Artificial intelligence-based recruitment interview evaluation system

Abstract

The invention provides a recruitment interview evaluation system based on artificial intelligence. The recruitment interview evaluation system comprises an information acquisition module which acquires enterprise recruitment information and corresponding applicant resume information and applicant interview information; the skill evaluation module extracts position skill features and applicant skill features, performs skill learning path analysis based on the established industry skill knowledge graph, and further obtains skill potential evaluation results of the applicants; the culture adaptation evaluation module extracts enterprise culture characteristics and applicant culture characteristics, and performs culture adaptation analysis based on the enterprise culture characteristics and the applicant culture characteristics to obtain a culture adaptation evaluation result of an applicant; the interview evaluation module analyzes the interview performance of the applicant according to the interview video record data of the applicant to obtain an interview evaluation result; and the report module is used for obtaining a comprehensive evaluation result of the applicant for the enterprise position. According to the invention, an enterprise can be helped to improve the intelligent level of evaluation of an applicant for a specific position.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence and enterprise management technology, in particular to a recruitment interview evaluation system based on artificial intelligence. Background Art

[0002] As talent resources are increasingly valued, companies are facing more and more challenges when recruiting talent.

[0003] When companies conduct traditional recruitment interviews, most of them are jointly screened and interviewed by the HR manager and the position manager. However, the screening effect depends largely on the vision and judgment of both HR and position management. That is, the HR manager needs to have a certain understanding of the relevant technology of the position so that he can screen out more suitable interview candidates (for example, when screening resumes based solely on the skills required for the position as a hard standard, many very potential candidates will be missed); and the position manager also needs to have awareness of corporate development and management to select employees who are in line with the company's culture and temperament to improve the adaptability of new employees. At the same time, the results of the interview mainly depend on the subjective judgment of the interviewer, and there may be a lack of objectivity in the recruitment results.

[0004] Therefore, based on the above situation, the current recruitment effect for enterprises is usually difficult to guarantee, which leads to a waste of recruitment human resources and an increase in employment costs, and cannot meet the needs of modern enterprises. Summary of the invention

[0005] In view of the above problems, the present invention aims to provide a recruitment interview evaluation system based on artificial intelligence.

[0006] The purpose of the present invention is achieved by the following technical solutions: The present invention proposes a recruitment interview evaluation system based on artificial intelligence, including an information acquisition module, a skill evaluation module, a cultural adaptation evaluation module, an interview evaluation module and a report module; wherein, The information acquisition module is used to obtain the company's recruitment information and the corresponding applicant's resume information and applicant's interview information; The skill assessment module is used to extract job skill characteristics and candidate skill characteristics based on the company recruitment information and candidate resume information respectively; and to perform skill learning path analysis on job skill characteristics and candidate skill characteristics based on the established industry skill knowledge graph, and further obtain the candidate's skill potential assessment results based on the learning path analysis results; The cultural adaptation assessment module is used to extract corporate culture characteristics and candidate culture characteristics based on corporate recruitment information and candidate resume information, and conduct cultural adaptation analysis based on corporate culture characteristics and candidate culture characteristics to obtain the candidate's cultural adaptation assessment results; The interview evaluation module is used to analyze the interview performance of the applicant based on the interview video record data of the applicant to obtain the interview evaluation result; The report module is used to obtain a comprehensive assessment result of the applicant for the company's position based on the applicant's skill potential assessment results, cultural adaptability assessment results and interview assessment results.

[0007] Preferably, the information acquisition module includes an enterprise recruitment information input unit, an applicant resume acquisition unit and an applicant interview acquisition unit; wherein, The enterprise recruitment information input unit is used to obtain enterprise recruitment information by scanning or inputting, wherein the enterprise recruitment information includes recruitment positions, job descriptions corresponding to the recruitment positions, and corporate culture introduction information; wherein the job description includes job responsibilities, qualifications and other requirements; The applicant resume acquisition unit is used to acquire the applicant's resume information, wherein the resume information includes the applicant's basic information, skills introduction, work experience and personal introduction, etc.; The candidate interview acquisition unit is used to record the candidate's interview video record data during the candidate's job interview.

[0008] Preferably, the skill assessment module includes a skill feature extraction unit, a skill path analysis unit and a skill potential assessment unit; wherein, The skill feature extraction unit is used to extract job skill features and candidate skill features according to the enterprise recruitment information and candidate resume information respectively; The skill path analysis unit is used to analyze the skill learning path of the job skill characteristics and the candidate skill characteristics based on the established industry skill knowledge graph, and obtain the best learning path between the candidate skill characteristics and the job skill characteristics; The skill potential estimation unit is used to calculate the skill potential value of the applicant as a skill potential evaluation result based on the obtained optimal learning path.

[0009] Preferably, the skill assessment module further includes a graph building unit; Among them, the graph building unit is used to build a skill knowledge graph based on the job skill database, industry technical manual database, skill training course database, and talent resume database, and to build a learning relationship network between industry skills through the skill knowledge graph; specifically, it includes: 1) Entity extraction is performed based on the job skill database, industry technical manual database, and skill training course database to obtain network node information corresponding to the skills and build a skill knowledge graph network; the extracted entities are industry skills, and each point in the constructed knowledge graph network corresponds to an industry skill; 2) Calculate the weights between network points based on the constructed skill knowledge graph network, where the weight calculation function used is: ; in, ω(i, j) Represents a skill node i and skill nodes j The connection weights between sim(i, j) Indicates skills i and skills j The correlation between them is obtained based on the talent resume database, where , NumJ Indicates the total number of talent resumes. Numj(i, j) Indicates skills i and skills j Total number of items appearing on the same resume; Numz(i, j) In the process of entity extraction, skills i and skills j The number of times the same content appears. NumZ Indicates the total number of contents; DDEC(i, j) Indicates skills i and skills j The time difference factor between them is obtained based on the difference between the time information when the two are first recorded; DHOT(i, j) Indicates skills i and skills j The heat difference factor between them is obtained based on the difference in their current frequencies of occurrence in the content.

[0010] Preferably, the skill path analysis unit includes: 1) Based on the job skills obtained under the same category a and candidate skills b ; 2) Retrieve candidates’ skills based on the established industry skills knowledge graph b To Skills a The learning paths of , respectively, calculate the learning path score of each path, where the learning path score function used is: ; In the formula, Point(i, b, a) Indicates that from the skill node b To skill node a The first i Learning path scores for learning paths, c, d ∈ i Representation variablesc, d Belong to i The adjacent skill nodes in the learning path, where the path direction is from c arrive d ; ω(c, d) Represents the skill node in the industry skill knowledge graph c and skill nodes d The connection weights between sim(c, d) Indicates skills c and skills d The correlation between β(d) Indicates skills d The learning cost factor is β(d) The value is to master the skill d The average time required; 3) Based on the learning path scores of each learning path, select the path with the largest learning path score as the optimal learning path from skill b to skill a.

[0011] Preferably, the skill potential estimation unit comprises: According to the obtained optimal learning path, the potential value of the candidate's skills corresponding to the recruitment position skills is calculated, and the potential value calculation function used is: ; In the formula, Qua(a, b) Indicates the candidate's skills b Skills for the corresponding job positions a Potential value, variable d Representing skills in the skill knowledge graph b To Skills a The best learning path I skill nodes on d ≠ b ; ω(d) Represents a skill node d The node weight in the skill knowledge graph, sim(d) Represents the best learning path I Medium Skill Node d The degree of correlation with the previous skill node; β(d) Indicates skills d The learning cost factor of ; represents the set cost attenuation coefficient, where ; σ 2 Indicates corresponding skills a and skills b The node weight variance of all skill nodes under the same category; According to the potential value of the applicant's skills corresponding to each recruitment position, the recruiter's skill potential assessment result is obtained.

[0012] Preferably, the cultural adaptation assessment model includes: The company culture introduction information and the applicant's personal introduction information are used to construct an input set, and the input set is input into the trained cultural adaptation analysis model to obtain the applicant's cultural adaptation evaluation result output by the cultural adaptation analysis model; The input set is in the form of {corporate culture introduction text, applicant personal introduction text}; Among them, the cultural adaptation analysis model is built based on the BERT model, including an input layer, an encoding layer and an output layer connected in sequence, wherein the input layer performs word segmentation, text completion and character encoding processing on the corporate culture introduction text and the applicant's personal introduction text in the input set, respectively, and obtains the character encoding vector sequence of the two texts respectively; the encoding layer adopts 12 consecutive Transformers, each of which adopts a multi-head attention mechanism for feature extraction, and a fully connected layer is added after the last Transformer to output a 256-dimensional cultural feature vector; the encoding layer extracts cultural feature vectors from the character encoding vector sequences of the two texts respectively, and the output layer compares the similarity of the two cultural feature vectors obtained, wherein the similarity comparison function adopted is the cos function, and the similarity result of the two is obtained as the cultural adaptation evaluation result.

[0013] Preferably, the interview assessment module includes a video extraction unit, an emotion assessment unit, a stress assessment unit and an interview assessment unit; wherein, The video extraction unit is used to obtain the interview video recording data of the applicant, and obtain the corresponding video image information and audio information according to the obtained interview video recording data; wherein the interview video recording data is based on the video recording obtained by the smart camera during the interview of the applicant; The emotion evaluation unit is used to track the facial features of the applicant based on the video image data of the applicant in the first key time period based on the face tracking model, and calculate the expression energy value of the applicant based on the tracking results of the facial feature points, wherein the expression energy value calculation function used is: ; In the formula, EVM The variable represents the energy value of the candidate's expression in response to the interview questions. The higher the energy value, the more unstable the candidate's emotions. t = 1, 2, … T , t Indicates the moment in the critical time period. T Indicates the total duration of the critical time period; ∆P(t) express t The average change of the facial key point coordinates at each moment compared to the previous moment, ∆T t express t The optical flow change of the applicant’s face area at each moment; The stress evaluation unit is used to obtain the stress response value of the applicant based on the audio data of the applicant in the second key time period based on the audio analysis technology, wherein the stress response value calculation function used is: ; In the formula, SRC Indicates the stress response value of the candidate to the interview questions. The higher the stress response value, the greater the stress of the candidate. σ(f t ) and μ(f t ) They represent the standard deviation and mean of the frequency of the audio data of the applicant at each moment in the second key time period, N pauce Indicates the number of abnormal pauses of the applicant in the second key time period. The audio data is detected based on the VAD algorithm. When the audio gap is greater than the preset standard time T pauce When an abnormal pause is recorded, T pauce ∈ [0.5s, 1s] ; T Indicates the total duration of the audio data; α Represents the preset sound fluctuation weight, where α∈ [0.4,0.6] ; β represents the preset pause frequency weight, where β∈[0.3,0.7] ; The interview evaluation unit is used to calculate the interview evaluation value as the interview evaluation result according to the obtained expression energy value and pressure response value, wherein the interview evaluation value calculation function used is: ; in FC Represents the interview evaluation value of the candidate, γ and δ Respectively represent the normalized weight factors set, respectively according to EVM and SRC The historical maximum value is obtained by statistics. EVM Indicates the energy value of the candidate's expression in response to interview questions. SRC Indicates the stress response value of the candidate to the interview questions.

[0014] Preferably, the reporting module includes: Output the candidate's skill potential assessment results, cultural fit assessment results, and interview assessment results separately to generate a comprehensive assessment result of the candidate; or, A comprehensive evaluation is conducted based on the candidate's skill potential assessment results, cultural fit assessment results and interview assessment results to obtain a comprehensive assessment result of the candidate.

[0015] The beneficial effects of the present invention are as follows: the present invention proposes a recruitment interview evaluation system based on artificial intelligence, which can intelligently evaluate the adaptability of applicants to positions from three dimensions of skill potential, cultural adaptability and interview evaluation based on the recruitment information of the enterprise and the resume information of the applicant, and the artificial intelligence technology, wherein the skill characteristics of the applicant are combined with the idea of ​​skill development, thereby improving the adaptability of job skill screening, which is helpful to find more potential personnel; by evaluating the adaptability of the applicant to the corporate culture, the values ​​of the applicant and the degree of suitability of the corporate culture can be quantified, which is helpful to avoid the situation of personnel loss due to insufficient recognition of the corporate culture by new employees due to personality and other reasons. And finally, combined with the interview video recording data of the applicant, objective feedback is given to the on-the-spot interview situation of the applicant, which is helpful to improve the objective level of feedback on the interview status of the applicant. The comprehensive display of the adaptability of the applicant to the enterprise through the multi-dimensional evaluation results helps the enterprise to improve the comprehensiveness, pertinence and objectivity of the evaluation of applicants for specific positions, and also improves the efficiency and intelligence level of enterprise recruitment. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The present invention is further described using the accompanying drawings, but the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative work.

[0017] Figure 1 This is a framework diagram of a recruitment interview evaluation system based on artificial intelligence shown in an embodiment of the present invention; Figure 2 for Figure 1 A schematic diagram of the framework structure of the information acquisition module in the embodiment; Figure 3 for Figure 1 Schematic diagram of the framework structure of the skill assessment module in the embodiment. DETAILED DESCRIPTION

[0018] The present invention is further described in conjunction with the following application scenarios.

[0019] See also Figure 1 The embodiment shows an artificial intelligence-based recruitment interview evaluation system, including an information acquisition module, a skill evaluation module, a cultural adaptation evaluation module, an interview evaluation module and a report module; wherein, The information acquisition module is used to obtain the company's recruitment information and the corresponding applicant's resume information and applicant's interview information; The skill assessment module is used to extract job skill characteristics and candidate skill characteristics based on the company recruitment information and candidate resume information respectively; and to perform skill learning path analysis on job skill characteristics and candidate skill characteristics based on the established industry skill knowledge graph, and further obtain the candidate's skill potential assessment results based on the learning path analysis results; The cultural adaptation assessment module is used to extract corporate culture characteristics and candidate culture characteristics based on corporate recruitment information and candidate resume information, and conduct cultural adaptation analysis based on corporate culture characteristics and candidate culture characteristics to obtain the candidate's cultural adaptation assessment results; The interview evaluation module is used to analyze the interview performance of the applicant based on the interview video record data of the applicant to obtain the interview evaluation result; The report module is used to obtain a comprehensive assessment result of the applicant for the company's position based on the applicant's skill potential assessment results, cultural adaptability assessment results and interview assessment results.

[0020] The above-mentioned embodiment of the present invention proposes a recruitment interview evaluation system based on artificial intelligence, which can be based on the recruitment information of the enterprise and the resume information of the applicant, and based on artificial intelligence technology, intelligently evaluate the adaptability of the applicant to the position from three dimensions of skill potential, cultural adaptability and interview evaluation, wherein the skill characteristics of the applicant are combined with the idea of ​​skill development, thereby improving the adaptability of job skill screening, which is helpful to find more potential personnel; by evaluating the adaptability of the applicant to the corporate culture, the values ​​of the applicant and the degree of suitability of the corporate culture can be quantified, which helps to avoid the situation of personnel loss caused by the lack of recognition of the corporate culture by new employees due to personality and other reasons. And finally, combined with the interview video recording data of the applicant, objective feedback is given to the on-the-spot interview situation of the applicant, which helps to improve the objective level of feedback on the interview status of the applicant. The comprehensive display of the adaptability of the applicant to the enterprise through the multi-dimensional evaluation results helps the enterprise to improve the comprehensiveness, pertinence and objectivity of the evaluation of applicants for specific positions, and also improves the efficiency and intelligence level of enterprise recruitment.

[0021] Among them, the system proposed in the present invention can be built based on a cloud server or a local server built within an enterprise. By establishing a corresponding database as support, and combining the above-mentioned functional modules and on-site equipment to obtain the required data, the system can provide complete recruitment interview evaluation services.

[0022] Preferably, the system also includes a database module; wherein, The database module includes the databases required for building or imaging the system, including job skill database, industry technical manual database, skill training course database, talent resume database, etc. The database can be built based on specific data resources, or by grabbing relevant content from existing open source databases.

[0023] Preferably, see Figure 2 The information acquisition module includes an enterprise recruitment information input unit and an applicant resume acquisition unit; wherein, The enterprise recruitment information input unit is used to obtain enterprise recruitment information by scanning or inputting, wherein the enterprise recruitment information includes recruitment positions, job descriptions corresponding to the recruitment positions, and corporate culture introduction information; wherein the job description includes job responsibilities, qualifications, other requirements, etc.; The applicant resume acquisition unit is used to acquire the applicant's resume information, wherein the resume information includes the applicant's basic information, skills introduction, work experience and personal introduction.

[0024] When a recruitment interview assessment is needed, first enter the relevant information of the company's recruitment and the relevant information of the applicant as a basis, so that the system can evaluate the applicant's skill potential and cultural adaptability based on the above information.

[0025] In one scenario, the personal introduction in the applicant's resume information includes a description of the applicant's personality, interests, hobbies, work attitude, etc.

[0026] Preferably, the information acquisition module also includes an applicant interview acquisition unit; wherein, The candidate interview acquisition unit is used to record the candidate's interview video record data during the candidate's job interview.

[0027] For the interview evaluation of candidates, the video data of the interview process is recorded during the interview of the candidates, which serves as the basis for subsequent interview evaluation.

[0028] Among them, for traditional recruitment screening technologies, such as some artificial intelligence-based screening, it is usually based on set skill conditions, "for example, Java engineers need more than 5 years of Java development experience" to make hard judgments on skill conditions, that is, when screening resumes, the resumes that meet "more than 5 years of Java development experience" are used as screening results, but in the field of programming, if there is another industry veteran who is only proficient in Python, then the veteran will be excluded. But in actual situations, an experienced senior programmer can quickly master the use of Java on the basis of being proficient in Python. Therefore, the screening results will not meet the industry situation, and potential talents will be missed. Therefore, the present invention proposes a method that can analyze the potential value between the skills required for the recruitment position based on the skills that the applicant has mastered, so as to improve the adaptability and effectiveness of talent matching and selection.

[0029] Preferably, see Figure 3 The skill assessment module includes a skill feature extraction unit, a skill path analysis unit and a skill potential assessment unit; wherein, The skill feature extraction unit is used to extract job skill features and candidate skill features according to the enterprise recruitment information and candidate resume information respectively; The skill path analysis unit is used to analyze the skill learning path of the job skill characteristics and the candidate skill characteristics based on the established industry skill knowledge graph, and obtain the best learning path between the candidate skill characteristics and the job skill characteristics; The skill potential estimation unit is used to calculate the skill potential value of the applicant as a skill potential evaluation result based on the obtained optimal learning path.

[0030] The above-mentioned embodiment of the present invention proposes a technical solution for evaluating the skill potential of applicants, wherein skill features are first extracted for the energy-saving requirements of the recruitment position and the skills that the applicant has mastered. Based on the extracted features, the best learning path between the two is obtained through the skill knowledge map. By obtaining the learning path, the fastest path for applicants to learn the skills required for the position with the skills they have currently mastered can be screened, so as to further evaluate the potential value of the applicant for the skills that have not met the requirements based on the best learning path. By quantifying the potential value, it can help to further screen out applicants who have a certain foundation and can quickly master the skills required for the position, thereby improving the adaptability of the skill potential assessment to the position requirements. Compared with the traditional resume screening method based on hard standards, by adding skill potential assessment, the intelligent level of job skill screening can be improved.

[0031] In one scenario, based on the Java skills mastered by the candidate, it is calculated that the best learning path for distributed system construction is from Java to Spring Cloud to distributed system. The calculated skill potential value is 0.92, which is higher than the preset standard. Therefore, it can be considered that the candidate can master "distributed system construction" with less learning cost based on the current technology "Java proficiency" and meet the requirements of the corresponding position.

[0032] Preferably, the skill feature extraction unit includes: Based on the BERT model, the recruitment positions and the corresponding job descriptions are input into the BERT model, the BERT model encodes the job description text, and extracts the corresponding skill keywords as the job skill features by classification; and the applicant's resume information is input into the BERT model, and the skill keywords extracted by the BERT model are obtained as the applicant's skill features.

[0033] Among them, in addition to the BERT model, the skill feature extraction unit can also select other feature word extraction models to extract position skill features and applicant skill features, such as based on the TextRank model, TF-IDF model, SingleRank model, RAKE model, YAKE model and other keyword extraction models that have been disclosed in the prior art to realize the extraction of skill feature words, and the present invention does not make specific limitations here.

[0034] Preferably, the skill assessment module further includes a graph building unit; Among them, the graph building unit is used to build a skill knowledge graph based on the job skill database, industry technical manual database, skill training course database, and talent resume database, and to build a learning relationship network between industry skills through the skill knowledge graph; specifically, it includes: 1) Entity extraction is performed based on the job skill database, industry technical manual database, and skill training course database to obtain network node information corresponding to the skills and build a skill knowledge graph network; the extracted entities are industry skills, and each point in the constructed knowledge graph network corresponds to an industry skill; 2) Calculate the weights between network points based on the constructed skill knowledge graph network, where the weight calculation function used is: ; in, ω(i, j) Represents a skill node i and skill nodes j The connection weights between sim(i, j) Indicates skills i and skills j The correlation between them is obtained based on the talent resume database, where ,NumJ Indicates the total number of talent resumes. Numj(i, j) Indicates skills i and skills j Total number of items appearing on the same resume; Numz(i, j) In the process of entity extraction, skills i and skills j The number of times the same content appears. NumZ Indicates the total number of contents; DDEC(i, j) Indicates skills i and skills j The time difference factor between them is obtained based on the difference between the time information when the two are first recorded; DHOT(i, j) Indicates skills i and skills j The popularity difference factor between them is obtained based on the difference in the current (for example, this year) appearance frequency of the two in the content.

[0035] In one scenario, entity extraction can be implemented using an existing data extraction model, such as a data extraction model based on HMM, HERT, etc., which is not specifically limited in this application.

[0036] Among them, the above-mentioned embodiment of the present invention also proposes a technical solution for building a skill knowledge graph. First, the node information of the knowledge graph network is determined based on data extraction technology, and then the connection weights between the skill nodes are determined through the proposed weight calculation function to reflect the correlation and heat relationship between the skills, thereby completing the construction of the skill knowledge graph and laying the foundation for the subsequent skill potential value evaluation.

[0037] Preferably, the skill path analysis unit includes: 1) Based on the job skills obtained under the same category a and candidate skills b ; 2) Retrieve candidates’ skills based on the established industry skills knowledge graph b To Skills a The learning paths of , respectively, calculate the learning path score of each path, where the learning path score function used is: ; In the formula, Point(i, b, a) Indicates that from the skill node b To skill node a The first i Learning path scores for learning paths, c, d ∈ i Representation variables c, d Belong to i The adjacent skill nodes in the learning path, where the path direction is from c arrive d ;ω(c, d) Represents the skill node in the industry skill knowledge graph c and skill nodes d The connection weights between sim(c, d) Indicates skills c and skills d The correlation between β(d) Indicates skills d The learning cost factor is β(d) The value is to master the skill d Average time required (months); 3) Based on the learning path scores of each learning path, select the path with the largest learning path score as the optimal learning path from skill b to skill a.

[0038] After obtaining the characteristics of the skills currently mastered by the applicant and the skills required for the position, the optimal learning path between the two is calculated based on the network structure in the skill knowledge graph. The present invention also specifically proposes a learning path scoring function that can be quantified based on the degree of association (difficulty) and learning cost of different learning paths, thereby serving as the basis for the evaluation of the optimal learning path. In this way, the optimal path from the skills currently mastered by the applicant to the skills required for the position can be accurately represented based on the skill knowledge graph, thereby providing support for subsequent potential value evaluation.

[0039] Preferably, the skill potential estimation unit comprises: According to the obtained optimal learning path, the potential value of the candidate's skills corresponding to the recruitment position skills is calculated, and the potential value calculation function used is: ; In the formula, Qua(a, b) Indicates the candidate's skills b Skills for the corresponding job positions a Potential value, variable d Representing skills in the skill knowledge graph b To Skills a The best learning path I skill nodes on d ≠ b ; ω(d) Represents a skill node d The node weight in the skill knowledge graph, sim(d) Represents the best learning path I Medium Skill Node d The degree of correlation with the previous skill node; β(d) Indicates skills d The learning cost factor of ; represents the set cost attenuation coefficient, where ; σ 2 Indicates corresponding skills aand skills b The node weight variance of all skill nodes under the same category; According to the potential value of the applicant's skills corresponding to each recruitment position, the recruiter's skill potential assessment results are obtained QU .

[0040] Usually, after obtaining the potential value of the candidate's skills for each recruitment position, the corresponding skill potential assessment result is obtained by adding them up. However, in other scenarios, the corresponding skill potential assessment result can also be determined based on the minimum value (only considering the weakest board) or average value (considering the overall situation) of the potential value of each recruitment skill, which can be reasonably set according to different recruitment ideas and recruitment scenarios.

[0041] Among them, the skills under the same category mentioned above in the present invention, wherein the classification referred to is the skill field, for example, corresponding to programming skills, such as Java, distributed systems, cloud computing, etc., all belong to the skills under the same category, and for positions that meet the requirements, foreign language skills, such as English, Spanish, etc., may also be required, which belong to skills under another category.

[0042] The above-mentioned embodiment of the present invention, after obtaining the best learning path between the applicant's current skills and the skills required for the position, further quantifies the cost and difficulty of the applicant's learning to obtain the skills required for the position based on the proposed potential value calculation function and the importance and learning cost of each skill node in the optimal path as consideration factors, thereby objectively and accurately evaluating the applicant's skill potential. The reliability and intelligence level of the applicant's potential consideration are improved. It can serve as an important consideration factor to provide a basis for the subsequent recruitment evaluation of applicants.

[0043] Taking into account the cultural preferences of applicants, such as attitude towards work, etc., which will become an important factor in whether the applicant can adapt to the position and corporate culture after joining the company, the present invention also proposes a method of evaluating the compatibility of applicants and corporate culture, thereby improving the pertinence and reliability of recruitment interview evaluation through the dimension of cultural adaptation.

[0044] Preferably, the cultural adaptation assessment model includes: The company culture introduction information and the applicant's personal introduction information are used to construct an input set, and the input set is input into the trained cultural adaptation analysis model to obtain the applicant's cultural adaptation evaluation result output by the cultural adaptation analysis model; The input set is in the form of {corporate culture introduction text, applicant personal introduction text}; The cultural adaptation analysis model is built based on the BERT model, including an input layer, an encoding layer and an output layer connected in sequence, wherein the input layer performs word segmentation, text completion and character encoding processing on the corporate culture introduction text and the applicant's personal introduction text in the input set, respectively, to obtain the character encoding vector sequence of the two texts; the encoding layer adopts 12 consecutive Transformers, each of which adopts a multi-head attention mechanism for feature extraction, and adds a fully connected layer after the last Transformer to output a 256-dimensional cultural feature vector; the encoding layer extracts cultural feature vectors from the character encoding vector sequences of the two texts, and the output layer compares the similarity of the two cultural feature vectors obtained, wherein the similarity comparison function adopted is the cos function, to obtain the similarity between the two. CA The results are used as cultural fit assessment results.

[0045] In one scenario, taking a 120-word corporate culture introduction text as an example, the corporate culture introduction text is input into the cultural adaptation analysis model, and 80 words are obtained through word segmentation based on the input layer. After character encoding processing, an 80×768 character encoding vector sequence is obtained; and after 12 consecutive Transformer attention processing, the corporate culture semantic features are obtained, and the cultural feature vector is extracted through the fully connected layer, where the activation function used in the fully connected layer is Sigmoid, and the final output is a 256-dimensional corporate culture feature vector; similarly, the above processing is performed on the applicant's personal introduction text to obtain the 256-dimensional applicant cultural feature vector corresponding to the applicant's personal introduction text; finally, the similarity of the two (cultural feature vectors) is calculated based on the similarity calculation function, thereby obtaining the similarity between the applicant's cultural inclination and the corporate culture inclination; the corporate culture inclination can be understood as including, for example, "work-life balance" or "pursuit of competition".

[0046] The above-mentioned implementation mode of the present invention can evaluate the cultural attributes of applicants from the perspective of corporate culture adaptation by extracting cultural tendencies from corporate culture introduction information and applicant personal introduction information, thereby assisting in finding culturally compatible applicants, thereby improving the applicants' subsequent adaptability to the company, reducing the probability of personnel mismatch and short-term personnel loss due to cultural conflicts, and reducing the overall cost of talent recruitment.

[0047] Preferably, the interview assessment module includes a video extraction unit, an emotion assessment unit, a stress assessment unit and an interview assessment unit; wherein, The video extraction unit is used to obtain the interview video recording data of the applicant, and obtain the corresponding video image information and audio information according to the obtained interview video recording data; wherein the interview video recording data is based on the video recording obtained by the smart camera during the interview of the applicant; The emotion evaluation unit is used to track the facial features of the applicant based on the video image data of the applicant in the first key time period based on the face tracking model, and calculate the expression energy value of the applicant based on the tracking results of the facial feature points, wherein the expression energy value calculation function used is: ; In the formula, EVM The variable represents the energy value of the candidate's expression in response to the interview questions. The higher the energy value, the more unstable the candidate's emotions. t = 1, 2, … T , t Indicates the moment in the critical time period. T Indicates the total duration of the critical time period; ∆P(t) express t The average change of the facial key point coordinates at each moment compared to the previous moment, ∆T t express t The optical flow change of the applicant’s face area at each moment; The stress evaluation unit is used to obtain the stress response value of the applicant based on the audio data of the applicant in the second key time period based on the audio analysis technology, wherein the stress response value calculation function used is: ; In the formula, SRC Indicates the stress response value of the candidate to the interview questions. The higher the stress response value, the greater the stress of the candidate. σ(f t ) and μ(f t ) They represent the standard deviation and mean of the frequency of the audio data of the applicant at each moment in the second key time period, N pauce Indicates the number of abnormal pauses of the applicant in the second key time period. The audio data is detected based on the VAD algorithm. When the audio gap is greater than the preset standard time T pauce When an abnormal pause is recorded, T pauce ∈ [0.5s, 1s] ; T Indicates the total duration of the audio data; α Represents the preset sound fluctuation weight, where α∈ [0.4,0.6] ; β represents the preset pause frequency weight, where β∈[0.3,0.7] ; The interview evaluation unit is used to calculate the interview evaluation value as the interview evaluation result according to the obtained expression energy value and pressure response value, wherein the interview evaluation value calculation function used is: ; in FC Represents the interview evaluation value of the candidate, γ and δ Respectively represent the normalized weight factors set, respectively according to EVM and SRC The historical maximum value is obtained by statistics. EVM Indicates the energy value of the candidate's expression in response to interview questions. SRC Indicates the stress response value of the candidate to the interview questions.

[0048] In one scenario, the face tracking model is implemented using a face recognition and tracking model based on tracking of 68 facial key points.

[0049] In the interview scenario, the first key time period is 1-3 seconds after the interviewer asks a question. By capturing the video recording data at the corresponding moment, the corresponding expression energy value calculation and emotion evaluation results can be completed. The second key time period is the time period from the beginning of the candidate's answer to the interviewer's question to the end of the answer.

[0050] In one scenario, the interview video record data may be obtained by setting up a camera at the interview site to record the interview process, or by recording and extracting the video interview process to obtain the interview video record data.

[0051] The above-mentioned embodiment of the present invention provides a technical solution for evaluating the interview status of an applicant based on the interview video recording data of the applicant, wherein firstly, based on the obtained interview video recording data, the audio of the image and the picture are extracted respectively, so as to lay the foundation for the subsequent emotional evaluation and stress evaluation. Wherein, based on the progress of the interview, the key time period is also marked, so as to carry out a targeted evaluation of the interview status of the applicant. Wherein, based on the proposed expression energy calculation function, the facial expression change characteristics of the applicant after the interview question is raised are extracted, and the degree of sudden change of the applicant's expression is reflected by the expression energy, so as to reflect the emotional stability of the applicant in the face of the recruitment question (the applicant is in a nervous state, because the facial muscles will have an instantaneous reaction, by capturing the characteristics of this instantaneous facial expression change, it can reflect the applicant's nervousness for the interview, thereby reflecting the applicant's interview status, such as nervousness or concealment), further, based on the proposed stress response value calculation function, it can be analyzed based on the audio data of the applicant in the process of answering the interview question, and the stress state of the interviewer in answering the interview question can be evaluated by the frequency characteristic change and pause characteristics of the audio characteristic signal. Finally, the interview status is evaluated based on the obtained expression energy value and stress response value, so as to provide objective feedback on the candidate's interview status, avoiding the deviation of the candidate's interview evaluation due to the interviewer's subjective evaluation, which helps to improve the objectivity of recruitment interview evaluation.

[0052] Preferably, the reporting module includes: Output the candidate's skill potential assessment results, cultural fit assessment results, and interview assessment results separately to generate a comprehensive assessment result of the candidate; or, A comprehensive evaluation is performed based on the obtained candidate's skill potential assessment results, cultural adaptation assessment results and interview assessment results to obtain the candidate's comprehensive assessment result; wherein the comprehensive evaluation function used is: ; In the formula, Score Indicates the comprehensive evaluation value of the applicant. QU Indicates the result of the candidate's skill potential assessment. CA Indicates the cultural fit assessment result of the candidate. FC Indicates the interview evaluation results of the candidate. ω 1 , ω 2 and ω 3 They respectively represent the set skill potential weight factor, cultural adaptation weight factor and interview weight factor.

[0053] Among them, according to actual needs, the comprehensive evaluation results of the applicant can be output separately according to the three evaluation results of the applicant, so as to obtain the interview evaluation report of the applicant; among them, the larger the value of the three evaluation results, the better the result achieved by the applicant in the corresponding evaluation item.

[0054] In another scenario, the three evaluation results can be further comprehensively quantified to obtain interview evaluation scores in the same dimension, thereby providing an intuitive and direct representation of the applicant's situation.

[0055] By using the output evaluation results as a basis, it can provide data support for the company to determine the final candidates for job recruitment and assist managers in selecting the most suitable managers for the company's positions.

[0056] In one scenario, the three assessment results are normalized into percentage outputs, for example, a skill potential assessment result of 90, a cultural fit assessment result of 20, and an interview assessment result of 80. If the lowest score is lower than the minimum standard (for example, 30), the applicant is judged to be unsuitable for the position in the company; or if the skill potential assessment result is 70, the cultural fit assessment result is 75, and the interview assessment result is 80, all of which meet the set minimum standards (for example, not less than 70), the applicant is hired.

[0057] In another scenario, the three evaluation results can also be used as part of many evaluation results, that is, combined with the evaluation results obtained by other means to comprehensively evaluate the degree of suitability of the applicant for the position.

[0058] In another scenario, the comprehensive evaluation value of the applicant can be used as a hard indicator (for example, it needs to be higher than the minimum standard) or a reference indicator to influence the final decision on whether to hire the applicant, so as to adapt to the application requirements of different recruitment scenarios and recruitment positions.

[0059] It should be noted that each functional unit / module in each embodiment of the present invention may be integrated into one processing unit / module, or each unit / module may exist physically separately, or two or more units / modules may be integrated into one unit / module. The above-mentioned integrated unit / module may be implemented in the form of hardware or in the form of software functional unit / module.

[0060] Through the description of the above implementation modes, it can be clearly understood by those skilled in the art that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code or any appropriate combination thereof. For hardware implementation, the processor can be implemented in one or more of the following units: an application specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a processor, a controller, a microcontroller, a microprocessor, other electronic units designed to implement the functions described herein or a combination thereof. For software implementation, part or all of the processes of the embodiment can be completed by instructing the relevant hardware through a computer program. When implemented, the above program can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein the communication media include any medium that facilitates the transmission of computer programs from one place to another. The storage medium can be any available medium that can be accessed by a computer. Computer-readable media can include, but are not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer.

[0061] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention, rather than to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should analyze that the technical solution of the present invention can be modified or replaced by equivalents without departing from the essence and scope of the technical solution of the present invention.

Claims

1. A recruitment interview evaluation system based on artificial intelligence, characterized in that: It includes information acquisition module, skill assessment module, cultural adaptation assessment module, interview assessment module and report module; among them, The information acquisition module is used to obtain the company's recruitment information and the corresponding applicant's resume information and applicant's interview information; The skill assessment module is used to extract job skill characteristics and candidate skill characteristics based on the company recruitment information and candidate resume information respectively; and to perform skill learning path analysis on job skill characteristics and candidate skill characteristics based on the established industry skill knowledge graph, and further obtain the candidate's skill potential assessment results based on the learning path analysis results; The cultural adaptation assessment module is used to extract corporate culture characteristics and candidate culture characteristics based on corporate recruitment information and candidate resume information, and conduct cultural adaptation analysis based on corporate culture characteristics and candidate culture characteristics to obtain the candidate's cultural adaptation assessment results; The interview evaluation module is used to analyze the interview performance of the applicant based on the interview video record data of the applicant to obtain the interview evaluation result; The report module is used to obtain a comprehensive assessment result of the applicant for the company's position based on the applicant's skill potential assessment results, cultural adaptability assessment results and interview assessment results.

2. The artificial intelligence-based recruitment interview evaluation system according to claim 1, characterized in that: The information acquisition module includes an enterprise recruitment information input unit, an applicant resume acquisition unit, and an applicant interview acquisition unit; among them, The enterprise recruitment information input unit is used to obtain enterprise recruitment information by scanning or inputting, wherein the enterprise recruitment information includes recruitment positions, job descriptions corresponding to the recruitment positions, and corporate culture introduction information; wherein the job description includes job responsibilities, qualifications and other requirements; The applicant resume acquisition unit is used to acquire the applicant's resume information, wherein the resume information includes the applicant's basic information, skills introduction, work experience and personal introduction; The candidate interview acquisition unit is used to record the candidate's interview video record data during the candidate's job interview.

3. The recruitment interview evaluation system based on artificial intelligence according to claim 1 is characterized in that: The skill assessment module includes a skill feature extraction unit, a skill path analysis unit, and a skill potential assessment unit; in, The skill feature extraction unit is used to extract job skill features and candidate skill features according to the enterprise recruitment information and candidate resume information respectively; The skill path analysis unit is used to analyze the skill learning path of the job skill characteristics and the candidate skill characteristics based on the established industry skill knowledge graph, and obtain the best learning path between the candidate skill characteristics and the job skill characteristics; The skill potential estimation unit is used to calculate the skill potential value of the applicant as a skill potential evaluation result based on the obtained optimal learning path.

4. The recruitment interview evaluation system based on artificial intelligence according to claim 3 is characterized in that: The skills assessment module also includes a map-building unit; Among them, the graph building unit is used to build a skill knowledge graph based on the job skill database, industry technical manual database, skill training course database, and talent resume database, and to build a learning relationship network between industry skills through the skill knowledge graph; specifically, it includes: 1) Entity extraction is performed based on the job skill database, industry technical manual database, and skill training course database to obtain network node information corresponding to the skills and build a skill knowledge graph network; the extracted entities are industry skills, and each point in the constructed knowledge graph network corresponds to an industry skill; 2) Calculate the weights between network points based on the constructed skill knowledge graph network, where the weight calculation function used is: ； in, ω(i,j) Represents a skill node i and skill nodes j The connection weights between sim(i,j) Indicates skills i and skills j The correlation between them is obtained based on the talent resume database, where , NumJ Indicates the total number of talent resumes. Numj(i,j) Indicates skills i and skills j Total number of items appearing on the same resume; Numz(i,j) In the process of entity extraction, skills i and skills j The number of times the same content appears. NumZ Indicates the total number of contents; DDEC(i,j) Indicates skills i and skills j The time difference factor between them is obtained based on the difference between the time information when the two are first recorded; DHOT(i,j) Indicates skills i and skills j The heat difference factor between them is obtained based on the difference in their current frequencies of occurrence in the content.

5. The recruitment interview evaluation system based on artificial intelligence according to claim 3 is characterized in that: The Skill Path Analysis Unit includes: 1) Based on the job skills obtained under the same category a and candidate skills b ; 2) Retrieve candidates’ skills based on the established industry skills knowledge graph b To Skills a The learning paths of , respectively, calculate the learning path score of each path, where the learning path score function used is: ； In the formula, Point(i,b,a) Indicates that from the skill node b To skill node a The first i Learning path scores for learning paths, c,d∈i Representation variables c. d. Belong to i The adjacent skill nodes in the learning path, where the path direction is from c arrive d ; ω(c,d) Represents the skill node in the industry skill knowledge graph c and skill nodes d The connection weights between sim(c,d) Indicates skills c and skills d The correlation between β(d) Indicates skills d The learning cost factor is β(d) The value is to master the skill d The average time required; 3) Based on the learning path scores of each learning path, select the path with the largest learning path score as the optimal learning path from skill b to skill a.

6. The artificial intelligence-based recruitment interview evaluation system according to claim 5, characterized in that: The Skill Potential Estimation Unit includes: According to the obtained optimal learning path, the potential value of the candidate's skills corresponding to the recruitment position skills is calculated, and the potential value calculation function used is: ； In the formula, Qua(a,b) Indicates the candidate's skills b Skills for the corresponding job positions a Potential value, variable d Representing skills in the skill knowledge graph b To Skills a The best learning path I skill nodes on d≠b ; ω(d) Represents a skill node d The node weight in the skill knowledge graph, sim(d) Represents the best learning path I Medium Skill Node d The degree of correlation with the previous skill node; β(d) Indicates skills d The learning cost factor; represents the set cost attenuation coefficient, where ; σ 2 Indicates corresponding skills a and skills b The node weight variance of all skill nodes under the same category; According to the potential value of the applicant's skills corresponding to each recruitment position, the recruiter's skill potential assessment result is obtained.

7. The recruitment interview evaluation system based on artificial intelligence according to claim 1 is characterized in that: The cultural fit assessment model includes: The company culture introduction information and the applicant's personal introduction information are used to construct an input set, and the input set is input into the trained cultural adaptation analysis model to obtain the applicant's cultural adaptation evaluation result output by the cultural adaptation analysis model; The input set is in the form of {corporate culture introduction text, applicant personal introduction text}; Among them, the cultural adaptation analysis model is built based on the BERT model, including an input layer, an encoding layer and an output layer connected in sequence, wherein the input layer performs word segmentation, text completion and character encoding processing on the corporate culture introduction text and the applicant's personal introduction text in the input set, respectively, and obtains the character encoding vector sequence of the two texts respectively; the encoding layer adopts 12 consecutive Transformers, each of which adopts a multi-head attention mechanism for feature extraction, and a fully connected layer is added after the last Transformer to output a 256-dimensional cultural feature vector; the encoding layer extracts cultural feature vectors from the character encoding vector sequences of the two texts respectively, and the output layer compares the similarity of the two cultural feature vectors obtained, wherein the similarity comparison function adopted is the cos function, and the similarity result of the two is obtained as the cultural adaptation evaluation result.

8. The recruitment interview evaluation system based on artificial intelligence according to claim 1, characterized in that: The interview assessment module includes a video extraction unit, an emotion assessment unit, a stress assessment unit and an interview assessment unit; among them, The video extraction unit is used to obtain the interview video recording data of the applicant, and obtain the corresponding video image information and audio information according to the obtained interview video recording data; wherein the interview video recording data is based on the video recording obtained by the smart camera during the interview of the applicant; The emotion evaluation unit is used to track the facial features of the applicant based on the video image data of the applicant in the first key time period based on the face tracking model, and calculate the expression energy value of the applicant based on the tracking results of the facial feature points, wherein the expression energy value calculation function used is: ； In the formula, EVM Indicates the energy value of the candidate's expression in response to the interview questions. The higher the expression energy value, the more unstable the candidate's emotions. t=1,2,…T , t Indicates the moment in the critical time period. T Indicates the total duration of the critical time period; ∆P(t) express t The average change of the facial key point coordinates at each moment compared to the previous moment, ∆T t express t The optical flow change of the applicant’s face area at each moment; The stress evaluation unit is used to obtain the stress response value of the applicant based on the audio data of the applicant in the second key time period based on the audio analysis technology, wherein the stress response value calculation function used is: ； In the formula, SRC Indicates the stress response value of the candidate to the interview questions. The higher the stress response value, the greater the stress of the candidate. σ(f t ) and μ(f t ) They represent the standard deviation and mean of the frequency of the audio data of the applicant at each moment in the second key time period, N pauce Indicates the number of abnormal pauses of the applicant in the second key time period. The audio data is detected based on the VAD algorithm. When the audio gap is greater than the preset standard time T pauce When an abnormal pause is recorded, T pauce ∈[0.5s,1s] ; T Indicates the total duration of the audio data; α Represents the preset sound fluctuation weight, where α∈[0.4, 0.6] ; β represents the preset pause frequency weight, where β∈[0.3,0.7] ; The interview evaluation unit is used to calculate the interview evaluation value as the interview evaluation result according to the obtained expression energy value and pressure response value, wherein the interview evaluation value calculation function used is: ; in FC Represents the interview evaluation value of the candidate, γ and δ Respectively represent the normalized weight factors set, respectively according to EVM and SRC The historical maximum value is obtained by statistics. EVM Indicates the energy value of the candidate's expression in response to interview questions. SRC Indicates the stress response value of the candidate to the interview questions.

9. The recruitment interview evaluation system based on artificial intelligence according to claim 1, characterized in that: Reporting modules include: Output the candidate's skill potential assessment results, cultural fit assessment results, and interview assessment results separately to generate a comprehensive assessment result of the candidate; or, A comprehensive evaluation is performed based on the candidate's skill potential assessment results, cultural adaptability assessment results and interview assessment results to obtain a comprehensive assessment result of the candidate.

Images