated work results information according to the work results quantification algorithm, it is conducive to improving the accuracy of the quantitative score of each predicted candidate for recruitment, and by being able to generate and push evaluation recommendation information only when it is judged that the work results score of the candidate for recruitment is greater than or equal to the work results threshold, it is conducive to improving the accuracy and reliability of determining the candidate for recruitment, and is conducive to improving the accuracy and reliability of pushing evaluation recommendation information, and is also conducive to improving the accuracy and reliability of the push evaluation recommendation information received by the target employer and each candidate for recruitment.
[0200] In another optional embodiment, for each target object, the Metaverse intelligent assessment system obtains the response processing result information of the target object for each assessment factor, inputs all the response processing result information corresponding to the target object into the target deep learning model, and before obtaining the response processing result analysis result, the method further includes:
[0201] For each target object, the Metaverse intelligent assessment system determines the facial image information of the target object based on the image information corresponding to the target object, and determines the target voice information of the target object based on the voice information corresponding to the target object;
[0202] Among them, all the corresponding processing result information corresponding to the target object is input into the target deep learning model to obtain the corresponding processing result analysis results, including:
[0203] The facial image information of the target object and the voice information corresponding to the target object are input into the target deep learning model to obtain the corresponding processing result analysis result.
[0204] In this optional embodiment, optionally, the facial image information corresponding to the target object includes the target object's eye information, the target object's expression information, the target object's eye movement feature information, the target object's head movement feature information, and the target object's micro-expression information; the target voice information corresponding to the target object includes the target object's voice semantic information, the target object's volume information, the target object's intonation information, the target object's voice intonation information, the target object's voice tone information, the target object's voice speed information, and the target object's voice text information.
[0205] In this optional embodiment, the target speech information optionally includes semantic information marked by the existing speech-to-text conversion and other information (such as tone information, emotional information) that is not marked by the existing speech-to-text conversion.
[0206] In this optional embodiment, the processing result analysis result optionally includes the processing result analysis result of the target object, wherein the processing result analysis result includes the processing result text analysis result, the processing result expression analysis result, and the processing result speed analysis result of the target object.
[0207] It can be seen that implementing this optional embodiment can determine the facial image information and target voice information of each target object, and convert the target voice information into target text information, input the facial image information and target text information of the target object into the target deep learning model, and obtain the corresponding processing result analysis result, which can perform targeted analysis on the image information and voice information, and can improve the accuracy and reliability of the facial image information and target voice information, and analyze the facial image information and target voice information in combination with the target deep learning model to obtain the corresponding processing result analysis result, which is conducive to improving the accuracy and reliability of the corresponding processing result analysis result, as well as improving the intelligence and efficiency of the corresponding processing result analysis result, and then It is beneficial to improve the accuracy and reliability of generating the evaluation results for each target object, and it is beneficial to improve the intelligence and efficiency of generating the evaluation results for each target object.
[0208] In yet another optional embodiment, for each target object, after the Metaverse intelligent assessment system determines the facial image information of the target object based on the image information corresponding to the target object, and determines the target voice information of the target object based on the voice information corresponding to the target object, the method further includes:
[0209] The Metaverse Intelligent Assessment System collects a micro-expression and micro-movement training data set and a voice intonation and / or tone data training set, and performs a training operation on the micro-expression and micro-movement training data set and the voice intonation and / or tone data training set through a predetermined deep learning algorithm to construct a multimodal personality assessment model;
[0210] The Metaverse Intelligent Assessment System calculates the model loss parameters of the multimodal personality assessment model and determines whether the model loss parameters meet the preset model generation conditions;
[0211] When the Metaverse Intelligent Assessment System determines that the model loss parameters meet the preset model generation conditions, for each target object, the facial image information, behavior information and target voice information of the target object are input into the multimodal personality assessment model to obtain a comprehensive assessment result of the target object;
[0212] For each target object, the Metaverse Intelligent Assessment System determines the assessment result of the target object based on the analysis result of the response processing result of the target object, including:
[0213] For each target object, the Metaverse Intelligent Assessment System determines the assessment result of the target object based on the analysis results of the response processing results of the target object and the multimodal personality assessment results of the target object.
[0214] In this optional embodiment, further optionally, when it is determined that the model loss parameters do not meet the preset model generation conditions, the operation of re-executing the training operation of the micro-expression training data set and the intonation and / or tone data training set by a predetermined deep learning algorithm to construct an operation of obtaining a multimodal personality assessment model, and triggering the operation of calculating the model loss parameters of the multimodal personality assessment model to determine whether the model loss parameters meet the preset model generation conditions; or,
[0215] According to the model loss parameters, the model update parameters are generated, and the model update operation is performed on the multimodal personality assessment model based on the model update parameters, and the operation of calculating the model loss parameters of the multimodal personality assessment model is re-triggered to determine whether the model loss parameters meet the preset model generation conditions.
[0216] In this optional embodiment, optionally, the micro-expression training data set includes a plurality of micro-expression training data; the intonation and / or tone data training set includes a plurality of intonation and / or tone training data.
[0217] In this optional embodiment, optionally, the multimodal personality assessment result of each target object includes personality assessment information of the target object; such as an extroverted personality, an introverted personality, etc.
[0218] In this optional embodiment, the evaluation result of the target object may include an analysis result of the coping result of the target object and a personality evaluation result of the target object.
[0219] In this optional embodiment, optionally, determining whether the model loss parameter satisfies a preset model generation condition includes:
[0220] Determine whether the model loss parameter is greater than or equal to the preset model loss threshold;
[0221] When it is determined that the model loss parameter is greater than or equal to a preset model loss threshold, determining that the model loss parameter does not meet a preset model generation condition;
[0222] When it is determined that the model loss parameter is less than a preset model loss threshold, it is determined that the model loss parameter meets the preset model generation condition.
[0223] It can be seen that the implementation of this optional embodiment can collect a micro-expression training data set and a voice intonation data training set to construct a multimodal personality assessment model, calculate the model loss parameters of the multimodal personality assessment model and determine whether the model loss parameters meet the preset model generation conditions. If they meet the conditions, the facial image information and target voice information of the target object are input into the multimodal personality assessment model to obtain the multimodal personality assessment result of the target object, and the assessment result of the target object is determined according to the analysis result of the target object's response processing result and the multimodal personality assessment result. The assessment result of the target object can be comprehensively determined by combining various assessment results, which is conducive to improving the accuracy and reliability of determining the assessment result of the target object, as well as improving the intelligence and efficiency of determining the assessment result of the target object.
[0224] In yet another optional embodiment, the method further includes:
[0225] The Metaverse Intelligent Assessment System determines the employment demand information of the target position of at least one employer, and generates a set of alternative assessment questions and inspection tasks based on the employment demand information of the target position of each employer, wherein the set of alternative assessment questions and inspection tasks includes a number of alternative assessment questions and inspection tasks;
[0226] Among them, based on the employment demand information of each employer's target position, a set of alternative assessment questions and inspection tasks are generated, including:
[0227] The Metaverse Intelligent Assessment System obtains an assessment question and inspection task training data set, wherein the assessment question and inspection task training data set includes a plurality of assessment training questions and a training answer corresponding to each assessment training question, and includes a plurality of inspection task training data and a training inspection processing result corresponding to each inspection task training data;
[0228] The Metaverse Intelligent Assessment System uses a predetermined deep learning algorithm and assessment question and inspection task training data set to train a predetermined candidate question generation model to obtain a question generation model;
[0229] The Metaverse Intelligent Assessment System inputs the employment demand information of the target positions of all employers into the question generation model, obtains the model output results, and generates a set of alternative assessment questions and inspection tasks based on the model output results, wherein the model output results include at least one alternative assessment question and inspection task;
[0230] And, based on the multimodal analysis result of the target object, determining at least one investigation factor that the target object needs to correspond to, including:
[0231] Based on the multimodal analysis results of the target object, the Metaverse Intelligent Assessment System determines the assessment factors of the target object from a set of alternative assessment questions and assessment tasks, wherein the assessment factors include assessment questions and / or assessment tasks.
[0232] In this optional embodiment, optionally, after performing model training on the predetermined candidate question generation model to obtain the question generation model, the method further includes:
[0233] Calculate the model loss parameter of the problem generation model, and determine whether the model loss parameter is greater than or equal to a preset model loss threshold;
[0234] When it is determined that the model loss parameter is greater than or equal to a preset model loss threshold, re-executing the operation of performing model training on the predetermined candidate question generation model to obtain the question generation model;
[0235] When it is determined that the model loss parameter is less than the preset model loss threshold, an operation is triggered to input the employment demand information of the employer's target position into the problem generation model to obtain the model output result.
[0236] It can be seen that the implementation of this optional embodiment can determine the employment demand information of the target position of the employer to generate a set of alternative evaluation questions and inspection tasks, obtain the evaluation question and inspection task training data set and train the alternative question generation model through a predetermined deep learning algorithm to obtain a question generation model, and input the employment demand information of the target position of the employer into the question generation model to obtain the model output result and then generate a set of alternative evaluation questions and inspection tasks, and determine the target question or inspection element of the target object from the set of alternative evaluation questions and inspection tasks based on the multimodal analysis result of the target object, which can improve the comprehensiveness and accuracy of the generated set of alternative evaluation questions and inspection tasks, and by combining the evaluation question and inspection task The inspection task training data set and the deep learning algorithm are used together to train the alternative question generation model to obtain the question generation model, which can improve the intelligence and efficiency of model training as well as the intelligence and efficiency of the question generation model, and is also beneficial to improving the accuracy and reliability of the question generation model. By combining the multimodal analysis results of the target object and the alternative evaluation questions and inspection task set to determine the target problem or inspection element of the target object, the accuracy and reliability of the target problem or inspection element of the target object can be determined in a targeted manner, and it can be beneficial to improve the matching degree between the determined target problem or inspection element and the target object, thereby helping to improve the accuracy and intelligence of the user's intelligent evaluation through the Metaverse intelligent evaluation system.
[0237] In another optional embodiment, the Metaverse Intelligent Assessment System determines the assessment factors of the target object from the set of candidate assessment questions and assessment tasks based on the multimodal analysis results of the target object, including:
[0238] The Metaverse Intelligent Assessment System determines at least one assessment scoring factor corresponding to the target object based on the multimodal analysis results;
[0239] The Metaverse Intelligent Assessment System performs a weight allocation operation on each assessment scoring factor according to all assessment scoring factors corresponding to the target object, and obtains the weight value corresponding to each assessment scoring factor;
[0240] The Metaverse Intelligent Assessment System determines the set of assessment factors for the target object from the set of candidate assessment questions and assessment tasks according to the weight value corresponding to each assessment scoring factor;
[0241] And, for each target object, the Metaverse Intelligent Assessment System obtains the response processing result information of the target object for each assessment factor, inputs all the response processing result information corresponding to the target object into the target deep learning model, and obtains the response processing result analysis result. The method also includes:
[0242] The Metaverse Intelligent Evaluation System determines the amount of training data for all evaluation data pairs included in the evaluation data set, and determines whether the amount of training data is greater than or equal to a preset quantity threshold;
[0243] When the Metaverse Intelligent Assessment System determines that the amount of training data is greater than or equal to the preset threshold, it triggers the execution of the Metaverse Intelligent Assessment System for each target object, obtains the response processing result information of the target object for each assessment factor, inputs all the response processing result information corresponding to the target object into the target deep learning model, and obtains the response processing result analysis result.
[0244] In this optional embodiment, optionally, the evaluation scoring factors corresponding to the target object include one or more of expression scoring factors, intonation scoring factors, speaking speed scoring factors, answer time scoring factors, answer content scoring factors, answer speed scoring uniformity, tone scoring factors, and micro-expression scoring factors.
[0245] In this optional embodiment, optionally, for example, if the weight value of the answer content scoring factor is the highest, then an examination element that matches the answer content scoring factor is determined from the set of alternative evaluation questions and examination tasks.
[0246] In this optional embodiment, further optionally, when it is determined that the amount of training data is less than a preset threshold value, the process can be terminated, or the training data can be reacquired and the operation of determining whether the amount of training data is greater than or equal to the preset threshold value can be re-executed.
[0247] It can be seen that the implementation of this optional embodiment can determine at least one evaluation and scoring factor of the target object based on the multimodal analysis results, and determine the weight value corresponding to each evaluation and scoring factor, and determine the target problem or examination element according to the weight value corresponding to each evaluation and scoring factor, and can only execute the operation of inputting the corresponding response processing result information of the target object into the target deep learning model to obtain the response processing result analysis result when it is judged that the number of training data is greater than or equal to the preset number threshold, and can determine the target problem or examination element according to the evaluation and scoring factors of the target object and the corresponding weight value, which is conducive to improving the accuracy and reliability of determining the target problem or examination element, and can determine the target problem or examination element in combination with the weight value of each evaluation and scoring factor, which is conducive to improving the pertinence and intelligence of determining the target problem or examination element; and can only execute the operation of inputting a large amount of information into the target deep learning model to obtain the response processing result analysis result when it is judged that the number of training data is greater than or equal to the preset number threshold, which can improve the accuracy and reliability of the target deep learning model, and can improve the intelligence of the target deep learning model, thereby helping to improve the accuracy and reliability of the response processing result analysis result obtained based on the target deep learning model.
[0248] In yet another optional embodiment, before the Metaverse intelligent assessment system determines the employment demand information of the target position of at least one employer, the method further includes:
[0249] The Metaverse Intelligent Assessment System collects recruitment information of target positions of at least one employer, and for each employer, determines the recruitment position of the employer's target position based on the recruitment information of the employer's target position, and obtains the target unit information of the employer's target position, and determines the job requirement information of the recruitment position based on the target unit information of the employer's target position; wherein the target unit information includes one or more of the work scenario information of the employer's target position, the job requirement information of the employer's target position, and the skill requirement information of the employer's target position;
[0250] For each employer, the Metaverse Intelligent Assessment System generates job analysis results for the employer's target position based on the job demand information of the employer's target position;
[0251] Among them, the Metaverse Intelligent Assessment System determines the employment demand information of at least one employer's target position, including:
[0252] For each employer, the Metaverse Intelligent Assessment System determines the employment demand information of the employer's target position based on the job analysis results of the employer's target position;
[0253] The employment demand information includes one or more of work task information, job qualification information, salary range information, employment time information, and work location information.
[0254] In this optional embodiment, the number of recruitment positions for the target positions of each employer may be one or more, which is not specifically limited in the embodiment of the present invention.
[0255] In this optional embodiment, the job requirement information of each recruitment position may optionally include one or more of the job working hours information of the recruitment position, the job education information of the recruitment position, the job professional background information of the recruitment position, the work experience information of the recruitment position, and the work location information of the recruitment position.
[0256] In this optional embodiment, optionally, the job analysis results of the employer's target job include job quantity requirement information, job skill requirement information, and job professional background requirement information of the employer's target job.
[0257] It can be seen that the implementation of this optional embodiment can collect the recruitment information of at least one employer's target position and determine the recruitment position of the employer's target position, and obtain the target unit information of the employer's target position to determine the job requirement information of the recruitment position, generate the job analysis results of the employer's target position according to the job requirement information of the recruitment position of each employer's target position, and determine the employment requirement information of the employer's target position according to the job analysis results of the employer's target position, and can comprehensively generate the job analysis results of the employer's target position through recruitment information and job requirement information, which is conducive to improving the accuracy and reliability of generating the job analysis results of the employer's target position, and is conducive to improving the intelligence of generating the job analysis results of the employer's target position, so that it can also determine the employment requirement information of the employer's target position according to the job analysis results of the employer's target position, which is conducive to improving the accuracy and reliability of the employment requirement information of the employer's target position, and further It is also conducive to improving the accuracy and reliability of the subsequent generation of the matching degree between each target object and the employer.
[0258] Embodiment 3
[0259] See also Figure 3 , Figure 3 Schematic diagram of the structure of a metaverse intelligent evaluation system disclosed in an embodiment of the present invention. Figure 3 As shown, the Metaverse intelligent evaluation system includes a multimodal intelligent evaluation platform with deep learning capabilities, a target device and a set of talent selection elements. The target device includes one or more of a VR device, an AR device, and an MR device. The Metaverse intelligent evaluation system includes a multimodal intelligent evaluation platform and / or a target device. The Metaverse intelligent evaluation system includes:
[0260] The training module 301 is used to input the multimodal data set into a predetermined deep learning model for training based on the multimodal data set collected from several users when using the multimodal intelligent evaluation platform for recruitment evaluation, so as to obtain a target deep learning model trained to convergence; wherein the multimodal evaluation data set includes a multimodal evaluation data pair corresponding to each of several multimodal evaluations obtained in real time from the Metaverse Intelligent Evaluation System, and the evaluation data pair includes evaluation multimodal data and evaluation result data corresponding to the evaluation multimodal data, wherein the evaluation multimodal data includes real-time multimodal data of the evaluator and real-time multimodal data of the evaluation object, and the evaluation result data includes the evaluation result output by at least one evaluator for at least one evaluation object, and the Metaverse Intelligent Evaluation System includes the target deep learning model;
[0261] The acquisition module 302 is used to acquire real-time evaluation multimodal data of a plurality of target objects to be evaluated;
[0262] The analysis module 303 is used to analyze the real-time evaluation multimodal data of each target object to obtain a multimodal analysis result of the target object;
[0263] A determination module 304 is used to determine at least one investigation element that needs to correspond to the target object based on the multimodal analysis result of the target object;
[0264] An output module 305, configured to output all the examination factors determined for the target object to a multimodal evaluation terminal corresponding to the target object, so that the multimodal evaluation terminal outputs all the examination factors for the target object to solve;
[0265] An acquisition module 306 is used to acquire, for each target object, the processing result information of the target object for each inspection element;
[0266] An input module 307 is used to input all the corresponding processing result information corresponding to the target object into the target deep learning model to obtain the corresponding processing result analysis result; wherein the corresponding processing result information includes image information obtained by collecting images of the target object, voice information obtained by collecting voices of the target object, and a variety of information contained in a data stream formed by the image information and voice information of the target object within a specific unit time;
[0267] The determination module 304 is further used to determine, for each target object, an evaluation result of the target object according to the analysis result of the corresponding processing result of the target object;
[0268] The training module 301 is also used to perform iterative training operations on the target deep learning model in the Metaverse intelligent evaluation system based on the collected evaluation result data corresponding to multiple evaluated objects that have undergone multimodal evaluation within a certain time period and the real-time evaluation multimodal data corresponding to each evaluated object, so as to update the target deep learning model, wherein all evaluated objects include at least one evaluated object that has undergone autonomous multimodal evaluation through the Metaverse intelligent evaluation system within the certain time period and / or at least one evaluated object that has undergone non-autonomous multimodal evaluation through the Metaverse intelligent evaluation system within the certain time period; autonomous multimodal evaluation includes multimodal evaluation without the participation of evaluators, and non-autonomous multimodal evaluation includes multimodal evaluation with the participation of corresponding evaluators.
[0269] It can be seen that the implementation Figure 3 The described device can be trained by the Metaverse Intelligent Evaluation System based on a predetermined evaluation data set and input into a predetermined deep learning model to obtain a target deep learning model trained to convergence, collect real-time evaluation multimodal data of multiple target objects to be evaluated, analyze the real-time evaluation multimodal data of each target object to obtain a multimodal analysis result of each target object, and determine the target problem or inspection element that requires the target object to respond to the processing result based on the multimodal analysis result and output all the target problems or inspection elements to the multimodal evaluation terminal or account corresponding to the target object, and obtain the target object's response processing result information for each target problem or inspection element, and generate a response processing result analysis result based on the response processing result information and then generate an evaluation result, and Based on the evaluation result data and real-time evaluation multimodal data, continuous iterative training is carried out to update the target deep learning model. Online intelligent evaluation can be realized based on the Metaverse intelligent evaluation system including the target deep learning model. Online real-time evaluation by multiple people can be realized, which is conducive to improving the efficiency and accuracy of evaluation. In addition, real-time evaluation can be performed by training the target deep learning model to convergence, which is conducive to improving the objectivity and equality of multi-person evaluation and preventing the phenomenon that unequal evaluation results are obtained for the target object to be evaluated due to subjective reasons of the evaluators. This is conducive to improving the fairness of evaluation while improving the efficiency and accuracy of evaluation, and thus is conducive to achieving the efficiency and accuracy of evaluation by multiple people at the same time through the Metaverse intelligent evaluation system and the fairness of multi-person evaluation.
[0270] In an optional embodiment, if Figure 4As shown, the acquisition module 306 is further used to obtain the recruitment demand information of the target position of the target employer corresponding to the target object for each target object after the determination module 304 determines the evaluation result of the target object according to the analysis result of the response processing result of the target object for each target object;
[0271] The input module 307 is further used to input the recruitment demand information of the target position of the target employer corresponding to the target object into a predetermined demand analysis model to obtain the demand analysis result of the target position of the target employer corresponding to the target object;
[0272] The Metaverse Intelligent Assessment System also includes:
[0273] A calculation module 308 is used to calculate, for each target object, the matching degree between the evaluation result of the target object and the demand analysis result of the target position of the target employer corresponding to the target object by using a predetermined matching algorithm, so as to obtain the person-position matching degree between the target object and the target employer corresponding to the target object;
[0274] A generation module 309 is used to generate a job matching result and a report between the target object and the target employer corresponding to the target object based on the person-job matching;
[0275] The judgment module 310 is used to judge, for each target object, whether the job matching degree between the target object and the target employer corresponding to the target object meets the preset job matching condition according to the job matching result between the target object and the target employer corresponding to the target object;
[0276] The determination module 304 is further configured to determine, for each target object, when the determination module 310 determines that the job matching degree between the target object and the target employer corresponding to the target object meets a preset job matching condition, determine the target object as a candidate for recruitment of the target job of the target employer corresponding to the target object;
[0277] The generating module 309 is further used to generate evaluation recommendation information for each target employer according to the evaluation result of each candidate recruitment object corresponding to the target employer;
[0278] The push module 311 is used to push the evaluation recommendation information to the terminal corresponding to each candidate recruitment object and the terminal corresponding to the target employer.
[0279] It can be seen that the implementation Figure 4The described device can obtain the recruitment demand information of the target position of the target employer corresponding to each target object, and obtain the demand analysis result of the target position of the target employer through a predetermined demand analysis model, calculate the matching degree between the evaluation result of the target object and the demand analysis result of the target position of the target employer corresponding to the target object through a predetermined matching algorithm, and generate the position matching degree based on the position matching degree, and if it is judged that the position matching degree meets the position matching condition, determine the target object as the candidate recruitment object for the target position of the target employer corresponding to the target object, and generate evaluation recommendation information based on the evaluation result of each candidate recruitment object, and push the evaluation recommendation information to the terminal corresponding to each candidate recruitment object or the terminal corresponding to the target employer, and can combine the demand analysis model to match the recruitment demand of the target position of the target employer. The needs analysis is beneficial to improving the accuracy and reliability of the needs analysis results, as well as the intelligence and efficiency of the needs analysis results; by combining a predetermined matching algorithm, it is beneficial to improving the accuracy and reliability of calculating the matching degree between the target object and the target employer, as well as the intelligence and efficiency of obtaining the matching degree; and then, by generating evaluation recommendation information according to the evaluation results of each candidate recruitment object, it can improve the accuracy and reliability of generating evaluation recommendation information, as well as the intelligence and efficiency of generating evaluation recommendation information, and then it is also beneficial to improve the experience and comfort of each target object and each target employer in using the Metaverse intelligent evaluation system, and it is also beneficial to improve the reliability, flexibility and convenience of each target object and each target employer in performing evaluations using the Metaverse intelligent evaluation system.
[0280] In another optional embodiment, Figure 4 As shown, the determination module 304 is also used to, for each target object, after determining the target object as a candidate for recruitment of the target position of the target employer corresponding to the target object, for each associated object that matches each target object, the Metaverse Intelligent Evaluation Platform determines the value parameters of the evaluation information corresponding to the associated object based on the selection factor reply information of the associated object, and determines the remuneration information of the intelligent selection optimization contribution that matches the associated object based on the information value parameters and a predetermined incentive cost accounting algorithm, and pays the associated object the remuneration fee that matches the associated object based on the remuneration information.
[0281] It can be seen that the implementation Figure 4The described device can determine the corresponding value parameters based on the selection factor response information of each associated object, and determine the matching remuneration information in combination with a predetermined incentive fee calculation algorithm, and pay the matching remuneration fee to the associated object based on the remuneration information. It can motivate the associated objects corresponding to each target object to participate in the selection factor response process through self-motivation, and can improve the accuracy and reliability of the evaluation results of each target object, as well as help improve the objectivity and effectiveness of the evaluation results of each target object.
[0282] In another optional embodiment, Figure 4 As shown, the Metaverse Intelligent Assessment System also includes:
[0283] The simulation module 312 is used for, after the determination module 304 determines, for each target object, that the target object is a candidate for recruitment of the target position of the target employer corresponding to the target object, for each candidate for recruitment of the target position of the target employer, the Metaverse intelligent evaluation system performs a remote Metaverse immersive simulated work test on the candidate recruitment object according to the evaluation result of the candidate recruitment object, obtains the simulated work result information of the candidate recruitment object, and performs quantitative scoring on the simulated work result information of the candidate recruitment object according to a preset work result quantification algorithm, to obtain the Metaverse immersive remote simulated work result score of the candidate recruitment object;
[0284] The judgment module 310 is further used to judge whether the remote simulation work achievement score of the candidate recruitment object is greater than or equal to a preset work achievement threshold;
[0285] The generation module 309 is also used to generate a work simulation test report and recommendation information based on the work simulation test results of each candidate recruitment object corresponding to the target employer when the judgment module 310 determines that the remote simulation work performance score of the candidate recruitment object is greater than or equal to a preset work performance threshold, and push the work simulation test report and recommendation information to the terminal corresponding to each candidate recruitment object and the terminal corresponding to the target employer.
[0286] It can be seen that the implementation Figure 4The described device can perform remote computer simulated work tests on the candidate recruitment objects according to the evaluation results of each candidate recruitment object to obtain simulated work achievement information, and perform quantitative scoring on the candidate recruitment objects according to a preset work achievement quantification algorithm to obtain a work achievement score. If it is judged that the work achievement score of the candidate recruitment object is greater than or equal to the work achievement threshold, the device generates evaluation recommendation information according to the evaluation results of the candidate recruitment object corresponding to the target employer, and pushes the evaluation recommendation information to the terminal or account corresponding to each candidate recruitment object and the terminal or account corresponding to the target employer. The device can determine the simulated work achievement information of each candidate recruitment object in combination with the remote computer simulated work test, and can realize work simulation for each candidate recruitment object. It is conducive to improving the intelligence and convenience of predicting the work results of candidate recruitment objects, and is conducive to improving the accuracy and reliability of obtaining the work results of predicted candidate recruitment objects, and by quantitatively scoring the simulated work results information according to the work results quantification algorithm, it is conducive to improving the accuracy of the quantitative score of each predicted candidate recruitment object, and by being able to generate and push evaluation recommendation information only when it is judged that the work result score of the candidate recruitment object is greater than or equal to the work result threshold, it is conducive to improving the accuracy and reliability of determining the candidate recruitment objects, and is conducive to improving the accuracy and reliability of pushing evaluation recommendation information, and is also conducive to improving the accuracy and reliability of the target employer and each candidate recruitment object receiving the pushed evaluation recommendation information.
[0287] In yet another optional embodiment, Figure 4 As shown, the determination module 304 is also used for the Metaverse Intelligent Assessment System to obtain the response processing result information of the target object for each assessment factor for each target object in the acquisition module 306, input all the response processing result information corresponding to the target object into the target deep learning model, and before obtaining the response processing result analysis result, for each target object, determine the facial image information of the target object according to the image information corresponding to the target object, and determine the target voice information of the target object according to the voice information corresponding to the target object;
[0288] Among them, the input module 307 inputs all the corresponding processing result information corresponding to the target object into the target deep learning model, and the specific method of obtaining the corresponding processing result analysis result includes:
[0289] The facial image information of the target object and the voice information corresponding to the target object are input into the target deep learning model to obtain the corresponding processing result analysis result.
[0290] It can be seen that the implementation Figure 4The described device can determine the facial image information and target voice information of each target object, and convert the target voice information into target text information, input the facial image information and target text information of the target object into the target deep learning model, and obtain the corresponding processing result analysis result, and can perform targeted analysis on the image information and voice information, and can improve the accuracy and reliability of the facial image information and target voice information, and analyze the facial image information and target voice information in combination with the target deep learning model to obtain the corresponding processing result analysis result, which is conducive to improving the accuracy and reliability of the corresponding processing result analysis result, as well as improving the intelligence and efficiency of the corresponding processing result analysis result, and then It is beneficial to improve the accuracy and reliability of generating the evaluation results of each target object, and it is beneficial to improve the intelligence and efficiency of generating the evaluation results of each target object.
[0291] In yet another optional embodiment, Figure 4 As shown, the training module 301 is also used for collecting a micro-expression and micro-movement training data set and a tone and / or tone data training set after the determination module 304 determines the facial image information of the target object according to the image information corresponding to the target object, and determines the target voice information of the target object according to the voice information corresponding to the target object for each target object, and performs a training operation on the micro-expression and micro-movement training data set and the tone and / or tone data training set through a predetermined deep learning algorithm to construct a multimodal personality assessment model;
[0292] The calculation module 308 is also used to calculate the model loss parameters of the multimodal personality assessment model;
[0293] The judgment module 310 is also used to judge whether the model loss parameter meets the preset model generation condition;
[0294] The input module 307 is also used to input the facial image information, behavior information and target voice information of each target object into the multimodal personality assessment model to obtain a comprehensive assessment result of the target object when the judgment module 310 judges that the model loss parameter meets the preset model generation condition;
[0295] The specific manner in which the determination module 304 determines the evaluation result of each target object according to the analysis result of the corresponding processing result of the target object includes:
[0296] For each target object, an assessment result of the target object is determined according to the analysis result of the coping treatment result of the target object and the personality assessment result of the target object.
[0297] It can be seen that the implementation Figure 4The described device can collect a micro-expression training data set and a voice intonation data training set to construct a multimodal personality assessment model, calculate the model loss parameters of the multimodal personality assessment model and determine whether the model loss parameters meet the preset model generation conditions. If they meet the conditions, the facial image information and the target voice information of the target object are input into the multimodal personality assessment model to obtain the personality assessment result of the target object, and the assessment result of the target object is determined according to the analysis result of the response processing result and the personality assessment result of the target object. The assessment result of the target object can be comprehensively determined by combining the assessment results of various aspects, which is conducive to improving the accuracy and reliability of the assessment result of the target object, as well as improving the intelligence and efficiency of determining the assessment result of the target object.
[0298] In yet another optional embodiment, Figure 4 As shown, the determination module 304 is also used to determine the employment demand information of the target position of at least one employer;
[0299] The generating module 309 is further used to generate a set of candidate evaluation questions and inspection tasks based on the employment demand information of the target position of each employer, wherein the set of candidate evaluation questions and inspection tasks includes a plurality of candidate evaluation questions and inspection tasks;
[0300] The specific method of generating the candidate evaluation questions and the inspection task set based on the employment demand information of the target position of each employer by the generating module 309 includes:
[0301] Acquire an evaluation question and inspection task training data set, wherein the evaluation question and inspection task training data set includes a plurality of evaluation training questions and a training answer corresponding to each evaluation training question, and includes a plurality of inspection task training data and a training inspection processing result corresponding to each inspection task training data;
[0302] Through the predetermined deep learning algorithm and the evaluation question and inspection task training data set, the predetermined candidate question generation model is trained to obtain the question generation model;
[0303] Inputting the employment demand information of the target positions of all employers into the question generation model to obtain the model output results, and generating a set of alternative evaluation questions and inspection tasks based on the model output results, wherein the model output results include at least one alternative evaluation question and inspection task;
[0304] Furthermore, the specific manner in which the determination module 304 determines at least one investigation element corresponding to the target object based on the multimodal analysis result of the target object includes:
[0305] Based on the multimodal analysis result of the target object, the inspection elements of the target object are determined from a set of alternative evaluation questions and inspection tasks, wherein the inspection elements include evaluation questions and / or inspection tasks.
[0306] It can be seen that the implementation Figure 4 The described device can determine the employment demand information of the target position of the employer to generate a set of alternative evaluation questions and inspection tasks, obtain the evaluation question and inspection task training data set and train the alternative question generation model through a predetermined deep learning algorithm to obtain a question generation model, and input the employment demand information of the target position of the employer into the question generation model to obtain the model output result and then generate a set of alternative evaluation questions and inspection tasks, and determine the target question or inspection element of the target object from the set of alternative evaluation questions and inspection tasks based on the multimodal analysis result of the target object, which can improve the comprehensiveness and accuracy of the generated set of alternative evaluation questions and inspection tasks, and by combining the evaluation questions and inspection tasks The training data set and the deep learning algorithm are used together to train the alternative question generation model to obtain the question generation model, which can improve the intelligence and efficiency of model training as well as the intelligence and efficiency of the question generation model, and is also beneficial to improving the accuracy and reliability of the question generation model. By combining the multimodal analysis results of the target object and the set of alternative evaluation questions and inspection tasks to determine the target problem or inspection element of the target object, the accuracy and reliability of the target problem or inspection element of the target object can be determined in a targeted manner, and it can be beneficial to improve the matching degree between the determined target problem or inspection element and the target object, thereby helping to improve the accuracy and intelligence of the user's intelligent evaluation through the Metaverse intelligent evaluation system.
[0307] In yet another optional embodiment, Figure 4 As shown, the specific manner in which the determination module 304 determines the inspection elements of the target object from the candidate evaluation questions and inspection task sets based on the multimodal analysis results of the target object includes:
[0308] Based on the multimodal analysis results, determining at least one evaluation scoring factor corresponding to the target object;
[0309] According to all the evaluation and scoring factors corresponding to the target object, a weight allocation operation is performed on each evaluation and scoring factor to obtain a weight value corresponding to each evaluation and scoring factor;
[0310] According to the weight value corresponding to each evaluation scoring factor, a set of evaluation factors for the target object is determined from a set of candidate evaluation questions and evaluation tasks;
[0311] And, the determination module 304 is further used to obtain the response processing result information of the target object for each inspection factor for each target object in the input module 307, input all the response processing result information corresponding to the target object into the target deep learning model, and determine the number of training data of all evaluation data pairs included in the evaluation data set before obtaining the response processing result analysis result;
[0312] The judgment module 310 is also used to judge whether the amount of training data is greater than or equal to a preset threshold value; when it is judged that the amount of training data is greater than or equal to the preset threshold value, the triggering module 306 executes the operation of obtaining the response processing result information of the target object for each assessment factor for each target object, and the triggering module 307 executes the operation of inputting all the response processing result information corresponding to the target object into the target deep learning model to obtain the response processing result analysis result.
[0313] It can be seen that the implementation Figure 4 The described device can determine at least one evaluation and scoring factor of the target object based on the multimodal analysis results, and determine the weight value corresponding to each evaluation and scoring factor, and determine the target problem or examination element according to the weight value corresponding to each evaluation and scoring factor, and can only execute the operation of inputting the corresponding response processing result information of the target object into the target deep learning model to obtain the response processing result analysis result when it is judged that the number of training data is greater than or equal to the preset number threshold, and can determine the target problem or examination element according to the evaluation and scoring factors of the target object and the corresponding weight value, which is conducive to improving the accuracy and reliability of determining the target problem or examination element, and can determine the target problem or examination element in combination with the weight value of each evaluation and scoring factor, which is conducive to improving the pertinence and intelligence of determining the target problem or examination element; and can only execute the operation of inputting a large amount of information into the target deep learning model to obtain the response processing result analysis result when it is judged that the number of training data is greater than or equal to the preset number threshold, which can improve the accuracy and reliability of the target deep learning model, and can improve the intelligence of the target deep learning model, thereby helping to improve the accuracy and reliability of the response processing result analysis result obtained based on the target deep learning model.
[0314] In yet another optional embodiment, Figure 4 As shown, the determination module 304 is further used to collect recruitment information of the target position of at least one employer before determining the employment demand information of the target position of at least one employer, and for each employer, determine the recruitment position of the target position of the employer according to the recruitment information of the target position of the employer;
[0315] The acquisition module 306 is also used to obtain the target unit information of the target position of the employer;
[0316] The determination module 304 is further used to determine the job requirement information of the recruitment position according to the target unit information of the target position of the employer; wherein the target unit information includes one or more of the work scene information of the target position of the employer, the job requirement information of the target position of the employer, and the skill requirement information of the target position of the employer;
[0317] The generating module 309 is further used to generate, for each employer, a job analysis result of the employer's target position according to the job requirement information of the employer's target position recruitment position;
[0318] The specific manner in which the determination module 304 determines the employment demand information of the target position of at least one employer includes:
[0319] For each employer, according to the job analysis results of the employer's target job, determine the employment demand information of the employer's target job;
[0320] The employment demand information includes one or more of work task information, job qualification information, salary range information, employment time information, and work location information.
[0321] It can be seen that the implementation Figure 4 The described device can collect recruitment information of at least one employer's target position and determine the recruitment position of the employer's target position, and obtain the target unit information of the employer's target position to determine the job requirement information of the recruitment position, generate the job analysis results of the employer's target position according to the job requirement information of the recruitment position of each employer's target position, and determine the employment requirement information of the employer's target position according to the job analysis results of the employer's target position, and can comprehensively generate the job analysis results of the employer's target position through recruitment information and job requirement information, which is conducive to improving the accuracy and reliability of generating the job analysis results of the employer's target position, and is conducive to improving the intelligence of generating the job analysis results of the employer's target position, so that it can also determine the employment requirement information of the employer's target position according to the job analysis results of the employer's target position, which is conducive to improving the accuracy and reliability of the employment requirement information of the employer's target position, and further conducive to improving the accuracy and reliability of the subsequent generation of the matching degree between each target object and the employer.
[0322] Embodiment 4
[0323] See also Figure 5 , Figure 5 FIG. 1 is a schematic diagram of the structure of another metaverse intelligent evaluation system disclosed in an embodiment of the present invention. Figure 5As shown, the Metaverse Intelligent Assessment System may include:
[0324] A memory 401 storing executable program codes;
[0325] a processor 402 coupled to the memory 401;
[0326] The processor 402 calls the executable program code stored in the memory 401 to execute the steps in the intelligent multimodal evaluation implementation method described in the first embodiment of the present invention or the second embodiment of the present invention.
[0327] Embodiment 5
[0328] An embodiment of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute the steps in the intelligent multimodal evaluation implementation method described in Embodiment 1 or Embodiment 2 of the present invention.
[0329] Embodiment 6
[0330] An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to enable a computer to execute the steps in the intelligent multimodal evaluation implementation method described in Embodiment 1 or Embodiment 2.
[0331] The system embodiment described above is only illustrative, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, that is, they may be located in one place, or they may be distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without paying creative labor.
[0332] Through the specific description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solution can be essentially or partly contributed to the prior art in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.
[0333] Finally, it should be noted that the intelligent multimodal evaluation implementation method and the metaverse intelligent evaluation system disclosed in the embodiments of the present invention are only preferred embodiments of the present invention, which are only used to illustrate the technical scheme of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical schemes described in the aforementioned embodiments can still be modified, or some of the technical features can be replaced by equivalents. These modifications or replacements do not deviate the essence of the corresponding technical schemes from the spirit and scope of the technical schemes of the embodiments of the present invention.
Claims
1. A method for implementing intelligent multimodal evaluation, characterized in that: The application is to a Metaverse intelligent assessment system, wherein the Metaverse intelligent assessment system includes a multimodal intelligent assessment platform with deep learning capabilities, a target device, and a talent selection element set, wherein the target device includes one or more of a VR device, an AR device, and an MR device, and the method is implemented by the multimodal intelligent assessment platform and / or the target device included in the Metaverse intelligent assessment system; the method includes: The Metaverse intelligent assessment system inputs the multimodal data set into a predetermined deep learning model for training based on the multimodal data set collected from several users when using the multimodal intelligent assessment platform for recruitment and assessment, so as to obtain a target deep learning model trained to convergence; wherein, the multimodal assessment data set includes a multimodal assessment data pair corresponding to each of the multimodal assessments in several multimodal assessments acquired in real time from the Metaverse intelligent assessment system, and the assessment data pair includes assessment multimodal data and assessment result data corresponding to the assessment multimodal data, wherein the assessment multimodal data includes real-time multimodal data of assessors and real-time multimodal data of assessment objects, and the assessment result data includes an assessment result output by at least one of the assessors for at least one of the assessment objects, and the Metaverse intelligent assessment system includes the target deep learning model; The Metaverse intelligent evaluation system collects real-time evaluation multimodal data of a number of target objects to be evaluated, analyzes the real-time evaluation multimodal data of each target object, obtains a multimodal analysis result of the target object, determines at least one assessment factor corresponding to the target object based on the multimodal analysis result of the target object, and outputs all the assessment factors determined for the target object to the multimodal evaluation terminal corresponding to the target object, so that the multimodal evaluation terminal outputs all the assessment factors for the target object to solve; For each of the target objects, the Metaverse intelligent assessment system obtains the response processing result information of the target object for each of the assessment factors, inputs all the response processing result information corresponding to the target object into the target deep learning model, and obtains the response processing result analysis result; wherein, the response processing result information includes image information obtained by image acquisition of the target object, voice information obtained by voice acquisition of the target object, and a variety of information contained in the data stream formed by the image information and voice information of the target object within a specific unit time; For each of the target objects, the Metaverse Intelligent Assessment System determines an assessment result of the target object based on an analysis result of the response processing result of the target object; The Metaverse intelligent assessment system performs an iterative training operation on the target deep learning model in the Metaverse intelligent assessment system based on the collected assessment result data corresponding to multiple assessed objects that have undergone multimodal assessment within a certain time period and the real-time assessment multimodal data corresponding to each of the assessed objects, so as to update the target deep learning model, wherein all the assessed objects include at least one assessed object that has undergone an autonomous multimodal assessment through the Metaverse intelligent assessment system within the certain time period and / or at least one assessed object that has undergone a non-autonomous multimodal assessment through the Metaverse intelligent assessment system within the certain time period; the autonomous multimodal assessment includes a multimodal assessment without the participation of an assessor, and the non-autonomous multimodal assessment includes a multimodal assessment with the participation of corresponding assessors; For each of the target objects, after the Metaverse Intelligent Assessment System determines the assessment result of the target object based on the analysis result of the response processing result of the target object, the method further includes: For each of the target objects, the Metaverse intelligent assessment system obtains the recruitment demand information of the target position of the target employer corresponding to the target object, and inputs the recruitment demand information of the target position of the target employer corresponding to the target object into a predetermined demand analysis model to obtain the demand analysis result of the target position of the target employer corresponding to the target object; For each of the target objects, the Metaverse intelligent assessment system calculates the matching degree between the assessment result of the target object and the demand analysis result of the target position of the target employer corresponding to the target object through a predetermined matching algorithm, obtains the person-position matching degree between the target object and the target employer corresponding to the target object, and generates a position matching result and report between the target object and the target employer corresponding to the target object based on the person-position matching degree; And, the method further comprises: For each of the target objects, the Metaverse intelligent assessment system determines whether the job matching degree between the target object and the target employer corresponding to the target object meets the preset person-job matching condition based on the job matching degree result between the target object and the target employer corresponding to the target object; For each of the target objects, when the Metaverse intelligent assessment system determines that the job matching degree between the target object and the target employer corresponding to the target object meets the preset person-job matching condition, the target object is determined as a candidate for recruitment of the target job of the target employer corresponding to the target object; For each of the target employers, the Metaverse intelligent assessment system generates assessment recommendation information based on the assessment results of each candidate recruitment object corresponding to the target employer, and pushes the assessment recommendation information to the terminal corresponding to each candidate recruitment object and the terminal corresponding to the target employer.
2. The intelligent multimodal evaluation implementation method according to claim 1, characterized in that: After determining, for each target object, the target object as a candidate for recruitment of a target position of a target employer corresponding to the target object, the method further includes: For each associated object that matches the target object, the Metaverse Intelligent Evaluation Platform determines the value parameters of the evaluation information corresponding to the associated object based on the selection factor reply information of the associated object, and determines the remuneration information of the intelligent selection optimization contribution that matches the associated object based on the information value parameters and a predetermined incentive fee accounting algorithm, and pays the associated object the remuneration fee that matches the associated object based on the remuneration information.
3. The intelligent multimodal evaluation implementation method according to claim 1, characterized in that: After determining, for each target object, the target object as a candidate for recruitment of a target position of a target employer corresponding to the target object, the method further includes: For each candidate for recruitment of a target position of the target employer, the Metaverse intelligent assessment system performs a remote Metaverse immersive simulated work test on the candidate based on the assessment result of the candidate, obtains the simulated work performance information of the candidate, and performs quantitative scoring on the simulated work performance information of the candidate according to a preset work performance quantification algorithm, obtains the Metaverse immersive remote simulated work performance score of the candidate, and determines whether the remote simulated work performance score of the candidate is greater than or equal to a preset work performance threshold; For each candidate recruitment object for the target position of the target employer, when the Metaverse intelligent assessment system determines that the remote simulation work performance score of the candidate recruitment object is greater than or equal to the preset work performance threshold, the Metaverse intelligent assessment system generates a work simulation test report and recommendation information based on the work simulation test results of each candidate recruitment object corresponding to the target employer, and pushes the work simulation test report and recommendation information to the terminal corresponding to each candidate recruitment object and the terminal corresponding to the target employer.
4. The intelligent multimodal evaluation implementation method according to any one of claims 1 to 3, characterized in that: For each of the target objects, the Metaverse intelligent assessment system obtains the response processing result information of the target object for each of the assessment elements, inputs all the response processing result information corresponding to the target object into the target deep learning model, and before obtaining the response processing result analysis result, the method further includes: For each of the target objects, the Metaverse intelligent assessment system determines the facial image information of the target object based on the image information corresponding to the target object, and determines the target voice information of the target object based on the voice information corresponding to the target object; Wherein, the inputting of all the corresponding processing result information corresponding to the target object into the target deep learning model to obtain the corresponding processing result analysis result includes: The facial image information of the target object and the voice information corresponding to the target object are input into the target deep learning model to obtain the corresponding processing result analysis result.
5. The intelligent multimodal evaluation implementation method according to claim 4, characterized in that: For each of the target objects, the Metaverse intelligent assessment system determines the facial image information of the target object according to the image information corresponding to the target object, and determines the target voice information of the target object according to the voice information corresponding to the target object, the method further includes: The Metaverse intelligent assessment system collects a micro-expression and micro-movement training data set and a voice intonation and / or tone data training set, and performs a training operation on the micro-expression and micro-movement training data set and the voice intonation and / or tone data training set through the predetermined deep learning algorithm to construct a multimodal personality assessment model; The Metaverse intelligent assessment system calculates the model loss parameter of the multimodal personality assessment model, and determines whether the model loss parameter meets the preset model generation condition; When the Metaverse intelligent assessment system determines that the model loss parameter satisfies the preset model generation condition, for each target object, the facial image information, behavior information and target voice information of the target object are input into the multimodal personality assessment model to obtain a comprehensive assessment result of the target object; For each of the target objects, the Metaverse Intelligent Assessment System determines the assessment result of the target object according to the analysis result of the response processing result of the target object, including: For each of the target objects, the Metaverse intelligent assessment system determines the assessment result of the target object based on the analysis result of the response processing result of the target object and the multimodal personality assessment result of the target object.
6. The intelligent multimodal evaluation implementation method according to any one of claims 1 to 3, characterized in that: The method further comprises: The Metaverse intelligent assessment system determines the employment demand information of the target position of at least one employer, and generates a set of alternative assessment questions and inspection tasks based on the employment demand information of the target position of each employer, wherein the set of alternative assessment questions and inspection tasks includes a plurality of alternative assessment questions and inspection tasks; The generation of the candidate evaluation questions and the inspection task set based on the employment demand information of the target positions of each employer includes: The Metaverse Intelligent Assessment System obtains an assessment question and inspection task training data set, wherein the assessment question and inspection task training data set includes a plurality of assessment training questions and a training answer corresponding to each of the assessment training questions, and includes a plurality of inspection task training data and a training inspection processing result corresponding to each of the inspection task training data; The Metaverse Intelligent Assessment System performs model training on a predetermined candidate question generation model through a predetermined deep learning algorithm and the assessment question and inspection task training data set to obtain a question generation model; The Metaverse Intelligent Assessment System inputs the employment demand information of the target positions of all the employers into the question generation model to obtain the model output results, and based on the model output results, generates a set of alternative assessment questions and inspection tasks, wherein the model output results include at least one alternative assessment question and inspection task; And, the determining, based on the multimodal analysis result of the target object, at least one investigation factor corresponding to the target object includes: The Metaverse Intelligent Assessment System determines the assessment elements of the target object from the set of alternative assessment questions and assessment tasks based on the multimodal analysis results of the target object, wherein the assessment elements include assessment questions and / or assessment tasks.
7. The intelligent multimodal evaluation implementation method according to claim 6, characterized in that: The Metaverse Intelligent Assessment System determines the assessment factors of the target object from the set of candidate assessment questions and assessment tasks based on the multimodal analysis results of the target object, including: The Metaverse intelligent evaluation system determines at least one evaluation scoring factor corresponding to the target object based on the multimodal analysis result; The Metaverse Intelligent Assessment System performs a weight allocation operation on each of the assessment scoring factors according to all the assessment scoring factors corresponding to the target object, and obtains a weight value corresponding to each of the assessment scoring factors; The Metaverse Intelligent Assessment System determines a set of assessment factors for the target object from the set of candidate assessment questions and assessment tasks according to the weight value corresponding to each of the assessment scoring factors; And, for each of the target objects, the Metaverse intelligent assessment system obtains the response processing result information of the target object for each of the assessment elements, inputs all the response processing result information corresponding to the target object into the target deep learning model, and before obtaining the response processing result analysis result, the method further includes: The Metaverse intelligent evaluation system determines the amount of training data of all the evaluation data pairs included in the evaluation data set, and determines whether the amount of training data is greater than or equal to a preset quantity threshold; When the Metaverse Intelligent Assessment System determines that the amount of training data is greater than or equal to the preset quantity threshold, it triggers the execution of the operation in which, for each of the target objects, the Metaverse Intelligent Assessment System obtains the response processing result information of the target object for each of the assessment factors, inputs all the response processing result information corresponding to the target object into the target deep learning model, and obtains the response processing result analysis result.
8. The method for realizing intelligent multimodal data evaluation according to claim 6, characterized in that: Before the Metaverse intelligent assessment system determines the employment demand information of the target position of at least one employer, the method further includes: The Metaverse intelligent assessment system collects recruitment information of target positions of at least one employer, and for each of the employers, determines the recruitment position of the employer's target position based on the recruitment information of the employer's target position, and obtains the target unit information of the employer's target position, and determines the position requirement information of the recruitment position based on the target unit information of the employer's target position; wherein the target unit information includes one or more of the work scenario information of the employer's target position, the position requirement information of the employer's target position, and the skill requirement information of the employer's target position; For each of the employers, the Metaverse intelligent assessment system generates a job analysis result of the employer's target position based on the job requirement information of the employer's target position; The Metaverse Intelligent Assessment System determines the employment demand information of the target position of at least one employer, including: For each of the employers, the Metaverse intelligent assessment system determines the employment demand information of the employer's target position based on the job analysis results of the employer's target position; The employment demand information includes one or more of work task information, job qualification information, salary range information, employment time information, and work location information.
9. A metaverse intelligent evaluation system, characterized in that: The Metaverse intelligent assessment system includes a multimodal intelligent assessment platform with deep learning capabilities, a target device and a talent selection element set, wherein the target device includes one or more of a VR device, an AR device, and an MR device, and the Metaverse intelligent assessment system includes the multimodal intelligent assessment platform and / or the target device, and the Metaverse intelligent assessment system includes: A training module, for inputting the multimodal data set into a predetermined deep learning model for training based on a multimodal data set collected from several users when using the multimodal intelligent evaluation platform for recruitment and evaluation, to obtain a target deep learning model trained to convergence; wherein the multimodal evaluation data set includes a multimodal evaluation data pair corresponding to each of the multimodal evaluations in several multimodal evaluations acquired in real time from the Metaverse Intelligent Evaluation System, the evaluation data pair includes evaluation multimodal data and evaluation result data corresponding to the evaluation multimodal data, wherein the evaluation multimodal data includes real-time multimodal data of an evaluator and real-time multimodal data of an evaluation object, the evaluation result data includes an evaluation result output by at least one of the evaluators for at least one of the evaluation objects, and the Metaverse Intelligent Evaluation System includes the target deep learning model; An acquisition module, used to acquire real-time evaluation multimodal data of a number of target objects to be evaluated; An analysis module, configured to analyze the real-time evaluation multimodal data of each target object to obtain a multimodal analysis result of the target object; A determination module, configured to determine at least one investigation element corresponding to the target object based on the multimodal analysis result of the target object; An output module, used for outputting all the examination factors determined for the target object to a multimodal evaluation terminal corresponding to the target object, so that the multimodal evaluation terminal outputs all the examination factors for the target object to solve; An acquisition module, used for acquiring, for each target object, information on the processing result of the target object for each of the inspection elements; An input module, used to input all the response processing result information corresponding to the target object into the target deep learning model to obtain the response processing result analysis result; wherein the response processing result information includes image information obtained by image acquisition of the target object, voice information obtained by voice acquisition of the target object, and multiple information contained in the data stream formed by the image information and voice information of the target object in a specific unit time; The determination module is further configured to determine, for each target object, an evaluation result of the target object according to an analysis result of a response processing result of the target object; The training module is further used to perform an iterative training operation on the target deep learning model in the Metaverse intelligent evaluation system according to the collected evaluation result data corresponding to multiple evaluated objects that have undergone multimodal evaluation within a certain time period and the real-time evaluation multimodal data corresponding to each of the evaluated objects, so as to update the target deep learning model, wherein all the evaluated objects include at least one evaluated object that has undergone autonomous multimodal evaluation through the Metaverse intelligent evaluation system within the certain time period and / or at least one evaluated object that has undergone non-autonomous multimodal evaluation through the Metaverse intelligent evaluation system within the certain time period; the autonomous multimodal evaluation includes a multimodal evaluation without the participation of an evaluator, and the non-autonomous multimodal evaluation includes a multimodal evaluation with the participation of corresponding evaluators; The acquisition module is further configured to obtain, for each target object, recruitment demand information of a target position of a target employer corresponding to the target object after the Metaverse intelligent evaluation system determines the evaluation result of the target object based on the analysis result of the response processing result of the target object in the determination module; The input module is further used to input the recruitment demand information of the target position of the target employer corresponding to the target object into a predetermined demand analysis model to obtain the demand analysis result of the target position of the target employer corresponding to the target object; The Metaverse Intelligent Assessment System further includes: A calculation module is used to calculate, for each target object, a matching degree between an evaluation result of the target object and a demand analysis result of a target position of a target employer corresponding to the target object by using a predetermined matching algorithm, so as to obtain a person-position matching degree between the target object and the target employer corresponding to the target object; A generation module, used to generate a job matching result and report between the target object and the target employer corresponding to the target object based on the person-job matching; A judgment module is used to judge, for each of the target objects, whether the job matching degree between the target object and the target employer corresponding to the target object meets a preset job matching condition according to the job matching result between the target object and the target employer corresponding to the target object; The determination module is further configured to, for each of the target objects, determine the target object as a candidate for recruitment of the target position of the target employer corresponding to the target object when the determination module determines that the job matching degree between the target object and the target employer corresponding to the target object meets the preset job matching condition; The generating module is further configured to generate evaluation recommendation information for each target employer according to the evaluation result of each candidate recruitment object corresponding to the target employer; The push module is used to push the evaluation recommendation information to the terminal corresponding to each of the candidate recruitment objects and the terminal corresponding to the target employer.
