Resume evaluation method and device based on artificial intelligence analysis, computer equipment and readable storage medium

Abstract

The invention discloses a resume evaluation method and device based on artificial intelligence analysis, computer equipment and a readable storage medium, and the method comprises the steps: firstly obtaining a to-be-processed resume, carrying out the structural processing of the to-be-processed resume into resume data, inputting the resume data into a pre-trained resume scoring model, and obtaining a multi-dimensional scoring result containing a plurality of key evaluation factors, and then obtaining weight distribution information of the target post for each key evaluation factor, and finally calculating an adaptation degree evaluation result of the resume to be processed for the target post according to the weight distribution information, thereby realizing accurate, efficient and objective resume evaluation.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a resume evaluation method, device, computer equipment and readable storage medium based on artificial intelligence analysis. Background Art

[0002] In today's recruitment field, processing a large number of resumes to screen out suitable candidates is a daunting task. The traditional manual resume evaluation method is inefficient and susceptible to subjective factors, and accuracy and objectivity are difficult to guarantee. Summary of the invention

[0003] The object of the present invention is to provide a resume evaluation method, device, computer equipment and readable storage medium based on artificial intelligence analysis.

[0004] In a first aspect, an embodiment of the present invention provides a resume evaluation method based on artificial intelligence analysis, comprising:

[0005] Obtaining a resume to be processed, and performing structured processing on the resume to be processed to obtain resume data;

[0006] Inputting the resume data into a pre-trained resume scoring model to obtain a resume multi-dimensional scoring result including multiple key evaluation factors;

[0007] Obtaining weight distribution information of the target position for each of the key evaluation factors;

[0008] Based on the weight distribution information and the resume multi-dimensional scoring result, the suitability evaluation result of the resume to be processed for the target position is calculated.

[0009] In a second aspect, an embodiment of the present invention provides a resume evaluation device based on artificial intelligence analysis, comprising:

[0010] The acquisition module is used to acquire the resume to be processed, and perform structured processing on the resume to be processed to obtain resume data; input the resume data into a pre-trained resume scoring model to obtain a resume multi-dimensional scoring result including multiple key evaluation factors; and obtain the weight distribution information of the target position for each of the key evaluation factors;

[0011] An evaluation module is used to calculate the suitability evaluation result of the resume to be processed for the target position based on the weight distribution information and the multi-dimensional scoring result of the resume.

[0012] In a third aspect, an embodiment of the present invention provides a computer device, comprising a processor and a non-volatile memory storing computer instructions, wherein when the computer instructions are executed by the processor, the computer device executes the method described in at least one possible implementation of the first aspect.

[0013] In a fourth aspect, an embodiment of the present invention provides a readable storage medium, wherein the readable storage medium includes a computer program, and when the computer program is run, the computer device where the readable storage medium is located is controlled to execute the method described in at least one possible implementation manner of the first aspect.

[0014] Compared with the prior art, the beneficial effects provided by the present invention include: adopting a resume evaluation method, device, computer equipment and readable storage medium based on artificial intelligence analysis disclosed in the present invention, by obtaining the resume to be processed and structuring it into resume data, inputting it into a pre-trained resume scoring model to obtain a multidimensional scoring result containing multiple key evaluation factors, and then obtaining the weight distribution information of the target position for each key evaluation factor, and finally calculating the suitability evaluation result of the resume to be processed for the target position based on this, thereby achieving accurate, efficient and objective resume evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only illustrate certain embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can also be obtained based on these drawings without creative work.

[0016] Figure 1 A schematic diagram of the steps of a resume evaluation method based on artificial intelligence analysis provided by an embodiment of the present invention;

[0017] Figure 2 A schematic block diagram of the structure of a resume evaluation device based on artificial intelligence analysis provided by an embodiment of the present invention;

[0018] Figure 3 A schematic block diagram of the structure of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions 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 part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0020] The specific implementation modes of the present invention are described in detail below in conjunction with the accompanying drawings.

[0021] In order to solve the technical problems in the aforementioned background technology, Figure 1 The present invention provides a flow chart of a resume evaluation method based on artificial intelligence analysis according to an embodiment of the present invention. The resume evaluation method based on artificial intelligence analysis is introduced in detail below.

[0022] Step S201, obtaining a resume to be processed, and performing structured processing on the resume to be processed to obtain resume data;

[0023] Step S202, inputting the resume data into a pre-trained resume scoring model to obtain a resume multi-dimensional scoring result including multiple key evaluation factors;

[0024] Step S203, obtaining the weight distribution information of the target position for each of the key evaluation factors;

[0025] Step S204, calculating the suitability evaluation result of the resume to be processed for the target position based on the weight distribution information and the resume multi-dimensional scoring result.

[0026] In the embodiment of the present invention, for example, suppose a large manufacturing company is recruiting for a production supervisor position, and the server receives resumes of many applicants. The content of one of the resumes is as follows:

[0027] "Name: Chen XX

[0028] Contact: 139xxxx5678

[0029] Email: chenhua@xxx.com

[0030] Personal profile: Many years of experience in production management, familiar with production process optimization and team management.

[0031] Work Experience:

[0032] From 2015 to 2018, he served as a production team leader in Company A, responsible for the daily management of the production line and personnel deployment, during which time he successfully improved the production line efficiency by 10%.

[0033] From 2018 to 2022, he worked as an assistant to the production supervisor in Company B, participated in production planning and quality control, and assisted the supervisor in solving many production problems.

[0034] Educational Background:

[0035] 2011-2015, XX University of Technology, major in mechanical manufacturing and automation, bachelor's degree

[0036] Project experience:

[0037] Led the production line automation transformation project, shortened the production cycle and reduced production costs.

[0038] Skills:

[0039] Proficient in production management software, such as ERP and MES systems.

[0040] Familiar with Six Sigma quality management methodology.

[0041] Professional Certificate:

[0042] Intermediate Certificate in Production Management

[0043] Awards:

[0044] Won the title of Company A's Outstanding Employee of the Year. "

[0045] The server structures the resume. First, identify the key parts of the resume, such as the name, contact information, and email address in the personal information section. For work experience, extract information such as the time period of work, the company you work for, the position you hold, the specific job responsibilities, and the results you have achieved. Sort out the enrollment time, graduation time, school name, major, and academic qualifications in the educational background separately. For project experience, clarify the name of the project, the specific work you are responsible for, and the results achieved. In terms of skill mastery, list the software and methods you have mastered in detail. The professional certificate section records the specific certificate names. Awards are also sorted out separately. After processing, the structured resume data obtained is as follows:

[0046] personal information:

[0047] -Name: Chen XX

[0048] -Contact information: 139xxxx5678

[0049] -Email: chenhua@xxx.com

[0050] Work Experience:

[0051] -2015-2018, Company A, Production Team Leader, responsible for daily management of production lines and personnel deployment, improving production line efficiency by 10%.

[0052] -2018-2022, Company B, Assistant to the Production Supervisor, participated in production planning and quality control, and assisted in solving production problems.

[0053] Educational Background:

[0054] -2011-2015, XX University of Technology, major in mechanical manufacturing and automation, bachelor's degree

[0055] Project experience:

[0056] -Lead the production line automation transformation project to shorten the production cycle and reduce production costs.

[0057] Skills:

[0058] - Proficient in production management software, such as ERP and MES systems.

[0059] -Familiar with Six Sigma quality management method.

[0060] Professional Certificate:

[0061] -Intermediate Certificate in Production Management

[0062] Awards:

[0063] -Won the title of Company A's Outstanding Employee of the Year.

[0064] The pre-trained resume scoring model in the server began to evaluate Chen XX's resume data. The model first analyzed the work experience. Chen XX's experience as a production team leader in Company A and as an assistant to the production supervisor in Company B was scored 85 points (out of 100) for work experience, considering the years of work, the complexity of job responsibilities and the results achieved.

[0065] Next, the model evaluates educational background. Chen XX graduated from XX University of Technology with a bachelor’s degree in mechanical manufacturing and automation, which is highly relevant to the position of production supervisor, and gave a score of 75 for educational background.

[0066] In terms of skill mastery, Chen XX is proficient in production management software and familiar with quality management methods. Based on the market demand and importance of these skills, the model gives a skill mastery score of 80 points.

[0067] Regarding project experience, the production line automation transformation project led by Chen XX achieved good results, and the model evaluation project experience score was 90 points.

[0068] Finally, since Chen XX holds an intermediate certificate for production management, the model gives the professional certificate a score of 70 points.

[0069] Based on the scores of the above key evaluation factors, the server obtains the multi-dimensional score results of Chen XX's resume:

[0070] Work experience: 85 points

[0071] Educational background: 75 points

[0072] Skill mastery: 80 points

[0073] Project experience: 90 points

[0074] Vocational certificate: 70 points

[0075] For the production supervisor position this time, the company has determined the following weight distribution based on the specific requirements and strategic priorities of the position:

[0076] Work experience: 0.3

[0077] Educational background: 0.2

[0078] Skill Mastery: 0.25

[0079] Project experience: 0.2

[0080] Professional certificate: 0.05

[0081] The server obtains this weight distribution information and uses it to subsequently calculate the compatibility between the resume and the target position.

[0082] Based on the multi-dimensional scoring results of Chen XX's resume and the weight distribution information of the target position, the server calculates the degree of suitability.

[0083] The score for work experience is 85 points, with a weight of 0.3. The weighted score = 85 × 0.3 = 25.5 points.

[0084] The score for educational background is 75 points, with a weight of 0.2. The weighted score = 75 × 0.2 = 15 points.

[0085] The score for skill mastery is 80 points, the weight is 0.25, and the weighted score = 80 × 0.25 = 20 points.

[0086] The score of project experience is 90 points, the weight is 0.2, and the weighted score = 90×0.2 = 18 points.

[0087] The score of the professional certificate is 70 points, with a weight of 0.05. The weighted score = 70 × 0.05 = 3.5 points.

[0088] Adding up all the weighted scores, the total suitability score of Chen XX's resume is: 25.5+15+20+18+3.5=82 points.

[0089] Assume that the preset suitability score threshold of the company is 75 points. Since Chen XX's suitability score of 82 points is higher than the threshold of 75 points, the suitability evaluation result of Chen XX's resume for the production supervisor position in this recruitment is suitable.

[0090] Suppose the company is recruiting for the position of marketing manager and receives another resume from the applicant Li XX.

[0091] "Name: Li XX

[0092] Contact information: 188xxxx9988

[0093] Email: lili@xxx.com

[0094] Personal profile: Have rich marketing experience, strong innovation ability, good communication and teamwork skills.

[0095] Work Experience:

[0096] From 2016 to 2019, he worked as a marketing specialist in Company C, responsible for market research and competitive product analysis, and successfully planned many small marketing activities.

[0097] From 2019 to 2022, he served as the marketing director of Ding Company, leading several large-scale marketing projects and effectively increasing brand awareness and market share.

[0098] Educational Background:

[0099] 2012-2016, XX University of Finance and Economics, Bachelor’s degree, major in Marketing

[0100] Project experience:

[0101] Led the marketing and promotion project for the launch of new products, and sales far exceeded expectations.

[0102] Planned a brand reshaping marketing campaign, which significantly enhanced the brand image.

[0103] Skills:

[0104] Proficient in market analysis tools, such as SPSS and SAS.

[0105] Familiarity with social media marketing and content marketing strategies.

[0106] Professional Certificate:

[0107] Intermediate Certificate in Marketing

[0108] Awards:

[0109] Won the Ding Company's annual best marketing creativity award. "

[0110] The server structures Li XX's resume and inputs it into the pre-trained resume scoring model to obtain the scores of each key evaluation factor. Assume that the scoring results are as follows:

[0111] Work experience: 80 points

[0112] Educational background: 70 points

[0113] Skill mastery: 85 points

[0114] Project experience: 95 points

[0115] Vocational certificate: 65 points

[0116] For the marketing manager position, the company sets the following weight distribution:

[0117] Work experience: 0.25

[0118] Educational background: 0.15

[0119] Skill Mastery: 0.3

[0120] Project experience: 0.25

[0121] Professional certificate: 0.05

[0122] Calculate the total fitness score of Li XX's resume:

[0123] Work experience weighted score = 80 × 0.25 = 20 points

[0124] Weighted score of educational background = 70 × 0.15 = 10.5 points

[0125] Skill mastery weighted score = 85 × 0.3 = 25.5 points

[0126] Project experience weighted score = 95 × 0.25 = 23.75 points

[0127] Weighted score of professional certificate = 65 × 0.05 = 3.25 points

[0128] The total score is: 20+10.5+25.5+23.75+3.25=83 points

[0129] Since 83 points is higher than the preset suitability score threshold of 75 points, the suitability assessment result of Li XX's resume for the marketing manager position is suitable.

[0130] If a large number of resumes for different positions are received during this recruitment, the server will process and evaluate them according to the above process. Suppose there are 30 people applying for the position of financial manager. After calculating the suitability score of each person, the server will sort them in descending order.

[0131] Assume that the compatibility scores of these 30 people are: 88, 82, 78, 85, 75, 80, 86, 70, 76, 83, 72, 79, 81, 68, 73, 77, 65, 69, 71, 87, 66, 63, 60, 58, 55, 50, 45, 40, 35, and 30 respectively.

[0132] The server sorts these scores and generates a candidate ranking table. At the same time, a detailed resume analysis report is generated for each candidate, including the scores of each key evaluation factor, strengths and weaknesses, etc., and the report is configured in the corresponding position of the ranking table. Finally, the ranking table is output to the preset interactive interface for recruiters to view and filter.

[0133] After the recruitment process is completed, the recruiter compares the actual recruitment results with the previous evaluation results and feeds the comparison data back to the server. For example, some candidates with higher scores did not perform well in the actual work, while some candidates with lower scores performed well.

[0134] After receiving this feedback, the server conducts a detailed analysis of the resume cases with incorrect predictions. For example, if candidate A has a high score but poor actual performance, it may be because the description of certain project experiences in the resume is too exaggerated, but the actual ability is insufficient; if candidate B has a low score but excellent performance, it may be because the resume does not fully reflect his in-depth application ability in certain key skills.

[0135] Based on these analysis results, the server adjusts the model's parameters or introduces new evaluation indicators and feature variables, such as the candidate's ability to withstand stress and adaptability in previous work, and then retrains the model using new data to continuously optimize and improve the model's performance, enabling it to more accurately evaluate the compatibility of resumes with target positions, providing a more reliable reference for corporate recruitment.

[0136] In the embodiment of the present invention, the obtaining of the resume to be processed and the structured processing of the resume to be processed to obtain the resume data may be implemented through the following example.

[0137] Obtaining the resume to be processed, and performing structured processing on the resume to be processed;

[0138] The structured resume to be processed is cleaned, denoised and standardized to obtain the resume data.

[0139] In the embodiment of the present invention, for example, it is assumed that a large financial company is recruiting risk control specialists, and the server receives a large number of application resumes.

[0140] First, the server obtains the resumes to be processed. The content of one of the resumes is as follows:

[0141] "Name: Wang XX

[0142] Contact: 137xxxx8888

[0143] Email: wangqiang@xxx.com

[0144] Personal profile: I have a strong interest in risk control in the financial industry and have strong data analysis and risk judgment capabilities.

[0145] Work Experience:

[0146] 2017.06-2019.08 Worked as a risk control assistant in A Financial Company

[0147] - Assist supervisor in credit risk assessment, collecting and analyzing customer information.

[0148] - Participate in the formulation of some risk control strategies.

[0149] 2019.09-Present Risk Control Specialist at Financial Company B

[0150] - Independently responsible for SME credit risk assessment and writing risk assessment reports.

[0151] -Monitor credit business risk indicators and issue timely warnings of potential risks.

[0152] Educational Background:

[0153] 2013.09-2017.06XX University of Finance and Economics, Bachelor of Finance

[0154] Project experience:

[0155] - Participated in a large-scale corporate credit risk assessment project for Financial Company A, responsible for data collection and preliminary analysis.

[0156] -Led the SME credit risk model optimization project of B Financial Company, which improved the accuracy of risk assessment.

[0157] Skills:

[0158] - Proficient in SPSS data analysis software.

[0159] - Familiar with the laws and regulations related to financial risk management.

[0160] Professional Certificate:

[0161] - Certified Financial Risk Manager (FRM) certificate

[0162] Awards:

[0163] -Won the Outstanding Employee of the Year Award from B Financial Company. "

[0164] The server begins to structure the resume, identifying key parts of the resume, such as personal information, work experience, education background, project experience, skills, professional certificates, and awards, and organizes the information into categories.

[0165] The personal information is organized as follows:

[0166] "Name: Wang XX

[0167] Contact: 137xxxx8888

[0168] Email: wangqiang@xxx.com”

[0169] The work experience is summarized as follows:

[0170] Working time: 2017.06-2019.08

[0171] Workplace: A Financial Company

[0172] Position: Risk Control Assistant

[0173] Job Description: Assist supervisor in credit risk assessment, collect and analyze customer information, and participate in the formulation of some risk control strategies.

[0174] Working time: 2019.09-Present

[0175] Workplace: B Financial Company

[0176] Position: Risk Control Specialist

[0177] Job Description: Independently responsible for the credit risk assessment of small and medium-sized enterprises, writing risk assessment reports, monitoring credit business risk indicators, and timely warning of potential risks.

[0178] The educational background is organized as follows:

[0179] "Enrollment time: 2013.09

[0180] Graduation date: 2017.06

[0181] School: XX University of Finance and Economics

[0182] Major: Finance

[0183] Education: Bachelor's degree

[0184] The project experience is summarized as follows:

[0185] "Project Name: A large enterprise credit risk assessment project of a financial company

[0186] Responsible for content: data collection and preliminary analysis

[0187] Project name: Optimization project of small and medium-sized enterprise credit risk model for B Financial Company

[0188] Responsible for content: Leading, improving the accuracy of risk assessment"

[0189] The skills mastery section is organized as follows:

[0190] Skill Name: SPSS Data Analysis Software

[0191] Level of mastery: Proficient

[0192] Skill name: Familiar with relevant laws and regulations on financial risk management

[0193] Level of mastery: Familiar

[0194] The professional certificates are organized as follows:

[0195] "Certificate Name: Certified Financial Risk Manager (FRM) Certificate"

[0196] The award-winning part is summarized as follows:

[0197] “Award title: Won the Outstanding Employee of the Year Award from Financial Company B”

[0198] After completing the structured processing, the server performs data cleaning, denoising and standardization on the resumes to be processed after the structured processing.

[0199] During the data cleaning process, the server checks and deletes possible duplicate and invalid data. For example, a resume may contain duplicate work experience or vague or unclear descriptions, and the server will identify and remove these contents.

[0200] De-noising is mainly to eliminate some possible interference information. For example, the resume may contain some descriptions of personal interests and hobbies that are irrelevant to the position of risk control specialist, or some self-evaluations that are too subjective and lack substantive content. The server will regard these contents as noise and remove them.

[0201] Standardization is to unify and standardize the information format in the resume. For example, the description of working time is standardized to the format of "year.month-year.month"; the description of the degree of skill mastery is standardized to clear words such as "proficient", "familiar", and "understanding".

[0202] After the above-mentioned data cleaning, denoising and standardization processing, we finally obtain clear, accurate and standardized resume data, which provides high-quality input for subsequent scoring and evaluation.

[0203] Through such a rigorous and meticulous processing flow, the server can generate accurate, clear and standardized resume data for each resume, ensuring that subsequent evaluation and analysis can be based on high-quality information, thereby improving the accuracy and efficiency of recruitment.

[0204] In an embodiment of the present invention, the resume scoring model is obtained in the following manner and can be implemented through the following examples.

[0205] Obtain historical recruitment data and labeled resume data;

[0206] The initial resume scoring model is trained based on the historical recruitment data and the labeled resume data, and when a preset training termination condition is reached, the resume scoring model for scoring multiple key evaluation factors is obtained; the multiple key evaluation factors include work experience, educational background, skill mastery, project experience and professional certificates.

[0207] In the embodiment of the present invention, for example, it is assumed that a well-known technology company has accumulated rich historical recruitment data and labeled resume data over a long period of time.

[0208] The server first obtains these historical recruitment data and labeled resume data. The historical recruitment data may include the recruitment process, recruitment requirements, information of the final hired personnel, and relevant information of the unhired personnel for each position of the company in the past few years.

[0209] For example, historical recruitment data for software engineer positions might include the following:

[0210] Job Description: Responsible for the development and maintenance of the company's core business systems, requiring a solid programming foundation, rich project experience and good teamwork skills.

[0211] Information of recruited person A:

[0212] Work experience: 5 years, participated in the development of multiple large-scale projects in well-known Internet companies.

[0213] Educational background: Master of computer-related major from a 985 university.

[0214] Skills: Proficient in programming languages ​​such as Java and Python, and familiar with database design and optimization.

[0215] Project experience: Led the architecture design and development of e-commerce platforms.

[0216] Professional Certificate: Advanced Certificate in Software Engineer.

[0217] Information of unemployed person B:

[0218] Work experience: 2 years, engaged in simple web development in a small company.

[0219] Educational background: Ordinary undergraduate computer major.

[0220] Skills: Familiar with Python programming, limited knowledge of databases.

[0221] Project experience: Participated in the development of the company's internal systems.

[0222] Professional certificate: None.

[0223] Labeled resume data marks and classifies the key information in these historical resumes. For example, work experience is divided into different years and marked, educational background is marked according to school level, major relevance, etc., skill mastery is marked according to proficiency and important skills, project experience is marked according to project scale, complexity and results, and professional certificates are marked according to the level and relevance of the certificate.

[0224] After obtaining these data, the server uses them to train the initial resume scoring model. The initial model may be a model based on deep learning or machine learning algorithms, such as a neural network model or a decision tree model.

[0225] During the training process, the model learns how to predict the suitability of a resume for a specific position based on key evaluation factors such as work experience, educational background, skills, project experience, and professional certificates in the resume.

[0226] Taking work experience as an example, the server inputs a large amount of resume data with work experience annotations into the model. If the resume is marked as "rich work experience", the model will learn its corresponding working years, job promotion, project complexity and other features; for resumes marked as "insufficient work experience", the model will also learn the corresponding feature patterns.

[0227] In terms of educational background, the model learns to distinguish the impact of factors such as different levels of schools, majors, and the relevance of positions on suitability.

[0228] In the skill mastery part, the model uses a large amount of data to learn the importance of different skills and the degree to which mastering these skills will improve the fitness.

[0229] In project experience, the model understands how the scale, complexity, innovation and results of the project affect the resume score.

[0230] In terms of professional certifications, the model understands the added value of different levels and types of certifications to candidates.

[0231] During the training process, the server continuously adjusts the model's parameters to improve the model's prediction accuracy.

[0232] When the model training reaches the preset training termination conditions, for example, the accuracy of the model on the validation set reaches 90% or the value of the loss function is less than a certain threshold, the server considers that the model training is completed and a resume scoring model capable of scoring multiple key evaluation factors is obtained.

[0233] For example, after training, for a resume for a new software engineer position, the model can comprehensively consider factors such as work experience (such as 3 years of development experience in a well-known company), educational background (undergraduate degree in computer science from a 211 university), skills (proficient in Java and C++, familiar with cloud computing technology), project experience (participated in the development of large-scale enterprise-level applications) and professional certificates (Intermediate Certificate in Software Engineer), and give a relatively accurate and objective score.

[0234] Another example is the training of resume scoring models for data analyst positions.

[0235] In the historical recruitment data, the information about the successfully hired data analyst C may be:

[0236] Work experience: 4 years, extensive practical experience in data analysis, processing large-scale data and providing valuable insights.

[0237] Educational background: Master of Statistics, graduated from a well-known university.

[0238] Skills: Proficient in Excel and SQL data analysis tools, skilled in using Python for data mining, and mastering data visualization technology.

[0239] Project experience: Participated in data analysis projects in multiple industries and can use a variety of analysis methods to solve practical problems.

[0240] Professional certificate: Certified Professional Data Analyst (CPDA).

[0241] The information of the unemployed data analyst D might be:

[0242] Work experience: 1 year, only engaged in simple data sorting work.

[0243] Educational background: Bachelor degree in non-statistics related major.

[0244] Skills: Familiar with basic Excel operations, but less familiar with SQL and Python.

[0245] Project experience: No relevant data analysis project experience.

[0246] Professional certificate: None.

[0247] The server uses this rich historical data and labeled information to train the initial model. By learning from a large number of such examples, the model gradually learns how to accurately assess the richness of work experience, the adaptability of educational background, the level of skill mastery, the value of project experience, and the importance of professional certification.

[0248] After multiple iterations and optimizations, when the model can accurately judge the degree of compatibility between a resume and a data analyst position on an independent test dataset and meets the preset accuracy requirements, a mature and usable resume scoring model for the data analyst position is obtained.

[0249] The resume scoring model obtained in this way can provide efficient and accurate support for the company's recruitment work, helping the company to quickly screen out candidate resumes that are highly suitable for the position.

[0250] In the embodiment of the present invention, the calculation of the suitability evaluation result of the resume to be processed for the target position based on the weight distribution information and the multi-dimensional scoring result of the resume can be implemented through the following example.

[0251] According to the weight distribution information and the resume multi-dimensional scoring result, a weighted calculation is performed to obtain a suitability score of the resume to be processed for the target position;

[0252] On the basis that the fitness score is higher than a preset fitness score threshold, the fitness evaluation result is fitness;

[0253] On the basis that the fitness score is lower than a preset fitness score threshold, the fitness evaluation result is unsuitable.

[0254] In an embodiment of the present invention, for example, assume that an Internet company is recruiting for a product manager position, and the server has obtained a resume to be processed, and through the previous steps, obtained the multi-dimensional scoring results of the resume in five key evaluation factors: work experience, educational background, skill mastery, project experience, and professional certificates. At the same time, the server also obtained the weight distribution information set by the company for these five key evaluation factors for the product manager position.

[0255] The multi-dimensional scoring results of this resume are as follows:

[0256] Work experience: 80 points

[0257] Educational background: 75 points

[0258] Skill mastery: 85 points

[0259] Project experience: 90 points

[0260] Vocational certificate: 70 points

[0261] The company sets the following weight distribution for product manager positions:

[0262] Work experience: 0.2

[0263] Educational background: 0.15

[0264] Skill Mastery: 0.25

[0265] Project experience: 0.3

[0266] Professional certificate: 0.1

[0267] The server performs weighted calculations based on the weight distribution information and the resume multi-dimensional scoring results.

[0268] Weighted score of work experience = 80 × 0.2 = 16 points

[0269] Weighted score of educational background = 75 × 0.15 = 11.25 points

[0270] Weighted score for skill mastery = 85 × 0.25 = 21.25 points

[0271] Weighted score of project experience = 90 × 0.3 = 27 points

[0272] Weighted score of professional certificate = 70 × 0.1 = 7 points

[0273] Adding up all the weighted scores, we get the suitability score of this resume for the product manager position: 16+11.25+21.25+27+7=82.5 points

[0274] The company's preset suitability score threshold is 80 points.

[0275] Since the suitability score of this resume is 82.5 points, which is higher than the preset suitability score threshold of 80 points, the server concludes that the suitability evaluation result is suitable. This means that according to the current evaluation criteria, the candidate represented by this resume is more suitable for the product manager position and has a greater chance of being qualified for the position.

[0276] Through such precise weighted calculation and comparison with the threshold, the server can give a clear suitability assessment result for each resume to be processed, providing a strong basis for the company's recruitment decisions and helping the company to more efficiently and accurately screen out candidates who meet the requirements of the target position.

[0277] In an embodiment of the present invention, there are multiple resumes to be processed, each of which corresponds to a candidate, and the following implementation method is also provided.

[0278] Obtaining a suitability score corresponding to each resume to be processed;

[0279] Sort each candidate according to the suitability score from high to low to obtain a candidate ranking table; the resume analysis report corresponding to each candidate is configured at a corresponding position in the candidate ranking table;

[0280] The candidate ranking list is output to a preset interactive interface.

[0281] In the embodiment of the present invention, for example, it is assumed that a large enterprise is recruiting for a financial manager position, and the server has received 50 resumes to be processed, each of which corresponds to a candidate.

[0282] The server first processes these 50 resumes and calculates a suitability score for each resume through the previous evaluation process. The following are examples of suitability scores for some candidates:

[0283] Candidate A: 88 points

[0284] Candidate B: 75 points

[0285] Candidate C: 92 points

[0286] Candidate D: 80 points

[0287] Candidate E: 70 points ...

[0289] Candidate X: 85 points

[0290] Candidate Y: 78 points

[0291] Candidate Z: 83 points

[0292] Next, the server sorts the 50 candidates in descending order of their suitability scores. The sorted results are as follows:

[0293] 1. Candidate C: 92 points

[0294] 2. Candidate A: 88 points

[0295] 3. Candidate X: 85 points

[0296] 4. Candidate D: 80 points

[0297] 5. Candidate Z: 83 points

[0298] 6. Candidate B: 78 points

[0299] 7. Candidate Y: 75 points

[0300] 8. Candidate E: 70 points ...

[0302] After sorting, the server generates a resume analysis report for each candidate. The resume analysis report for candidate C may describe in detail his work experience, including many years of experience in financial management of large enterprises, successful leadership of several important financial projects, and significant cost savings for the enterprise; in terms of education, he graduated from a well-known financial college with a master's degree; in terms of skills, he is proficient in a variety of financial software and analysis tools, and has advanced financial analysis capabilities; in terms of project experience, he has been responsible for financial optimization projects that have achieved remarkable results; in terms of professional certificates, he has a number of authoritative certifications, including a certified public accountant certificate.

[0303] Similarly, for other candidates, corresponding detailed analysis reports will be generated based on the specific content of their resumes, including their strengths, possible shortcomings, and highlights of matching with the target positions.

[0304] Then, the server configures the resume analysis report corresponding to each candidate in the corresponding position of the candidate ranking table. In this way, when the recruiter views the ranking table, he can directly view the detailed resume analysis report of each candidate by clicking on the name or serial number of the candidate.

[0305] Finally, the server outputs the candidate ranking table, which includes the candidate ranking and the corresponding resume analysis report, to the preset interactive interface. When the recruiter opens this interactive interface, he can clearly see the top-ranked candidates and quickly understand the comprehensive situation of each candidate. For example, the recruiter can see the advantages and outstanding performance of candidate C at a glance, and can also easily check the situation of the candidates ranked lower, so as to make recruitment decisions more comprehensively and efficiently and select the most suitable candidates for the position of financial director.

[0306] In this way, the server can provide systematic, orderly and detailed candidate information for the company's recruitment work, greatly improving recruitment efficiency and accuracy.

[0307] In the embodiments of the present invention, the following implementation modes are also provided.

[0308] Receiving feedback information from the user based on the fitness evaluation result, wherein the feedback information includes comparison data between the actual recruitment result and the predicted result;

[0309] The resume scoring model is iteratively optimized using the feedback information.

[0310] In the embodiment of the present invention, for example, it is assumed that after a large enterprise completes a round of recruitment, the server receives feedback information from the recruiter based on the suitability evaluation results.

[0311] The recruiter provides the server with detailed data comparing the actual recruitment results with the predicted results. For example, for the software development engineer position, the following feedback is available:

[0312] The resume suitability assessment result of Candidate A showed that he was highly suitable, but after he was actually recruited and hired, his work performance failed to meet expectations. The feedback information described in detail that the technical difficulties that Candidate A encountered in actual work could not be effectively solved, and there were also certain problems in teamwork.

[0313] Candidate B's resume fit assessment score was relatively low, but in actual work, he demonstrated outstanding technical and problem-solving skills, exceeding expectations.

[0314] After receiving this feedback, the server begins to analyze it. For Candidate A, the server will review his resume and the evaluation process. It may be found that although his work experience and skills are scored high in the evaluation, the model's evaluation of his ability to solve complex technical problems and soft skills of teamwork is not accurate enough.

[0315] For Candidate B, the server may find that his or her in-depth skills in certain specific areas or innovative capabilities in actual projects were not fully valued or accurately quantified in the initial assessment.

[0316] Based on these analyses, the server uses feedback information to iteratively optimize the resume scoring model.

[0317] First, the server identifies and marks the incorrectly predicted resume cases. The cases of Candidate A and Candidate B are explicitly marked as key cases that need to be improved for model evaluation.

[0318] The server then extracts feature information from these error cases. For candidate A, the server extracts features of specific technical and collaborative abilities that the candidate did not perform well in the actual work; for candidate B, the server extracts features of specific skills and innovative abilities that the candidate performed well in the actual work but were not fully reflected in the initial assessment.

[0319] Next, the server compares and analyzes these feature information with the correctly predicted cases to determine the source of model errors. For example, it was found that for the evaluation of technical problem-solving ability, the model overly relied on some superficial technical keywords and project descriptions, without in-depth analysis of specific technical depth and application scenarios. For the evaluation of soft skills, there is a lack of effective quantitative indicators and reference data.

[0320] Based on the analysis results, the server adjusts the model parameters. It may increase the weight of certain technical ability assessments, or introduce new evaluation indicators to measure soft skills such as teamwork. At the same time, it may also introduce new feature variables, such as the candidate's contribution to open source projects, activeness in the technical community, etc., to more comprehensively evaluate their technical strength and innovation capabilities.

[0321] The server retrained the resume scoring model using new data that included more annotations for soft skills and specific deep technical capabilities, as well as key features extracted from other success and failure cases.

[0322] During the retraining process, the server continuously monitors the model's performance indicators, such as accuracy, recall, etc., to ensure that the model can more accurately predict the candidate's actual work performance and the fit of the resume on new data.

[0323] After multiple iterations and optimizations, the server obtained a resume scoring model with better performance. In subsequent recruitment evaluations, it can more accurately predict the suitability of candidates for positions and provide companies with more reliable recruitment decision support.

[0324] For example, there is a similar feedback and optimization process for recruiting marketing positions.

[0325] Candidate C's resume assessment showed suitability, but the market development results in actual work were not good; Candidate D's resume assessment score was average, but he actually demonstrated strong market insight and customer development capabilities.

[0326] After receiving these feedbacks, the server analyzed and found that in evaluating the actual effect of market expansion, the model did not fully consider factors such as the market size and growth potential of the candidate's past projects; in evaluating market insight, there was a lack of in-depth exploration of the candidate's specific market analysis cases.

[0327] The server then makes corresponding adjustments and optimizations and retrains the model so that it can more accurately assess the suitability of candidates for marketing positions.

[0328] By continuously receiving user feedback and optimizing the model, the server can continuously improve the accuracy and reliability of resume evaluation and better serve the recruitment needs of enterprises.

[0329] In the embodiment of the present invention, the iterative optimization of the resume scoring model using the feedback information can be implemented through the following examples.

[0330] Based on the feedback information, identifying and marking incorrectly predicted resume cases;

[0331] Extract the characteristic information from these error cases, compare and analyze them with the correctly predicted cases, and determine the source of model error;

[0332] According to the analysis results, the model parameters are adjusted or new feature variables are introduced to retrain the resume scoring model until the model performance is improved to meet the preset standards.

[0333] In the embodiment of the present invention, for example, it is assumed that after a large enterprise completes a round of recruitment for human resource manager positions, the server receives detailed feedback information from the recruitment team.

[0334] First, based on this feedback information, the server identifies and marks the cases of incorrectly predicted resumes. For example, Candidate E's resume showed a high degree of fit in the initial assessment, but in actual work, he showed insufficient leadership and unskilled application of human resources expertise; Candidate J's resume assessment score was low, but after joining the company, he performed well in handling complex employee relationships and formulating human resources strategies. The server marked these two cases as incorrectly predicted cases.

[0335] Next, the server extracts feature information from these error cases. For candidate E, the extracted features include the lack of significant scale and results of the projects he led in his past work, and the limited professional human resources tools and methods he mastered. For candidate F, the extracted features include the experience of successfully resolving major employee disputes in his past work and the specific measures he took to participate in the formulation of efficient human resources strategies.

[0336] The server then compares and analyzes the characteristic information of these incorrect cases with the correctly predicted cases. The server found that the resumes of correctly predicted candidates usually have clear and rich descriptions of their successful experience in leading large teams, as well as systematic and in-depth learning and application of human resources expertise. Through comparison, the source of model error is determined. For example, the model may rely too much on job titles in the assessment of leadership and ignore actual leadership results; the assessment of human resources expertise may not fully take into account the application differences in different industries and enterprise sizes.

[0337] Based on the above analysis results, the server began to adjust the model parameters. It increased the weight of actual leadership results and the application of professional knowledge, and reduced the reliance on superficial factors such as job titles. At the same time, the server introduced new feature variables, such as the candidate's activity in well-known human resources forums in the industry, and the honors or awards related to human resources received.

[0338] Next, the server retrains the resume scoring model using the adjusted parameters and the newly introduced feature variables. During the training process, the server continuously monitors the model's performance indicators, such as accuracy, recall, and F1 value.

[0339] If the model performance improvement does not meet the preset standard after the initial retraining, for example, the accuracy is still below 90%, the server will analyze the reasons for the failure again. It may be found that the data quality of the newly introduced feature variables is not high, or that some parameters are adjusted too much, resulting in overfitting.

[0340] The server then further optimizes the data processing method, improves the data quality of the new feature variables, and fine-tunes the parameters. Then, the model is retrained again.

[0341] After multiple cycles of adjustment, training, and evaluation, until the model performance reaches the preset standard, such as 95% accuracy, 90% recall, and 92% F1 value, the server believes that the resume scoring model has reached a relatively ideal state after iterative optimization and can more accurately evaluate the resumes of candidates for the position of human resources manager.

[0342] Let's assume that another recruitment is for a sales manager position. The feedback information received by the server shows that candidate Geng's resume evaluation is good but his actual sales performance is poor, while candidate Xin's resume evaluation is average but his actual sales ability is outstanding.

[0343] After the server identified and marked the two error cases, it extracted feature information. For candidate Geng, it was found that although his resume had a lot of sales project experience, his customer satisfaction and sales channel expansion results were not significant; candidate Xin had outstanding new customer development results and high customer satisfaction records in previous work, but these were not fully valued in the initial evaluation.

[0344] Compared with the cases with correct predictions, the server determined that the source of the model error might be the insufficient assessment of customer satisfaction and the ability to develop new customers.

[0345] The server then adjusts the model parameters, increases the weight of customer satisfaction and new customer development capabilities in the scoring, and introduces new feature variables, such as the awards the candidate has won in sales competitions and the number of high-quality customer resources he or she has.

[0346] Continuously retrain and optimize the model until it can accurately distinguish candidates who are truly suitable for sales manager positions, thereby improving the accuracy and effectiveness of recruitment.

[0347] By continuously using feedback information for iterative optimization, the resume scoring model in the server can continuously adapt to the company's recruitment needs and market changes, providing strong support for the company to select more suitable talents.

[0348] In order to more clearly describe the solution provided by the embodiment of the present invention, a relatively complete implementation method is provided below.

[0349] System architecture: The system architecture of the present invention includes the following modules: resume data acquisition module, data cleaning and preprocessing module, AI analysis model module, multidimensional data analysis module, scoring calculation module, and result output module.

[0350] Resume data acquisition module: obtains the electronic resume documents submitted by candidates and parses them into a structured data format using natural language processing technology.

[0351] Data cleaning and preprocessing module: cleans, denoises and standardizes resume data to ensure data consistency and integrity.

[0352] AI analysis model module: Based on deep learning algorithms, AI models are trained using historical recruitment data and labeled resume data to build a model that can identify and evaluate key factors in resumes.

[0353] Multidimensional data analysis module: Use the trained AI model in combination with multidimensional data analysis methods to comprehensively consider factors such as the candidate's work experience, educational background, skill matching, project experience, etc. to generate various scores.

[0354] Scoring calculation module: Based on the output results of the AI ​​model, a weighted algorithm is used to calculate the comprehensive score of the resume and generate a ranked list of resumes according to job requirements.

[0355] Result output module: outputs the final scoring results to the recruitment system and generates an automated resume analysis report for recruitment personnel to make reference for decision-making.

[0356] System working process:

[0357] Step 1: Resume data acquisition: Candidates submit their resumes through the recruitment platform, and the system automatically acquires and stores the resume data.

[0358] Step 2: Data cleaning and preprocessing: Perform text analysis on the obtained resume data, remove irrelevant information, and ensure structuring and standardization.

[0359] Step 3: AI model training and application: Use historical recruitment data to train the AI ​​model so that the model can accurately evaluate key factors such as work experience, education background, and skills.

[0360] Step 4: Multidimensional Data Analysis: Combined with the output of the deep learning model, use multidimensional data analysis technology to conduct a comprehensive assessment of the resume to ensure that factors from different dimensions are taken into account.

[0361] Step 5: Comprehensive score calculation: Based on the evaluation results of the deep learning model, a weighted algorithm is used to calculate the comprehensive score of the candidate to determine the suitability of each candidate for a specific position.

[0362] Step 6: Result output: Output the scoring and sorting results to the recruitment management system to generate an automated resume analysis report to help recruiters quickly find suitable candidates.

[0363] Please refer to Figure 2 , Figure 2 An embodiment of the present invention provides a resume evaluation device 110 based on artificial intelligence analysis, comprising:

[0364] The acquisition module 1101 is used to acquire the resume to be processed, and perform structured processing on the resume to be processed to obtain resume data; input the resume data into a pre-trained resume scoring model to obtain a resume multi-dimensional scoring result including multiple key evaluation factors; and obtain the weight distribution information of the target position for each of the key evaluation factors;

[0365] The evaluation module 1102 is used to calculate the suitability evaluation result of the resume to be processed for the target position according to the weight distribution information and the resume multi-dimensional scoring result.

[0366] It should be noted that the implementation principle of the resume evaluation device 110 based on artificial intelligence analysis can refer to the implementation principle of the resume evaluation method based on artificial intelligence analysis, which will not be repeated here. It should be understood that the division of the various modules of the above device is only a division of logical functions. In actual implementation, they can be fully or partially integrated into a physical entity, or physically separated. And these modules can all be implemented in the form of software called by processing elements; they can also be all implemented in the form of hardware; some modules can also be implemented in the form of software called by processing elements, and some modules can be implemented in the form of hardware. For example, the resume evaluation device 110 based on artificial intelligence analysis can be a separately established processing element, or it can be integrated in a chip of the above device. In addition, it can also be stored in the memory of the above device in the form of program code, and called and executed by a processing element of the above device. The function of the resume evaluation device 110 based on artificial intelligence analysis. The implementation of other modules is similar. In addition, all or part of these modules can be integrated together, or they can be implemented independently. The processing element described here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each module above can be completed by an integrated logic circuit of hardware in a processor element or instructions in the form of software.

[0367] For example, the above modules may be one or more integrated circuits configured to implement the above methods, such as one or more application specific integrated circuits (ASIC), or one or more microprocessors (digital signal processors, DSP), or one or more field programmable gate arrays (FPGA), etc. For another example, when a module above is implemented in the form of a processing element scheduling program code, the processing element may be a general-purpose processor, such as a central processing unit (CPU) or other processor that can call program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0368] The embodiment of the present invention provides a computer device 100, which includes a processor and a non-volatile memory storing computer instructions. When the computer instructions are executed by the processor, the computer device 100 executes the resume evaluation device 110 based on artificial intelligence analysis. Figure 3 As shown, Figure 3 The computer device 100 is a block diagram of a computer device 100 according to an embodiment of the present invention. The computer device 100 includes a resume evaluation device 110 based on artificial intelligence analysis, a memory 111, a processor 112 and a communication unit 113.

[0369] To achieve data transmission or interaction, the memory 111, the processor 112 and the communication unit 113 are electrically connected to each other directly or indirectly. For example, these components can be electrically connected to each other through one or more communication buses or signal lines. The resume evaluation device 110 based on artificial intelligence analysis includes at least one software function module that can be stored in the memory 111 in the form of software or firmware or solidified in the operating system (OS) of the computer device 100. The processor 112 is used to execute the resume evaluation device 110 based on artificial intelligence analysis stored in the memory 111, such as the software function modules and computer programs included in the resume evaluation device 110 based on artificial intelligence analysis.

[0370] An embodiment of the present invention provides a readable storage medium, which includes a computer program. When the computer program is running, it controls the computer device where the readable storage medium is located to execute the aforementioned resume evaluation device 110 based on artificial intelligence analysis.

[0371] For illustrative purposes, the foregoing description is made with reference to specific embodiments. However, the above illustrative discussion is not intended to be exhaustive or to limit the present disclosure to the precise form disclosed. Numerous modifications and variations are possible in accordance with the above teachings. These embodiments are selected and described in order to best illustrate the principles of the present disclosure and its practical application, so that those skilled in the art can best utilize the present disclosure and utilize various embodiments with different modifications to suit the intended specific application.

Claims

1. A resume evaluation method based on artificial intelligence analysis, characterized in that: include: Obtaining a resume to be processed, and performing structured processing on the resume to be processed to obtain resume data; Inputting the resume data into a pre-trained resume scoring model to obtain a resume multi-dimensional scoring result including multiple key evaluation factors; Obtaining weight distribution information of the target position for each of the key evaluation factors; Based on the weight distribution information and the resume multi-dimensional scoring result, the suitability evaluation result of the resume to be processed for the target position is calculated.

2. The method according to claim 1, characterized in that The obtaining of the resume to be processed and structural processing of the resume to be processed to obtain resume data includes: Obtaining the resume to be processed, and performing structured processing on the resume to be processed; The structured resume to be processed is cleaned, denoised and standardized to obtain the resume data.

3. The method according to claim 1, characterized in that The resume scoring model is obtained by: Obtain historical recruitment data and labeled resume data; The initial resume scoring model is trained based on the historical recruitment data and the labeled resume data, and when a preset training termination condition is reached, the resume scoring model for scoring multiple key evaluation factors is obtained; the multiple key evaluation factors include work experience, educational background, skill mastery, project experience and professional certificates.

4. The method according to claim 1, characterized in that: The calculation of the suitability evaluation result of the resume to be processed for the target position based on the weight distribution information and the resume multi-dimensional scoring result includes: According to the weight distribution information and the resume multi-dimensional scoring result, a weighted calculation is performed to obtain a suitability score of the resume to be processed for the target position; On the basis that the fitness score is higher than a preset fitness score threshold, the fitness evaluation result is fitness; On the basis that the fitness score is lower than a preset fitness score threshold, the fitness evaluation result is unsuitable.

5. The method according to claim 4, characterized in that There are multiple resumes to be processed, each of which corresponds to a candidate, and the method further includes: Obtaining a suitability score corresponding to each resume to be processed; Sort each candidate according to the suitability score from high to low to obtain a candidate ranking table; the resume analysis report corresponding to each candidate is configured at a corresponding position in the candidate ranking table; The candidate ranking list is output to a preset interactive interface.

6. The method according to claim 1, characterized in that The method further comprises: Receiving feedback information from the user based on the fitness evaluation result, wherein the feedback information includes comparison data between the actual recruitment result and the predicted result; The resume scoring model is iteratively optimized using the feedback information.

7. The method according to claim 6, characterized in that The iterative optimization of the resume scoring model using the feedback information includes: Based on the feedback information, identifying and marking incorrectly predicted resume cases; Extract the characteristic information from these error cases, compare and analyze them with the correctly predicted cases, and determine the source of model error; According to the analysis results, the model parameters are adjusted or new feature variables are introduced to retrain the resume scoring model until the model performance is improved to meet the preset standards.

8. A resume evaluation device based on artificial intelligence analysis, characterized in that: include: An acquisition module is used to acquire the resume to be processed and perform structured processing on the resume to be processed to obtain resume data; Inputting the resume data into a pre-trained resume scoring model to obtain a resume multi-dimensional scoring result including multiple key evaluation factors; Obtaining weight distribution information of the target position for each of the key evaluation factors; An evaluation module is used to calculate the suitability evaluation result of the resume to be processed for the target position based on the weight distribution information and the multi-dimensional scoring result of the resume.

9. A computer device, characterized in that: The computer device comprises a processor and a non-volatile memory storing computer instructions. When the computer instructions are executed by the processor, the computer device executes the method according to any one of claims 1 to 7.

10. A readable storage medium, characterized in that: The readable storage medium includes a computer program, and when the computer program is executed, the computer device where the readable storage medium is located is controlled to execute the method according to any one of claims 1 to 7.

Images