Method, system and equipment for simulating interview scene based on large model and medium

Abstract

The invention belongs to the technical field of large models and artificial intelligence, and particularly relates to a method, a system and equipment for simulating an interview scene based on a large model and a medium, and the method comprises the steps: obtaining input resume information through a visual interface; preprocessing the input resume information, performing semantic analysis on the input resume by using a large model, and extracting key information of the resume in combination with recruitment information of a user company; generating a corresponding interview problem based on the analysis result and the key information of the resume, and recording generated related parameters in real time; calculating an interview evaluation index based on the reply information of the job seeker and the resume information until the set number of questions and answers are completed; comparing the interview evaluation index with a preset threshold range to obtain an evaluation scheme corresponding to the interview evaluation index; and presenting the generated question and the corresponding job seeker reply information and evaluation scheme to the user in a visual form. For job seekers, the interview success rate can be improved.

Description

Technical Field

[0001] The present invention belongs to the field of big models and artificial intelligence technology, and specifically relates to a method, system, device and medium for simulating interview scenarios based on big models. Background Art

[0002] In the current recruitment process, interviews are an important part of evaluating whether job applicants meet the job requirements. Interviews are not only a test of the interviewer's preparation of answers to test questions, but also a test of the job applicant's reaction speed and interview skills. However, the traditional interview method often relies on interviewers to interview one by one. With the increase in the number of job applicants, for companies, interviewers need to spend a lot of time and energy on interviews. It takes a lot of manpower and material resources to select the best performers and the most suitable candidates for the position from a large number of candidates.

[0003] In order to solve this problem, automated interview systems based on large models have emerged in recent years. These systems usually extract key information from resumes, generate interview questions, and make preliminary evaluations based on the applicants' responses. However, existing automated interview systems still have some shortcomings. For example, they can often only process simple text information and cannot deeply understand the semantic content in resumes; when generating interview questions, they lack close integration with recruitment information, resulting in inaccurate questions; when evaluating applicants' responses, they also lack comprehensiveness and accuracy, and are often not very helpful to applicants. Summary of the invention

[0004] In view of the above-mentioned problems existing in the existing automated interview system, the present invention provides a method, system, device and medium for simulating interview scenarios based on a large model.

[0005] In a first aspect, the technical solution of the present invention provides a method for simulating interview scenarios based on a large model, comprising the following steps: Get the input resume information through the visual interface; Preprocess the input resume information, use the big model to perform semantic analysis on the input resume, and extract key information of the resume in combination with the recruitment information of the user's company; Generate corresponding interview questions based on the analysis results and key information of the resume combined with the recruitment information of the corresponding position, and record the generated relevant parameters in real time; Receive the job seeker's answer information based on the questions, and generate interview questions again based on the answer information and key information of the resume; After the set number of questions and answers are completed, the interview evaluation index is calculated based on the job seeker's answer information and resume information; Compare the interview evaluation index with a preset threshold range to obtain an evaluation plan corresponding to the interview evaluation index; The generated questions, the corresponding job seeker response information and the evaluation scheme are presented to the user in a visual form.

[0006] As a preferred embodiment of the technical solution of the present invention, the method further comprises: Provide a feedback interface to receive users' comments and suggestions on the generated questions and evaluation solutions, and optimize and fine-tune the large model based on users' comments and suggestions.

[0007] As a preferred embodiment of the technical solution of the present invention, the input resume information is preprocessed, a semantic analysis is performed on the input resume using a large model, and the key information of the resume is extracted in combination with the recruitment information of the user company. The steps include: Use text preprocessing tools to remove noise and perform word segmentation on resume information data to obtain preprocessed text data; Perform a preliminary analysis on the preprocessed data using the TF-IDF algorithm to calculate the importance value of each word in the text data; Filter candidate keywords from text data based on calculated importance values: Combined with the recruitment information of the user's company, the candidate keywords are semantically analyzed through the big model to obtain the semantic analysis results; Determine the keywords of the resume based on the results of semantic analysis.

[0008] As a preferred embodiment of the technical solution of the present invention, the preprocessed data is preliminarily analyzed by the TF-IDF algorithm, and in the step of calculating the importance value of each word in the text data, the calculation formula is as follows: Importance value = word frequency × inverse document frequency; Word frequency = the number of times a word appears in an article / the total number of words in the article; Inverse document frequency = log(total number of documents in the corpus / (number of documents containing the term + 1)).

[0009] As a preferred embodiment of the technical solution of the present invention, the step of calculating the interview evaluation index based on the job seeker's reply information and resume information includes: According to the purpose and requirements of the interview, evaluation indicators are extracted from the applicant's responses and resumes, including professional skills, work experience, communication skills, teamwork, and problem-solving skills; Assign a weight to each evaluation indicator; Score each evaluation indicator based on the applicant's responses and resume information; Multiply the score of each indicator by its corresponding weight to get the weighted score; Sum up all weighted scores to get the total evaluation score of the interview; Convert the total evaluation score into an interview evaluation index according to the set index range.

[0010] As a preferred embodiment of the technical solution of the present invention, the method further comprises: Analyze candidate resumes and candidate questions to identify issues in candidate resumes and / or interview responses.

[0011] As a preferred embodiment of the technical solution of the present invention, the steps of analyzing the resume and questions of job seekers and pointing out the problems in the resume and / or interview answers of job seekers include: Collect and organize resume information and interview response data of job applicants; According to the composition of the interview evaluation index, set key analysis indicators and determine the evaluation criteria and thresholds for each analysis indicator; Compare and analyze applicants’ resume information and interview responses against pre-set evaluation criteria and thresholds; Identify analytical indicators where the candidate falls below the assessment criteria; Text mining of applicants’ resumes and interview responses to extract key information and keywords; Interpret the extracted information in combination with the identified analytical indicators and evaluation criteria; Summarize and summarize the problems obtained from the analysis to form a problem report; Provide specific suggestions for improvement to job seekers based on the problem report, and provide feedback on the problem report and improvement suggestions to job seekers.

[0012] When the mock interview is for job seekers, they upload their resumes and intended positions, and the big model can accurately simulate the interview scene based on the position and resume information. Generally, there will be 5-6 rounds of questions. After each question, the job seeker enters his or her answer by voice or text. The big model will make a simple evaluation of the answer and then enter the next round of questions and answers. After the last question, the big model mock interviewer will summarize the interview, including the strengths and weaknesses of the job seeker, and the degree of match between the job seeker's resume and the position, and put forward improvement suggestions to help job seekers identify their own shortcomings and develop targeted improvement plans. Through this multiple mock interview, job seekers have mastered the interviewer's questioning logic and answering skills, and can better deal with various problems encountered in the real interview, thereby improving the success rate of the interview.

[0013] In a second aspect, the technical solution of the present invention also provides a system for simulating interview scenarios based on a large model, including a resume acquisition module, a processing and analysis module, a question generation module, an interview evaluation index calculation module, an evaluation plan acquisition module and an interview result feedback module; Resume acquisition module, used to obtain input resume information through a visual interface; The processing and analysis module is used to pre-process the input resume information, use the big model to perform semantic analysis on the input resume, and extract the key information of the resume in combination with the recruitment information of the user's company; The question generation module is used to generate corresponding interview questions based on the analysis results and the key information of the resume combined with the recruitment information of the corresponding position, and record the generated relevant parameters in real time; receive the job seeker's reply information based on the question, and generate interview questions again based on the reply information and the key information of the resume; An interview evaluation index calculation module is used to calculate the interview evaluation index based on the job seeker's answer information and resume information after completing a set number of questions and answers; An evaluation scheme acquisition module is used to compare the interview evaluation index with a preset threshold range to obtain an evaluation scheme corresponding to the interview evaluation index; The interview result feedback module is used to present the generated questions, the corresponding job seeker's answer information and the evaluation plan to the user in a visual form.

[0014] As a preferred embodiment of the technical solution of the present invention, the system also includes a model fine-tuning module, which provides a feedback interface to receive users' comments and suggestions on the generated questions and evaluation solutions, and optimizes and fine-tunes the large model based on the users' comments and suggestions.

[0015] As a preferred embodiment of the technical solution of the present invention, the processing and analysis module includes a preprocessing unit, an analysis and calculation unit, a candidate word screening unit, a semantic analysis unit and a keyword determination unit; A preprocessing unit, used to perform noise removal and word segmentation preprocessing on the resume information data through a text preprocessing tool to obtain preprocessed text data; An analysis and calculation unit is used to perform preliminary analysis on the preprocessed data using the TF-IDF algorithm and calculate the importance value of each word in the text data; The candidate word screening unit is used to screen candidate keywords from text data based on the calculated importance values: The semantic analysis unit is used to combine the recruitment information of the user company, perform semantic analysis on the candidate keywords through a large model, and obtain the semantic analysis results; The keyword determination unit is used to determine the keywords of the resume according to the semantic analysis result.

[0016] As a preferred embodiment of the technical solution of the present invention, the analysis and calculation unit performs a preliminary analysis on the preprocessed data using the TF-IDF algorithm, and the calculation formula for calculating the importance value of each word in the text data is as follows: Importance value = word frequency × inverse document frequency; Word frequency = the number of times a word appears in an article / the total number of words in the article; Inverse document frequency = log(total number of documents in the corpus / (number of documents containing the term + 1)).

[0017] As a preferred embodiment of the technical solution of the present invention, an interview evaluation index calculation module is specifically used to extract evaluation indicators from the job applicant's replies and resumes according to the purpose and requirements of the interview, and the evaluation indicators include professional skills, work experience, communication skills, teamwork, and problem-solving ability; assign a weight to each evaluation indicator; score each evaluation indicator according to the job applicant's replies and resume information; multiply the score of each indicator by its corresponding weight to obtain a weighted score; sum all weighted scores to obtain a total evaluation score for the interview; and convert the total evaluation score into an interview evaluation index according to a set index range.

[0018] As a preferred embodiment of the technical solution of the present invention, the system further comprises a job seeker question feedback module, which is used to analyze the job seeker resume and job seeker questions and point out the problems in the job seeker resume and / or interview answers.

[0019] As a preferred embodiment of the technical solution of the present invention, a job seeker problem feedback module is specifically used to collect and organize the resume information and interview response data of job seekers; set key analysis indicators according to the composition of the interview evaluation index, and determine the evaluation criteria and thresholds for each analysis indicator; compare and analyze the resume information and interview responses of job seekers with preset evaluation criteria and thresholds; identify the analysis indicators of job seekers that are below the evaluation criteria; perform text mining on the resumes and interview responses of job seekers to extract key information and keywords; interpret the extracted information in combination with the identified analysis indicators and evaluation criteria; summarize and conclude the problems obtained from the analysis to form a problem report; provide specific improvement suggestions to job seekers based on the problem report, and feed back the problem report and improvement suggestions to the job seekers.

[0020] In a third aspect, the technical solution of the present invention provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; the memory stores computer program instructions executable by the at least one processor, and the computer program instructions are executed by the at least one processor so that the at least one processor can execute the method for simulating interview scenarios based on a large model as described in the first aspect.

[0021] In a fourth aspect, the technical solution of the present invention also provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions enable the computer to execute the method for simulating interview scenarios based on a large model as described in the first aspect.

[0022] It can be seen from the above technical solutions that the present invention has the following advantages: by using a large model to perform semantic analysis on the input resume, combined with the recruitment information of the user company, and extracting the key information of the resume, it is possible to have a deeper understanding of the background and experience of the job seeker, and provide strong support for generating accurate interview questions. Based on the analysis results and the key information of the resume, the corresponding interview questions are generated, which can ensure that the questions are highly relevant to the recruitment information and the job seeker's background, thereby improving the pertinence and effectiveness of the interview. After completing the set number of questions and answers, the interview evaluation index is calculated based on the job seeker's reply information and resume information, and compared with the preset threshold range to obtain the corresponding evaluation scheme. This method can comprehensively consider the performance of multiple aspects of the job seeker and improve the accuracy and comprehensiveness of the evaluation. The generated questions and the corresponding job seeker reply information and evaluation scheme are presented to the user in a visual form, which can facilitate the user to intuitively understand the interview process and results and improve the efficiency of the job seeker interview. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0024] Figure 1 is a schematic flow chart of a method according to an embodiment of the present invention.

[0025] Figure 2 is a schematic block diagram of a system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] In the application of simulated interview scenarios, Prompt prompt engineering technology has significant advantages. Traditional simulated interview systems often rely on preset question and answer libraries, lacking flexibility and adaptability. The Prompt prompt engineering technology based on large models can dynamically generate interview questions based on the user's actual resume and job requirements, and judge the applicant's answers. It can even provide intelligent guidance based on the interviewee's answers to achieve a more realistic and natural interview experience. In order to enable personnel in this technical field to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described 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 making creative work should fall within the scope of protection of the present invention.

[0027] like Figure 1As shown, an embodiment of the present invention provides a method for simulating interview scenarios based on a large model, comprising the following steps: Step 1: Obtain the input resume information through the visual interface; Step 2: Preprocess the input resume information, use the big model to perform semantic analysis on the input resume, and extract key information of the resume in combination with the recruitment information of the user's company; Step 3: Generate corresponding interview questions based on the analysis results and key information of the resume combined with the recruitment information of the corresponding position, and record the generated relevant parameters in real time; Step 4: Receive the job seeker's response information based on the question, and generate interview questions again based on the response information and key information of the resume; Step 5: After the set number of questions and answers are completed, the interview evaluation index is calculated based on the job seeker's answer information and resume information; Step 6: Compare the interview evaluation index with a preset threshold range to obtain an evaluation scheme corresponding to the interview evaluation index; Step 7: The generated questions, the corresponding job seeker response information and the evaluation plan are presented to the user in a visual form.

[0028] In some embodiments, the method further comprises: Provide a feedback interface to receive users' comments and suggestions on the generated questions and evaluation solutions, and optimize and fine-tune the large model based on users' comments and suggestions.

[0029] First, design the feedback interface. Specifically, design an intuitive and easy-to-use interface layout to ensure that users can easily find the entrance to submit comments and suggestions. Provide input fields such as text boxes, radio buttons, and multiple-choice buttons to facilitate users to enter different types of feedback. Set a clear submit button to ensure that users can easily submit feedback. Clearly inform users on the interface how their privacy will be protected and obtain necessary privacy authorization.

[0030] Then, collect user comments and suggestions through the feedback interface. Store the collected feedback data in a secure database to ensure data integrity and traceability. Clean the collected data to remove invalid or duplicate feedback.

[0031] The third step is to conduct text analysis on user comments to extract key information and opinions. Use sentiment analysis technology to determine user satisfaction with questions and evaluation plans. Categorize user feedback into different categories, such as inaccurate question generation, unreasonable evaluation plan, etc. Based on the analysis results, locate the problems of the large model in question generation and evaluation plan formulation. Develop specific optimization strategies for the located problems, such as adjusting model parameters, improving algorithms, etc. Prioritize the optimization strategies based on the severity and urgency of the problem.

[0032] The fourth step is to train and adjust the large model according to the optimization strategy. After the training is completed, the newly generated questions and evaluation plans are evaluated to ensure that the optimization effect meets the expectations. According to the evaluation results, the optimization strategy is continuously adjusted and iterative improvements are made. The optimized questions and evaluation plans are presented to the user, and the optimization results are informed.

[0033] In some embodiments, the steps of preprocessing the input resume information, using a large model to perform semantic analysis on the input resume, and combining the recruitment information of the user's company to extract key information of the resume include: Step 21: Use a text preprocessing tool to remove noise and perform word segmentation on the resume information data to obtain preprocessed text data; Step 22: Perform a preliminary analysis on the preprocessed data using the TF-IDF algorithm to calculate the importance value of each word in the text data; Step 23: Filter candidate keywords from the text data based on the calculated importance values: Step 24: Combined with the recruitment information of the user company, semantic analysis is performed on the candidate keywords through the big model to obtain semantic analysis results; Step 25: Determine the keywords of the resume based on the semantic analysis results. The resume includes relevant information about the job you are applying for.

[0034] It should be noted that the TF-IDF algorithm is used to perform a preliminary analysis on the preprocessed data and calculate the importance value of each word in the text data. The calculation formula is as follows: Importance value = word frequency × inverse document frequency; Word frequency = the number of times a word appears in an article / the total number of words in the article; Inverse document frequency = log(total number of documents in the corpus / (number of documents containing the term + 1)).

[0035] In some embodiments, the step of calculating the interview evaluation index based on the job applicant's response information and resume information includes: Step 51: Based on the purpose and requirements of the interview, extract evaluation indicators from the applicant's responses and resume, including professional skills, work experience, communication skills, teamwork, and problem-solving skills; Step 52: assign a weight to each evaluation indicator; Step 53: Score each evaluation indicator based on the job applicant's response and resume information; Step 54: Multiply the score of each indicator by its corresponding weight to obtain a weighted score; Step 55: Sum all weighted scores to obtain the total evaluation score of the interview; Step 56: Convert the total evaluation score into an interview evaluation index according to the set index range.

[0036] It should be noted that the specific process of calculating the interview evaluation index is as follows: First, organize the job seeker's reply information and resume information to ensure that the data format is consistent for subsequent processing. Standardize the key information in the reply and resume, such as converting text replies into numerical scores, or quantifying work experience, educational background, etc. in the resume according to unified standards.

[0037] According to the purpose and requirements of the interview, extract key evaluation indicators from the job applicant's response and resume. These indicators may include professional skills, work experience, communication skills, teamwork, problem-solving skills, etc. Assign a weight to each evaluation indicator, and the size of the weight reflects the importance of the indicator in the overall evaluation. The allocation of weights can be determined based on factors such as company needs, job requirements, and industry standards. Score each evaluation indicator based on the job applicant's response and resume information. The score can be a five-level scoring system (such as 1-5 points), a percentage system, or other appropriate scoring methods. Multiply the score of each indicator by its corresponding weight to obtain a weighted score. Sum all weighted scores to obtain the total evaluation score of the interview.

[0038] Convert the total evaluation score to an interview evaluation index. This can be achieved by mapping the total score to a specific index range, such as an index range of 0-100. The interview evaluation index reflects the overall performance level of the job seeker in the interview. Compare the interview evaluation index with the preset threshold range, and generate a corresponding evaluation plan based on the comparison result. The evaluation plan may include recommendations such as recommended for employment, pending inspection, and not recommended for employment.

[0039] The generated questions, the applicant's response information, the interview evaluation index and the corresponding evaluation plan are presented to the user in a visual form to facilitate the user's decision-making.

[0040] In some embodiments, the method further comprises: Analyze applicant resumes and applicant questions to identify issues in applicant resumes and / or interview responses. Specifically include: Step 81: Collect and organize the resume information and interview response data of job seekers; Step 82: According to the composition of the interview evaluation index, set key analysis indicators and determine the evaluation criteria and thresholds for each analysis indicator; Step 83: Compare and analyze the applicant's resume information and interview responses with the preset evaluation criteria and thresholds; Step 84: Identify the analytical indicators for which the applicant falls below the assessment criteria; Step 85: Perform text mining on the applicant's resume and interview responses to extract key information and keywords; Step 86: interpret the extracted information in combination with the identified analysis indicators and evaluation criteria; Step 87: Summarize and summarize the problems obtained through analysis to form a problem report; Step 88: Provide the job applicant with specific improvement suggestions based on the problem report, and provide the job applicant with feedback on the problem report and improvement suggestions.

[0041] Based on the steps of analyzing the resumes and questions of job seekers to point out problems in the resumes and / or interview responses of job seekers, another application of the present application includes: The job seeker enters his / her resume and the position he / she is applying for or the resume includes the position he / she is applying for. Based on the prompt guidance instruction, the big model will generate multiple interview questions that are highly suitable for the job seeker according to the position and resume. After the job seeker enters his / her answer to the question in the conversation, the big model will professionally judge the answer, summarize the advantages and disadvantages of the question, and continue the next round of questions and answers. When all questions are asked, the big model will professionally judge the overall performance of the job seeker, point out the advantages and disadvantages, and give optimization suggestions and subsequent improvement directions. By conducting multiple rounds of dialogue with job seekers, a highly simulated interview rehearsal platform is created for job seekers, which greatly improves the job seeker's interview skills, increases the probability of successful interviews and the job seeker's confidence in interviews, and assists in improving the employment rate on the social side.

[0042] To choose a suitable large model, you need to have the ability to prompt the project, which can be an open source large model or a self-developed large model. Try to choose a larger model for better results. After selecting the base model, you need to build a large model environment, including evaluating the configuration information of the GPU server according to the selected model, installing the GPU server, installing and deploying the base model, and configuring the environment. After building the large model environment, you need to collect and build test data for subsequent algorithm debugging and optimization, mainly including job seekers' resumes and recruitment positions. For resume information, it is necessary to include age, gender, education, major, skill direction, work experience, skill training, job search intentions, etc. Position information needs to include company name, company business scope, job responsibilities, skill requirements, education and professional requirements, etc. The more detailed the content provided, the better the effect of the simulated interview.

[0043] Prompt project writing - judgment position: On the basis of completing the above plan, the prompt project should be written according to actual needs. According to the needs and the format requirements of the prompt project, the instructions for each step to be completed by the large model should be refined to form text content that the large model can accurately understand: 1) Resume judgment prompt: Write a function to determine whether the content entered by the user is a resume in a formal format. If so, it returns true and the large model continues to the next step. If not, it returns false and prompts the user to re-enter the resume. Instruction example: isResume: You are an interview material judgement expert. You need to judge whether the user input is resume information. The principles to be followed are as follows: 1. When the input is information unrelated to the resume, output false; 2. When the input is a resume, output true; 3. You must output one of these two answers, and you are prohibited from outputting other answers. User input: {userInput}.

[0044] 2) Job judgment prompt: Please provide a piece of text, and I will judge whether the text contains relevant content of the job requirements. The judgment criteria are as follows: 1. If the text is irrelevant to the job requirements, output "False"; 2. If the text contains content of the job requirements, even if not all of the text is relevant to the job requirements, it will output "True"; 3. I will only output "True" or "False" and will not give other answers. Text: {userInput}.

[0045] 3) Generate interview question prompt: You are an interview expert. You can ask 6 different interview questions based on the job requirements and user resumes you input. Output format: "Question 1: Based on your resume, I have a question I want to know..."; "Question 2: Next, I want to know..."; "Question 3: Next, I want to know..."; "Question 4: Next, I want to know..."; "Question 5: Next, I want to know..."; "Question 6: The last question...". The principles to be followed are as follows: 1. It is forbidden to give questions beyond the scope of the resume. You can give some detailed questions about the technology involved in the resume; 2. You must strictly output to me in the output format. It is forbidden to output in other formats. Make sure that the output must include questions 1, 2, 3, 4, 5 and 6. Do not lose them; 3. The questions should cover the resume as comprehensively as possible. Job requirements: {demands}. User resume: {resume}.

[0046] 4) Evaluate the answers to each interview question prompt: You need to play the role of an interview coach and provide feedback based on the interview questions and answers provided. For each answer, you need to choose one of the following three options to output: 1. If the answer is obviously off topic, please reply directly: "Please answer the interview question and do not give irrelevant answers." 2. If the respondent requests a change of question or indicates that he cannot answer, reply: "The question has been changed for you." 3. If the answer gives specific content to the question, even if it is not comprehensive, a brief evaluation should be provided. Feedback should be one to two sentences, evaluating the response, using the second person "you" to express, pointing out the advantages of the answer, but avoiding asking for more information or asking new questions. Question: {question}. Answer: {answer}.

[0047] 5) Interview summary after all questions are answered prompt: You are an interview expert. You need to give an overall evaluation of the interview process based on the job requirements, the user's resume, and the questions and answers. The principles to be followed are as follows: 1. Evaluation format requirements: In the interview, your performance was... (advantages), in xx... (advantages), in xx... (advantages), but there is room for improvement in... (deficiencies). I suggest you... (improvement suggestions). ; 2. The evaluation must be conducted from two perspectives: the fit between the user's resume and the job requirements and the user's answers; 3. When "I don't know" appears in the question and answer, it means that the user's knowledge in the question is insufficient, and this part can be introduced in the insufficiency part of the template; 4. Some personalized suggestions for the user's interview answer are made to facilitate the user to make improvements and improve the user's interview level; 5. Words such as the user's name and resume are prohibited, and the second person "you" must be used to output the overall interview evaluation; 6. It must be detailed, and it is prohibited to be all-general. The evaluation must be combined with the user's resume and the user's answer. The overall evaluation must be given according to the user's answer. It is prohibited to give an evaluation based on each answer. It must be an overall evaluation of all answers; 7. When there are 4 or more "I don't know" in the user's answer, the mandatory overall evaluation output is: During today's interview, you seem to have encountered some obstacles in answering the questions and did not give specific answers to the questions asked. This may be due to communication, lack of preparation, or challenges in coping with interview pressure. Nevertheless, I still want to thank you for your participation and efforts today. Due to the lack of specific answers, it is difficult for me to directly evaluate your professional ability and fit with the position. We recommend that you improve your preparation for future interviews and your ability to perform under pressure. Job requirements: {demands}. User resume: {resume}. Questions and answers: {q&a}.

[0048] After completing the above solution, call the prompt related interface provided by the big model, execute the above content in sequence, and output the return content of the big model at each step.

[0049] like Figure 2 As shown, an embodiment of the present invention also provides a system for simulating interview scenarios based on a large model, including a resume acquisition module, a processing and analysis module, a question generation module, an interview evaluation index calculation module, an evaluation plan acquisition module and an interview result feedback module; Resume acquisition module, used to obtain input resume information through a visual interface; The processing and analysis module is used to pre-process the input resume information, use the big model to perform semantic analysis on the input resume, and extract the key information of the resume in combination with the recruitment information of the user's company; The question generation module is used to generate corresponding interview questions based on the analysis results and the key information of the resume combined with the recruitment information of the corresponding position, and record the generated relevant parameters in real time; receive the job seeker's reply information based on the question, and generate interview questions again based on the reply information and the key information of the resume; An interview evaluation index calculation module is used to calculate the interview evaluation index based on the job seeker's answer information and resume information after completing a set number of questions and answers; An evaluation scheme acquisition module is used to compare the interview evaluation index with a preset threshold range to obtain an evaluation scheme corresponding to the interview evaluation index; The interview result feedback module is used to present the generated questions, the corresponding job seeker's answer information and the evaluation plan to the user in a visual form.

[0050] In some embodiments, the system also includes a model fine-tuning module, which provides a feedback interface for receiving user comments and suggestions on the generated questions and evaluation solutions, and optimizes and fine-tunes the large model based on the user comments and suggestions.

[0051] In some embodiments, the processing and analysis module includes a preprocessing unit, an analysis and calculation unit, a candidate word screening unit, a semantic analysis unit, and a keyword determination unit; A preprocessing unit, used to perform noise removal and word segmentation preprocessing on the resume information data through a text preprocessing tool to obtain preprocessed text data; An analysis and calculation unit is used to perform preliminary analysis on the preprocessed data using the TF-IDF algorithm and calculate the importance value of each word in the text data; The candidate word screening unit is used to screen candidate keywords from text data based on the calculated importance values: The semantic analysis unit is used to combine the recruitment information of the user company, perform semantic analysis on the candidate keywords through a large model, and obtain the semantic analysis results; The keyword determination unit is used to determine the keywords of the resume according to the semantic analysis result.

[0052] In some embodiments, the analysis and calculation unit performs a preliminary analysis on the preprocessed data using the TF-IDF algorithm, and the calculation formula for calculating the importance value of each word in the text data is as follows: Importance value = word frequency × inverse document frequency; Word frequency = the number of times a word appears in an article / the total number of words in the article; Inverse document frequency = log(total number of documents in the corpus / (number of documents containing the term + 1)).

[0053] In some embodiments, the interview evaluation index calculation module is specifically used to extract evaluation indicators from the job applicant's replies and resumes according to the purpose and requirements of the interview, the evaluation indicators including professional skills, work experience, communication skills, teamwork, and problem-solving ability; assign a weight to each evaluation indicator; score each evaluation indicator based on the job applicant's replies and resume information; multiply the score of each indicator by its corresponding weight to obtain a weighted score; sum all weighted scores to obtain a total evaluation score for the interview; and convert the total evaluation score into an interview evaluation index according to a set index range.

[0054] In some embodiments, the system further includes a job applicant question feedback module, which is used to analyze the job applicant's resume and job applicant's questions and point out problems in the job applicant's resume and / or interview answers when the interview evaluation index is lower than the re-examination threshold.

[0055] In some embodiments, the job applicant problem feedback module is specifically used to collect and organize the resume information and interview response data of the job applicant; set key analysis indicators according to the composition of the interview evaluation index, and determine the evaluation criteria and thresholds for each analysis indicator; compare and analyze the resume information and interview responses of the job applicant with the preset evaluation criteria and thresholds; identify the analysis indicators of the job applicant that are below the evaluation criteria; perform text mining on the resume and interview responses of the job applicant to extract key information and keywords; interpret the extracted information in combination with the identified analysis indicators and evaluation criteria; summarize and conclude the problems obtained from the analysis to form a problem report; provide specific improvement suggestions to the job applicant based on the problem report, and feedback the problem report and improvement suggestions to the job applicant.

[0056] The embodiment of the present invention also provides an electronic device, the electronic device includes: a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete mutual communication through the communication bus. The communication bus can be used for information transmission between the electronic device and the sensor. The processor can call the logic instructions in the memory to execute the following method: Step 1: obtain the input resume information through the visual interface; Step 2: pre-process the input resume information, use the large model to perform semantic analysis on the input resume, and extract the key information of the resume in combination with the recruitment information of the user company; Step 3: generate corresponding interview questions based on the analysis results and the key information of the resume combined with the recruitment information of the corresponding position, and record the generated related parameters in real time; Step 4: receive the job seeker's reply information based on the question, and generate the interview question again based on the reply information and the key information of the resume; Step 5: until the set number of questions and answers are completed, the interview evaluation index is calculated based on the job seeker's reply information and resume information; Step 6: compare the interview evaluation index with the preset threshold range to obtain the evaluation scheme corresponding to the interview evaluation index; Step 7: present the generated questions and the corresponding job seeker's reply information and evaluation scheme to the user in a visual form.

[0057] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program code.

[0058] An embodiment of the present invention provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions enable a computer to execute the method provided by the above method embodiment, for example, including: Step 1: Obtain input resume information through a visual interface; Step 2: Preprocess the input resume information, use a large model to perform semantic analysis on the input resume, and extract key information of the resume in combination with the recruitment information of the user company; Step 3: Generate corresponding interview questions based on the analysis results and the key information of the resume in combination with the recruitment information of the corresponding position, and record the generated related parameters in real time; Step 4: Receive the job seeker's reply information based on the question, and generate the interview question again based on the reply information and the key information of the resume; Step 5: After completing the set number of questions and answers, calculate the interview evaluation index based on the job seeker's reply information and resume information; Step 6: Compare the interview evaluation index with the preset threshold range to obtain the evaluation plan corresponding to the interview evaluation index; Step 7: Present the generated questions and the corresponding job seeker's reply information and evaluation plan to the user in a visual form.

[0059] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0060] The embodiment of the system for simulating interview scenarios based on big models provided in an embodiment of the present invention belongs to the same inventive concept as the method for simulating interview scenarios based on big models in the above-mentioned embodiments. For details not fully described in the embodiment of the system for simulating interview scenarios based on big models, please refer to the embodiment of the method for simulating interview scenarios based on big models mentioned above.

[0061] Although the present invention has been described in detail with reference to the accompanying drawings and in combination with the preferred embodiments, the present invention is not limited thereto. Without departing from the spirit and essence of the present invention, a person of ordinary skill in the art may make various equivalent modifications or substitutions to the embodiments of the present invention, and these modifications or substitutions shall be within the scope of the present invention. Any person of ordinary skill in the art may easily think of changes or substitutions within the technical scope disclosed by the present invention, and these shall be within the scope of protection of the present invention.

Claims

1. A method for simulating interview scenarios based on a large model, characterized in that: include: Get the input resume information through the visual interface; Preprocess the input resume information, use the big model to perform semantic analysis on the input resume, and extract key information of the resume in combination with the recruitment information of the user's company; Generate corresponding interview questions based on the analysis results and key information of the resume combined with the recruitment information of the corresponding position, and record the generated relevant parameters in real time; Receive the job seeker's answer information based on the questions, and generate interview questions again based on the answer information and key information of the resume; After the set number of questions and answers are completed, the interview evaluation index is calculated based on the job seeker's answer information and resume information; Compare the interview evaluation index with a preset threshold range to obtain an evaluation plan corresponding to the interview evaluation index; The generated questions, the corresponding job seeker response information and the evaluation scheme are presented to the user in a visual form.

2. The method for simulating interview scenarios based on a large model according to claim 1 is characterized in that: The method further includes: Provide a feedback interface to receive users' comments and suggestions on the generated questions and evaluation solutions, and optimize and fine-tune the large model based on users' comments and suggestions.

3. The method for simulating interview scenarios based on a large model according to claim 2 is characterized in that: The steps of preprocessing the input resume information, using the big model to perform semantic analysis on the input resume, and combining it with the recruitment information of the user's company to extract the key information of the resume include: Use text preprocessing tools to remove noise and perform word segmentation on resume information data to obtain preprocessed text data; Perform a preliminary analysis on the preprocessed data using the TF-IDF algorithm to calculate the importance value of each word in the text data; Filter candidate keywords from text data based on calculated importance values: Combined with the recruitment information of the user's company, the candidate keywords are semantically analyzed through the big model to obtain the semantic analysis results; Determine the keywords of the resume based on the results of semantic analysis.

4. The method for simulating interview scenarios based on a large model according to claim 3 is characterized in that: The TF-IDF algorithm is used to perform a preliminary analysis on the preprocessed data and calculate the importance of each word in the text data. The calculation formula is as follows: Importance value = word frequency × inverse document frequency; Word frequency = the number of times a word appears in an article / the total number of words in the article; Inverse document frequency = log(total number of documents in the corpus / (number of documents containing the term + 1)).

5. The method for simulating interview scenarios based on a large model according to claim 4 is characterized in that: The steps for calculating the interview evaluation index based on the applicant's response information and resume information include: According to the purpose and requirements of the interview, evaluation indicators are extracted from the applicant's responses and resumes, including professional skills, work experience, communication skills, teamwork, and problem-solving skills; Assign a weight to each evaluation indicator; Score each evaluation indicator based on the applicant's responses and resume information; Multiply the score of each indicator by its corresponding weight to get the weighted score; Sum up all weighted scores to get the total evaluation score of the interview; Convert the total evaluation score into an interview evaluation index according to the set index range.

6. The method for simulating interview scenarios based on a large model according to claim 5 is characterized in that: The method further includes: Analyze candidate resumes and candidate questions to identify issues in candidate resumes and / or interview responses.

7. The method for simulating interview scenarios based on a large model according to claim 6 is characterized in that: Analyze Candidate Resumes and Candidate Questions Steps to identify issues in a candidate's resume and / or interview responses include: Collect and organize resume information and interview response data of job applicants; According to the composition of the interview evaluation index, set key analysis indicators and determine the evaluation criteria and thresholds for each analysis indicator; Compare and analyze applicants’ resume information and interview responses against pre-set evaluation criteria and thresholds; Identify analytical indicators where the candidate falls below the assessment criteria; Text mining of applicants’ resumes and interview responses to extract key information and keywords; Interpret the extracted information in combination with the identified analytical indicators and evaluation criteria; Summarize and summarize the problems obtained from the analysis to form a problem report; Provide specific suggestions for improvement to job seekers based on the problem report, and provide feedback on the problem report and improvement suggestions to job seekers.

8. A system for simulating interview scenarios based on a large model, characterized in that: It includes resume acquisition module, processing and analysis module, question generation module, interview evaluation index calculation module, evaluation plan acquisition module and interview result feedback module; Resume acquisition module, used to obtain input resume information through a visual interface; The processing and analysis module is used to pre-process the input resume information, use the big model to perform semantic analysis on the input resume, and extract the key information of the resume in combination with the recruitment information of the user's company; The question generation module is used to generate corresponding interview questions based on the analysis results and key information of the resume combined with the recruitment information of the corresponding position, and record the generated relevant parameters in real time; Receive the job seeker's answer information based on the questions, and generate interview questions again based on the answer information and key information of the resume; An interview evaluation index calculation module is used to calculate the interview evaluation index based on the job seeker's answer information and resume information after completing a set number of questions and answers; An evaluation scheme acquisition module is used to compare the interview evaluation index with a preset threshold range to obtain an evaluation scheme corresponding to the interview evaluation index; The interview result feedback module is used to present the generated questions, the corresponding job seeker's answer information and the evaluation plan to the user in a visual form.

9. An electronic device, characterized in that: The electronic device includes: at least one processor; and a memory communicatively connected to the at least one processor; the memory stores computer program instructions executable by the at least one processor, and the computer program instructions are executed by the at least one processor so that the at least one processor can execute the method based on large model simulation interview scenario as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores computer instructions, and the computer instructions enable the computer to execute the method for simulating interview scenarios based on a large model as described in any one of claims 1 to 7.

Images