Artificial intelligence-based recruitment interview evaluation method and system, and storage medium

Abstract

The invention discloses a recruitment interview evaluation method and system based on artificial intelligence, and a storage medium, relates to the technical field of recruitment interview evaluation, and solves the problems that in the prior art, lexical analysis and pragmatic reasoning are completed only through voice recognition and corpus comparison, dynamic features of voice signals are not deeply mined, and the efficiency of recruitment interview evaluation is improved. The technical problems of single evaluation dimension and difficulty in comprehensively and objectively reflecting the language expression ability, the psychological state and the character matching degree of the applicant due to the fact that the character description in the resume and the real-time emotional expression in the interview are not subjected to association analysis in the prior art are solved. The method comprises the following steps: acquiring voice data and emotion data of an applicant; the method comprises the following steps: preprocessing voice data to obtain a formant sequence corresponding to the voice data; calculating a voice evaluation coefficient based on the formant sequence; performing calculation based on the emotion data to obtain emotion similarity; calculating an interview evaluation coefficient based on the voice evaluation coefficient and the emotion similarity; and judging whether the applicant is qualified based on the interview evaluation coefficient. The technical problem is solved.

Description

Technical Field

[0001] The present invention belongs to the field of recruitment interview evaluation, and specifically relates to a recruitment interview evaluation method, system and storage medium based on artificial intelligence. Background Art

[0002] In today's highly competitive job market, recruitment interviews are a crucial step in talent selection, and the accuracy and fairness of their evaluations are crucial. Traditional recruitment interview evaluation methods rely primarily on the subjective judgment of interviewers, which presents numerous shortcomings. For example, interviewers may be influenced by personal biases, emotional states, and fatigue, resulting in less than objective evaluation results. Furthermore, traditional methods struggle to comprehensively and in real time record and analyze a candidate's various interactions during the interview process, such as verbal expression, body language, and emotional changes, resulting in a lack of depth and comprehensiveness in the evaluation. Furthermore, as the scale of recruitment expands, the inefficiency of traditional methods is becoming increasingly prominent.

[0003] The prior art (invention patent with application number 2021108786489) discloses an intelligent interview method based on expression analysis, including: the interview begins; real-time voice data collection; the server performs voice recognition on the applicant's answer to obtain text information; the server compares the text information with the vocabulary information in the corpus to complete lexical analysis; the server completes pragmatic reasoning based on the lexical analysis results; the server generates an evaluation result based on the lexical analysis and pragmatic reasoning results; the server pushes the evaluation result to the recruiter in real time. This invention patent realizes a practical and targeted talent assessment interview full-process product, which enables recruiters to screen out talents that match the recruitment position, so that each job applicant can get a more fair and objective personality evaluation. In addition, not only does the interview complete the traditional interview assessment of job seekers, but it also connects to the data processing process, making the personality assessment results more fair, objective and accurate, thereby improving the efficiency of corporate talent selection. However, existing technologies only complete lexical analysis and pragmatic reasoning through speech recognition and corpus comparison, without deeply exploring the dynamic characteristics of speech signals (such as the fractal dimension of the resonance peak sequence, Lyapunov exponent and other nonlinear dynamic characteristics), nor do they conduct correlation analysis between the personality description in the resume and the real-time emotional expression in the interview. As a result, the evaluation dimension is single and it is difficult to fully and objectively reflect the applicant's language expression ability, psychological state and personality matching.

[0004] Therefore, the present invention solves the above problems by proposing a recruitment interview evaluation method based on artificial intelligence. Summary of the Invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a recruitment interview evaluation method, system and storage medium based on artificial intelligence, which are used to solve the technical problem that the prior art only completes lexical analysis and pragmatic reasoning through speech recognition and corpus comparison, does not deeply explore the dynamic characteristics of the speech signal, and does not correlate the personality description in the resume with the real-time emotional expression in the interview, resulting in a single evaluation dimension and difficulty in comprehensively and objectively reflecting the applicant's language expression ability, psychological state and personality matching.

[0006] To achieve the above objectives, the first aspect of the present invention provides a recruitment interview evaluation method based on artificial intelligence, comprising:

[0007] Obtain the applicant's voice data and emotional data; emotional data includes resume personality data and interview emotional data;

[0008] By preprocessing the speech data, a formant sequence corresponding to the speech data is obtained;

[0009] The speech evaluation coefficient is calculated based on the formant sequence;

[0010] Emotional similarity is calculated based on the emotional data;

[0011] The interview evaluation coefficient is calculated based on the voice evaluation coefficient and the emotional similarity;

[0012] Determine whether the applicant is qualified based on the interview evaluation coefficient.

[0013] Preferably, the acquiring of the applicant's voice data and emotion data includes:

[0014] The data collection device collects the applicant's voice data, image data, and resume personality data; wherein the voice data refers to the applicant's voice answers during the interview; the image data refers to several images of the applicant during the interview;

[0015] By analyzing the image data, the interview emotion data is obtained.

[0016] It should be noted that the resume personality data refers to the personality type described by the applicant in his / her resume or the personality type answered by the applicant when the interviewer asks about the applicant's personality during the interview; the interview emotion data refers to the emotion type displayed by the applicant during the interview.

[0017] Preferably, the analyzing the image data includes:

[0018] Use the FER2013 dataset as training data;

[0019] Extract facial images and corresponding emotions from the FER2013 dataset;

[0020] Integrate facial images as standard input data, and integrate facial images and corresponding emotion types as standard output data;

[0021] Training an artificial intelligence model based on standard input data and standard output data to obtain an emotion recognition model; wherein the artificial intelligence model includes: a convolutional neural network or a deep belief network;

[0022] The image data is identified based on the emotion recognition model to obtain interview emotion data; wherein, the interview emotion data refers to the emotion type corresponding to the image data; the emotion types include: confidence, nervousness, happiness, fear, etc.

[0023] It should be noted that the FER2013 dataset is a dataset for facial expression recognition, which is provided by researchers from the University of California, Los Angeles and the International Face Recognition Competition (FERET). It contains 35,887 images, and the size of each image is 48x48 pixels.

[0024] Preferably, the preprocessing of the speech data to obtain a formant sequence corresponding to the speech data includes:

[0025] Convert voice data into digital signals;

[0026] Perform fast Fourier transform on the digital signal corresponding to each frame of speech data, and obtain the formant sequence corresponding to the speech data through a formant estimation method; wherein the formant estimation method includes: linear prediction cepstral coefficient method, autocorrelation method or cepstral filtering method.

[0027] Preferably, the calculation of the speech evaluation coefficient based on the formant sequence includes:

[0028] The formant sequence is analyzed by the box dimension algorithm to obtain the D value corresponding to the speech data;

[0029] The Wolf algorithm is used to analyze the formant sequence and obtain the maximum Lyapunov exponent corresponding to the speech data;

[0030] The D value is labeled D and the maximum Lyapunov exponent is labeled L;

[0031] The speech evaluation coefficient is calculated using the formula P=A1×ln(D+1)×tanh(D)+A2 / e^L; where P is the speech evaluation coefficient, and A1 and A2 are weight coefficients.

[0032] It should be noted that the weight coefficients A1 and A2 are set by those skilled in the art based on practical experience;

[0033] In the analysis of candidate speech, a larger D value (fractal dimension) indicates greater dynamic complexity and structural richness in the speech signal. A high D value means that the speech's time-domain waveform or frequency-domain characteristics (such as formant trajectories) exhibit more complex fluctuation patterns. This corresponds to the candidate's varied intonation, natural rhythmic changes, and even the diversity of spectral structures brought about by emotional engagement. For example, when expressing their views, confident and calm candidates often have richer rhythmic variations due to their rich emotions. This complexity, quantified by fractal dimension, is reflected as a larger D value. Conversely, nervous or monotonous speech, due to its stiff rhythm and flat intonation, typically has a smaller D value. Therefore, a larger D value indicates a candidate's strong language expression ability and a quantitative reflection of positive emotional expression, making it a positive indicator in interview assessments.

[0034] In an interview situation, if the maximum Lyapunov exponent of an applicant's speech is small, it means that the dynamic trajectory of their vocal system has strong convergence in phase space, that is, the fluctuations in intonation, speaking speed and rhythm when speaking are more stable, and there are fewer irregular changes caused by tension and anxiety. For example, the adjacent trajectory divergence rate of the speech signal of a calm and clear-minded applicant is low, and the corresponding exponent tends to be negative, reflecting the coherence of expression and stability of mentality. For applicants with large emotional fluctuations or nervousness, the speech dynamics system is more likely to be sensitive to initial conditions, causing the exponent to tend to be positive, manifested as sudden changes in intonation or disordered rhythm. Therefore, a small maximum Lyapunov exponent is usually regarded as a sign of a stable psychological state and calm expression of the applicant, which is more advantageous in interview assessment.

[0035] Preferably, calculating the emotion similarity based on the emotion data includes:

[0036] The keywords in the resume personality data and interview sentiment data are represented as fixed-dimensional vectors through a word vector model to obtain resume personality word vector data and interview sentiment word vector data; wherein the word vector model includes Word2Vec or GloVe;

[0037] The similarity between each word vector in the resume personality word vector data and each word vector in the interview emotion word vector data is calculated using a similarity algorithm, and the average is taken to obtain the emotion similarity; wherein, the similarity algorithm includes: cosine similarity, Pearson correlation coefficient or Jaccard similarity.

[0038] Preferably, the interview evaluation coefficient is calculated based on the voice evaluation coefficient and the emotion similarity, including:

[0039] Extract the speech evaluation coefficient P and mark the emotional similarity as Y;

[0040] The interview evaluation coefficient is calculated using the formula MS = α × ln (e^(P / (P+1))) + β × e^tanh (Y); where MS is the interview evaluation coefficient, and α and β are weight coefficients.

[0041] It should be noted that the weight coefficients α and β are set by those skilled in the art based on practical experience.

[0042] Preferably, the determining whether the applicant is qualified based on the interview evaluation coefficient includes:

[0043] Determine whether the interview evaluation coefficient is greater than the preset interview evaluation threshold; if yes, the applicant is deemed qualified; if not, the applicant is deemed unqualified.

[0044] It should be noted that the preset interview assessment threshold is set by those skilled in the art based on practical experience.

[0045] A second aspect of the present invention provides an artificial intelligence-based recruitment interview evaluation system, comprising: a data acquisition module, a data analysis module, and an interview evaluation module;

[0046] Data collection module: used to obtain applicants' voice data and emotional data;

[0047] Data analysis module: obtains the formant sequence corresponding to the speech data by preprocessing the speech data; calculates the speech evaluation coefficient based on the formant sequence; calculates the emotional similarity based on the emotional data; and,

[0048] The interview evaluation coefficient is calculated based on the voice evaluation coefficient and the emotional similarity;

[0049] Interview evaluation module: Determine whether the applicant is qualified based on the interview evaluation coefficient.

[0050] Preferably, the data analysis module is in communication and / or electrically connected with the data acquisition module and the interview evaluation module respectively.

[0051] A third aspect of the present invention provides a storage medium for recruitment interview evaluation based on artificial intelligence, wherein the storage medium stores computer program instructions, which implement the above-mentioned method steps when executed by a processor.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] The present invention obtains the applicant's voice data and multimodal emotional data including resume personality data and interview emotion data, pre-processes the voice data to obtain a formant sequence, and uses the box dimension algorithm and the Wolf algorithm to calculate the fractal dimension (D value) and the maximum Lyapunov exponent; obtains the voice evaluation coefficient by calculation to quantify the dynamic complexity of the voice signal and the stability of the vocal system, making up for the defect of the existing technology that does not mine the nonlinear dynamic characteristics of speech; at the same time, the resume personality and interview emotion keywords are converted into vectors through the word vector model, and the emotion similarity is calculated with the help of the similarity algorithm to realize the correlation analysis between the resume personality description and the real-time emotional expression of the interview, solving the existing The problem of a single technical evaluation dimension; finally, an interview evaluation coefficient is constructed based on the voice evaluation coefficient and emotional similarity, and the preset threshold is combined to judge whether the applicant is qualified, thereby comprehensively and objectively reflecting the applicant's language expression ability, psychological state and personality matching, improving the accuracy and fairness of recruitment interview evaluation, and solving the problem that the existing technology only completes lexical analysis and pragmatic reasoning through voice recognition and corpus comparison, without deeply exploring the dynamic characteristics of voice signals, and without correlating the personality description in the resume with the real-time emotional expression in the interview. This leads to a single evaluation dimension and makes it difficult to comprehensively and objectively reflect the applicant's language expression ability, psychological state and personality matching. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0055] Figure 1 Schematic diagram of the method steps of an embodiment of the present invention;

[0056] Figure 2 This is a schematic diagram of determining the eligibility of a job applicant according to an embodiment of the present invention;

[0057] Figure 3 Schematic diagram of system modules according to an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0059] See also Figure 1The first embodiment of the present invention provides a recruitment interview evaluation method based on artificial intelligence, comprising:

[0060] Obtain the applicant's voice data and emotional data; emotional data includes resume personality data and interview emotional data;

[0061] By preprocessing the speech data, a formant sequence corresponding to the speech data is obtained;

[0062] The speech evaluation coefficient is calculated based on the formant sequence;

[0063] Emotional similarity is calculated based on the emotional data;

[0064] The interview evaluation coefficient is calculated based on the voice evaluation coefficient and the emotional similarity;

[0065] Determine whether the applicant is qualified based on the interview evaluation coefficient.

[0066] Obtain candidate voice and emotion data, including:

[0067] The data collection device collects the applicant's voice data, image data, and resume personality data; wherein the voice data refers to the applicant's voice answers during the interview; the image data refers to several images of the applicant during the interview;

[0068] By analyzing the image data, the interview emotion data is obtained.

[0069] By analyzing the image data, including:

[0070] Use the FER2013 dataset as training data;

[0071] Extract facial images and corresponding emotions from the FER2013 dataset;

[0072] Integrate facial images as standard input data, and integrate facial images and corresponding emotion types as standard output data;

[0073] Training an artificial intelligence model based on standard input data and standard output data to obtain an emotion recognition model; wherein the artificial intelligence model includes: a convolutional neural network or a deep belief network;

[0074] The image data is identified based on the emotion recognition model to obtain interview emotion data; wherein, the interview emotion data refers to the emotion type corresponding to the image data; the emotion types include: confidence, nervousness, happiness, fear, etc.

[0075] By preprocessing the speech data, the formant sequence corresponding to the speech data is obtained, including:

[0076] Convert voice data into digital signals;

[0077] Perform fast Fourier transform on the digital signal corresponding to each frame of speech data, and obtain the formant sequence corresponding to the speech data through a formant estimation method; wherein the formant estimation method includes: linear prediction cepstral coefficient method, autocorrelation method or cepstral filtering method.

[0078] The speech evaluation coefficient is calculated based on the formant sequence, including:

[0079] The formant sequence is analyzed by the box dimension algorithm to obtain the D value corresponding to the speech data;

[0080] The Wolf algorithm is used to analyze the formant sequence and obtain the maximum Lyapunov exponent corresponding to the speech data;

[0081] The D value is labeled D and the maximum Lyapunov exponent is labeled L;

[0082] The speech evaluation coefficient is calculated using the formula P=A1×ln(D+1)×tanh(D)+A2 / e^L; where P is the speech evaluation coefficient, and A1 and A2 are weight coefficients.

[0083] Emotional similarity is calculated based on the emotional data, including:

[0084] The keywords in the resume personality data and interview sentiment data are represented as fixed-dimensional vectors through a word vector model to obtain resume personality word vector data and interview sentiment word vector data; wherein the word vector model includes Word2Vec or GloVe;

[0085] The similarity between each word vector in the resume personality word vector data and each word vector in the interview emotion word vector data is calculated using a similarity algorithm, and the average is taken to obtain the emotion similarity; wherein, the similarity algorithm includes: cosine similarity, Pearson correlation coefficient or Jaccard similarity.

[0086] The interview evaluation coefficient is calculated based on the voice evaluation coefficient and emotion similarity, including:

[0087] Extract the speech evaluation coefficient P and mark the emotional similarity as Y;

[0088] The interview evaluation coefficient is calculated using the formula MS = α × ln (e^(P / (P+1))) + β × e^tanh (Y); where MS is the interview evaluation coefficient, and α and β are weight coefficients.

[0089] See Figure 2 , judge whether the applicant is qualified based on the interview evaluation coefficient, including:

[0090] Determine whether the interview evaluation coefficient is greater than the preset interview evaluation threshold; if yes, the applicant is deemed qualified; if not, the applicant is deemed unqualified.

[0091] For example, in an interview for a software engineer position at an internet company, the candidate’s answer to the question “Please describe a distributed system optimization project you led” was given 2 minutes of voice and image data, while the personality description in the resume was “confident, logical, and a team player.”

[0092] 1. Voice data collection:

[0093] The sampling rate is 16kHz, the quantization is 16-bit, the noise reduction microphone is used for acquisition, the frame length is 25ms, and the frame shift is 10ms, resulting in a total of 480 frames of voice data.

[0094] 2. Image data acquisition:

[0095] The camera records at 720p resolution and 30fps, and extracts 100 frames of clear facial images.

[0096] 3. Resume personality data:

[0097] The key words are "confidence", "rigorous logic" and "teamwork".

[0098] 4. Interview sentiment data extraction:

[0099] The CNN model (ResNet-18) trained on the FER2013 dataset recognizes facial expressions, with the results being "confident (probability 0.75)" and "focused (0.68)".

[0100] 2. Calculation of speech evaluation coefficient;

[0101] 1. Formant sequence extraction;

[0102] Preprocessing: The speech is converted into a digital signal and pre-emphasized, and each frame of speech is subjected to a fast Fourier transform.

[0103] Formant estimation (linear prediction cepstral coefficient method, order p = 12):

[0104] Extract the F2 formant sequence (unit: Hz) of each frame. The first five frames of data are: 1300, 1350, 1400, 1380, 1420.

[0105] 2. Calculation of fractal dimension (D value);

[0106] Phase space reconstruction: embedding dimension m = 7, time delay τ = 12, generating 7-dimensional trajectory points.

[0107] Box dimension algorithm: Take δ = 0.05σ (σ = 80Hz), calculate the number of grids covering the trajectory at different scales, and linear fit to obtain D = 1.82 (which is in line with the range of 1.7-1.9 for excellent candidates).

[0108] It should be noted that σ here represents the standard deviation of the formant frequency, a statistic that measures the frequency fluctuation range of the speech signal. In this example, σ = 80 Hz indicates that the standard deviation of the fluctuation of the candidate's formant frequencies (such as F1 and F2) in the time domain is 80 Hz. δ is the grid size (or box side length) used to cover the phase space trajectory in the box dimension algorithm. In this example, δ = 0.05σ = 0.05 × 80 Hz = 4 Hz, which represents a grid divided into 4 Hz units on the frequency axis, used to count the number of boxes required to cover the formant trajectory.

[0109] 3. Calculation of the maximum Lyapunov exponent (L);

[0110] Wolf algorithm: Track the evolution of the distance between adjacent trajectory points, time window T = 100ms, and calculate L = 0.23 (L < 0.3, indicating that the speech dynamics system is stable).

[0111] 4. Calculation of speech evaluation coefficient P;

[0112] Assume that weight coefficients A1 = 0.6, A2 = 0.4;

[0113] The speech evaluation coefficient is calculated using the formula P = A1 × ln(D + 1) × tanh(D) + A2 / e^L;

[0114] Substituting the data into the data, we obtain P≈0.921.

[0115] 3. Emotional similarity calculation;

[0116] 1. Word vector conversion;

[0117] Resume personality word vector: Use Word2Vec (skip-gram model, dimension 100), "confidence" vector is V1, "logical rigor" is V2, and "teamwork" is V3.

[0118] Interview sentiment word vector: "confidence" is W1, and "focus" is W2.

[0119] 2. Cosine similarity calculation;

[0120] Calculate the cosine similarity of each pair of word vectors:

[0121] sim(V1,W1)=0.85, sim(V1,W2)=0.62;

[0122] sim(V2,W1)=0.58, sim(V2,W2)=0.71;

[0123] sim(V3,W1)=0.49, sim(V3,W2)=0.36;

[0124] Take the average:

[0125] Y=(0.85+0.62+0.58+0.71+0.49+0.36) / 6≈0.602.

[0126] 4. Calculation and judgment of interview evaluation coefficient;

[0127] 1. Calculation of interview evaluation coefficient MS:

[0128] Assume that the weight coefficients α = 0.7, β = 0.3;

[0129] The interview evaluation coefficient is calculated using the formula MS = α × ln (e^(P / (P+1))) + β × e^tanh (Y);

[0130] Substituting the data into the equation, we obtain: MS≈0.855.

[0131] 2. Eligibility judgment:

[0132] Assume that the preset threshold is 0.8, since 0.855>0.8, the candidate is determined to be qualified.

[0133] 5. Result analysis;

[0134] The voice evaluation coefficient P = 0.921: The D value of 1.82 indicates that the candidate's language expression is rich in rhythm (for example, the intonation rises and falls naturally when explaining technical details), and L = 0.23 reflects his stable mentality (no obvious sudden changes in intonation due to tension).

[0135] Emotional similarity Y = 0.602: The keywords of resume personality and interview emotion have a high match, especially the similarity of "confidence" reaches 0.85, which verifies the consistency between the candidate's self-description and actual performance.

[0136] Interview evaluation coefficient MS = 0.855: Comprehensively reflects the candidate's language expression ability and personality matching, exceeding the threshold, and meeting the requirements of the software engineer position for clear logic and stable mentality.

[0137] This example demonstrates, through full-process quantitative analysis, how the present invention breaks through the limitations of existing technologies in terms of voice dynamic features and emotion matching, achieving a more objective interview evaluation.

[0138] See Figure 3, the second embodiment of the present invention provides a recruitment interview evaluation system based on artificial intelligence, including: a data acquisition module, a data analysis module and an interview evaluation module;

[0139] Data collection module: used to obtain applicants' voice data and emotional data;

[0140] Data analysis module: obtains the formant sequence corresponding to the speech data by preprocessing the speech data; calculates the speech evaluation coefficient based on the formant sequence; calculates the emotional similarity based on the emotional data; and,

[0141] The interview evaluation coefficient is calculated based on the voice evaluation coefficient and the emotional similarity;

[0142] Interview evaluation module: Determine whether the applicant is qualified based on the interview evaluation coefficient.

[0143] Some of the data in the above formula are calculated by removing the dimensions and taking their numerical values. The formula is a formula that is closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0144] Working principle of the present invention:

[0145] The present invention obtains the voice data and emotion data of the applicant; obtains the resonance peak sequence corresponding to the voice data by preprocessing the voice data; calculates the voice evaluation coefficient based on the resonance peak sequence; calculates the emotion similarity based on the emotion data; calculates the interview evaluation coefficient based on the voice evaluation coefficient and the emotion similarity; and judges whether the applicant is qualified based on the interview evaluation coefficient. This solves the technical problem that the existing technology only completes lexical analysis and pragmatic reasoning through voice recognition and corpus comparison, does not deeply explore the dynamic characteristics of the voice signal, and does not correlate the personality description in the resume with the real-time emotion expression in the interview, resulting in a single evaluation dimension and difficulty in comprehensively and objectively reflecting the applicant's language expression ability, psychological state and personality matching.

[0146] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. The recruitment interview evaluation method based on artificial intelligence is characterized by: include: Obtain the applicant's voice data and emotional data; emotional data includes resume personality data and interview emotional data; By preprocessing the speech data, a formant sequence corresponding to the speech data is obtained; The speech evaluation coefficient is calculated based on the formant sequence; Emotional similarity is calculated based on the emotional data; The interview evaluation coefficient is calculated based on the voice evaluation coefficient and the emotional similarity; Determine whether the applicant is qualified based on the interview evaluation coefficient.

2. The artificial intelligence-based recruitment interview evaluation method according to claim 1, characterized in that: The acquisition of the candidate's voice data and emotion data includes: The data collection device collects the applicant's voice data, image data, and resume personality data; wherein the voice data refers to the applicant's voice answers during the interview; the image data refers to several images of the applicant during the interview; By analyzing the image data, the interview emotion data is obtained.

3. The artificial intelligence-based recruitment interview evaluation method according to claim 2, characterized in that: The analysis of the image data includes: Use the FER2013 dataset as training data; Extract facial images and corresponding emotions from the FER2013 dataset; Integrate facial images as standard input data, and integrate facial images and corresponding emotion types as standard output data; Training an artificial intelligence model based on standard input data and standard output data to obtain an emotion recognition model; wherein the artificial intelligence model includes: a convolutional neural network or a deep belief network; The image data is identified based on the emotion recognition model to obtain interview emotion data; wherein the interview emotion data refers to the emotion type corresponding to the image data.

4. The recruitment interview evaluation method based on artificial intelligence according to claim 1 is characterized in that: The method of preprocessing the speech data to obtain a formant sequence corresponding to the speech data includes: Convert voice data into digital signals; Perform fast Fourier transform on the digital signal corresponding to each frame of speech data, and obtain the formant sequence corresponding to the speech data through a formant estimation method; wherein the formant estimation method includes: linear prediction cepstral coefficient method, autocorrelation method or cepstral filtering method.

5. The recruitment interview evaluation method based on artificial intelligence according to claim 1 is characterized in that: The method of calculating the speech evaluation coefficient based on the formant sequence includes: The formant sequence is analyzed by the box dimension algorithm to obtain the D value corresponding to the speech data; The Wolf algorithm is used to analyze the formant sequence and obtain the maximum Lyapunov exponent corresponding to the speech data; The D value is labeled D and the maximum Lyapunov exponent is labeled L; The speech evaluation coefficient is calculated using the formula P=A1×ln(D+1)×tanh(D)+A2 / e^L; where P is the speech evaluation coefficient, and A1 and A2 are weight coefficients.

6. The artificial intelligence-based recruitment interview evaluation method according to claim 1, characterized in that: The calculating of the emotion similarity based on the emotion data includes: The keywords in the resume personality data and interview sentiment data are represented as fixed-dimensional vectors through a word vector model to obtain resume personality word vector data and interview sentiment word vector data; wherein the word vector model includes Word2Vec or GloVe; The similarity between each word vector in the resume personality word vector data and each word vector in the interview emotion word vector data is calculated using a similarity algorithm, and the average is taken to obtain the emotion similarity; wherein, the similarity algorithm includes: cosine similarity, Pearson correlation coefficient or Jaccard similarity.

7. The recruitment interview evaluation method based on artificial intelligence according to claim 1 is characterized in that: The interview evaluation coefficient is calculated based on the voice evaluation coefficient and the emotion similarity, including: Extract the speech evaluation coefficient P and mark the emotional similarity as Y; The interview evaluation coefficient is calculated using the formula MS = α × ln (e^(P / (P+1))) + β × e^tanh (Y); where MS is the interview evaluation coefficient, and α and β are weight coefficients.

8. The artificial intelligence-based recruitment interview evaluation method according to claim 1, characterized in that: The determination of whether an applicant is qualified based on the interview evaluation coefficient includes: Determine whether the interview evaluation coefficient is greater than the preset interview evaluation threshold; if yes, the applicant is deemed qualified; if not, the applicant is deemed unqualified.

9. An artificial intelligence-based recruitment interview evaluation system, which executes the artificial intelligence-based recruitment interview evaluation method according to any one of claims 1 to 8, characterized in that: include: Data collection module, data analysis module and interview evaluation module; Data collection module: used to obtain applicants' voice data and emotional data; Data analysis module: obtains the formant sequence corresponding to the speech data by preprocessing the speech data; The speech evaluation coefficient is calculated based on the formant sequence; Emotional similarity is calculated based on the emotional data; as well as, The interview evaluation coefficient is calculated based on the voice evaluation coefficient and the emotional similarity; Interview evaluation module: Determine whether the applicant is qualified based on the interview evaluation coefficient.

10. An artificial intelligence-based recruitment interview evaluation storage medium, characterized in that: The storage medium stores computer program instructions, which, when executed by a processor, implement the method steps described in any one of claims 1 to 8.

Images