Resume screening method and device fusing causal reasoning, equipment and medium
By constructing a causal directed graph and bias index, and combining preset rules to correct resume screening methods, the problem of lack of causal logic in traditional resume screening is solved, and a more accurate and fair resume screening is achieved.
Patent Information
- Application Number
- CN202510365129.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-26
- Publication Date
- 2025-07-04
AI Technical Summary
The traditional resume screening method lacks causal logic analysis, resulting in false correlations and data bias, affecting the accuracy and fairness of recruitment results.
By constructing a target causal directed graph, the causal relationship intensity and bias index between resume characteristics are determined, and data correction is carried out in combination with preset screening rules to achieve accurate screening.
Improve the accuracy and reliability of resume screening, eliminate false correlations and biases in the data, and ensure the fairness of the screening process.
Smart Images
Figure CN120258142A_ABST
Abstract
Description
Technical Field
[0001] The embodiments of the present application relate to the field of computer technology, and particularly to the field of information processing technology. Specifically, the present application relates to a resume screening method, device, equipment, and medium that integrates causal reasoning. Background Art
[0002] When traditional rule engines handle complex data relationships, they mainly rely on simple rule matching, lack in-depth insight into the internal causal logic of data, and are difficult to effectively identify false correlations. Most of the past fairness correction measures are only based on surface statistical adjustments and do not touch the causal relationships at the root of bias, so they cannot fundamentally solve the problem.
[0003] For example, in the human resources recruitment scenario, the graduation schools of candidates are often simply associated with their work capabilities. However, through in-depth causal analysis, it can be seen that there may not be a direct and necessary causal relationship between the graduation school and work ability. This incorrect association will lead to biases in the screening results, and enterprises may miss many excellent talents or introduce some inappropriate candidates. Summary of the Invention
[0004] The present application provides a resume screening method, device, equipment, and medium that integrates causal reasoning to improve the accuracy and reliability of resume screening.
[0005] According to one aspect of the present application, a resume screening method that integrates causal reasoning is provided. The method includes:
[0006] According to the preset resume screening rules for the target position, perform feature screening on the candidate structured data of at least one resume to be screened to obtain a target structured data set; wherein, the target structured data set includes the target structured data of the at least one resume to be screened; the candidate structured data refers to the structured data obtained after preprocessing the resume data of the resume to be screened;
[0007] According to the target structured data set, determine a target causal directed graph and a bias index set of the applicant characteristics in the target structured data set; wherein, the target causal directed graph is used to represent the causal relationship between the resume characteristics in the target structured data set; the resume characteristics include applicant characteristics and position matching characteristics;
[0008] Based on the target causal directed graph, according to the target structured data set, determine the causal relationship intensity set between the position characteristics in the target structured data set;
[0009] Determine the structured data to be matched for the at least one resume to be screened and the target resume screening rule for the target position according to the bias index set, the causal relationship strength set, the preset resume screening rule, and the target structured data of the at least one resume to be screened;
[0010] Screen the structured data to be matched of the at least one resume to be screened according to the target resume screening rule to obtain the target resume.
[0011] According to another aspect of the present application, there is provided a resume screening device integrating causal reasoning, and the device includes:
[0012] A feature screening module, configured to perform feature screening on the candidate structured data of at least one resume to be screened according to the preset resume screening rule of the target position to obtain a target structured data set; wherein, the target structured data set includes the target structured data of the at least one resume to be screened; the candidate structured data refers to the structured data obtained after preprocessing the resume data of the resume to be screened;
[0013] A directed graph determination module, configured to determine a target causal directed graph and a bias index set of the applicant characteristics in the target structured data set according to the target structured data set; wherein, the target causal directed graph is used to represent the causal relationship between the resume characteristics in the target structured data set; the resume characteristics include applicant characteristics and position matching characteristics;
[0014] A strength set determination module, configured to determine a causal relationship strength set between the position characteristics in the target structured data set according to the target structured data set based on the target causal directed graph;
[0015] A data correction module, configured to determine the structured data to be matched for the at least one resume to be screened and the target resume screening rule for the target position according to the bias index set, the causal relationship strength set, the preset resume screening rule, and the target structured data of the at least one resume to be screened;
[0016] A resume screening module, configured to screen the structured data to be matched of the at least one resume to be screened according to the target resume screening rule to obtain the target resume.
[0017] According to another aspect of the present application, there is provided an electronic device, and the electronic device includes:
[0018] One or more processors;
[0019] A memory for storing one or more programs;
[0020] When the one or more programs are executed by the one or more processors, the one or more processors implement any one of the resume screening methods integrating causal reasoning provided by the embodiments of the present application.
[0021] According to another aspect of the present application, there is provided a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements any one of the resume screening methods integrating causal reasoning provided by the embodiments of the present application.
[0022] According to another aspect of the present application, there is provided a computer program product, including a computer program, and when the computer program is executed by a processor, it implements any one of the resume screening methods integrating causal reasoning provided by the embodiments of the present application.
[0023] In the present application, by performing feature screening on the candidate structured data of at least one resume to be screened according to the preset resume screening rules of the target position, a target structured data set is obtained; wherein, the target structured data set includes the target structured data of at least one resume to be screened; the candidate structured data refers to the structured data obtained after preprocessing the resume data of the resume to be screened; according to the target structured data set, a target causal directed graph and a bias index set of the applicant characteristics in the target structured data set are determined; wherein, the target causal directed graph is used to represent the causal relationship between the resume characteristics in the target structured data set; the resume characteristics include applicant characteristics and position matching characteristics; based on the target causal directed graph, according to the target structured data set, a causal relationship strength set between the position characteristics in the target structured data set is determined; according to the bias index set, the causal relationship strength set, the preset resume screening rules and the target structured data of at least one resume to be screened, the to-be-matched structured data of at least one resume to be screened and the target resume screening rules of the target position are determined; according to the target resume screening rules, the to-be-matched resume data of at least one resume to be screened is screened to obtain a target resume. The above technical solution, by sorting out the causal relationship between each resume characteristic and combining the characteristic bias index, corrects the resume data of the user resume and the screening rules of the target position, realizes accurate identification and elimination of false correlations in the data, and comprehensively eliminates data bias, which helps to improve the accuracy and reliability of resume screening. Description of the Drawings
[0024] Figure 1 is a flowchart of a resume screening method integrating causal reasoning according to Embodiment 1 of the present application;
[0025] Figure 2 is a flowchart of a resume screening method integrating causal reasoning according to Embodiment 2 of the present application;
[0026] Figure 3It is a schematic structural diagram of a resume screening device integrating causal reasoning according to Embodiment 3 of the present application;
[0027] Figure 4 It is a schematic structural diagram of an electronic device for implementing the resume screening method integrating causal reasoning according to Embodiment 4 of the present application. Detailed implementation manners
[0028] In order to enable those skilled in the art to better understand the solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.
[0029] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data used in appropriate cases can be interchanged so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.
[0030] In addition, it should also be noted that in the technical solutions of the present application, the collection, storage, use, processing, transmission, provision, disclosure and other processing of relevant data such as the preset resume screening rules and candidate structured data comply with the provisions of relevant laws and regulations and do not violate public order and good customs.
[0031] Embodiment 1
[0032] Figure 1 It is a flowchart of a resume screening method integrating causal reasoning according to Embodiment 1 of the present application. This embodiment is applicable to the situation of uniformly screening the resumes of all applicants after offline tests, and can be executed by a resume screening device integrating causal reasoning. The resume screening device integrating causal reasoning can be implemented in the form of hardware and / or software, and the resume screening device integrating causal reasoning can be configured in a computer device, such as a server. As Figure 1 shown, the method includes:
[0033] S110. Screen the features of the candidate structured data of at least one resume to be screened according to the preset resume screening rules of the target position, and obtain the target structured data set.
[0034] In this embodiment, the target position refers to the position that needs to be recruited currently. The preset resume screening rules are screening rules set in advance according to the position requirements, such as rules that the working years are greater than 3 years, having a bachelor's degree in a specific major, etc. The resume to be screened refers to the resume of the applicant who has undergone an offline test after submitting the application. The candidate structured data refers to the structured data obtained after preprocessing the resume data of the resume to be screened; this data may include the applicant's age, applicant's education background, applicant's working years, applicant's work experience, applicant's project experience, whether the applicant has passed the interview, etc. The target structured data set refers to the data set that can be used for unified resume screening after preliminary screening of the candidate structured data; the target structured data set includes the target structured data of at least one resume to be screened.
[0035] Exemplarily, based on the standard features in the preset resume screening rules of the target position, the resume features in the candidate structured data of at least one resume to be screened that are inconsistent with the standard features are eliminated, and the target structured data of at least one resume to be screened is obtained.
[0036] Optionally, the determination method of the candidate structured data can be, for each resume to be screened, parse the resume data of the resume to be screened, extract key text data and resume structured data from the resume data of the resume to be screened; perform text extraction on the key text data to extract keywords, work content, skills, etc. from the key text data to obtain resume unstructured data; convert the resume unstructured data into a structured data representation form and merge it with the resume structured data to obtain the candidate structured data of the resume to be screened.
[0037] In this embodiment, the key text data refers to the text information in the resume data of the resume to be screened, which may include work experience, project experience, etc. The resume structured data refers to the structured data in the resume data of the resume to be screened, which may include education background, age, etc. The resume unstructured data refers to the text data after keyword processing of the key text data, which may include work experience, project experience, etc.
[0038] S120. Determine the target causal directed graph and the bias index set of the applicant features in the target structured data set according to the target structured data set.
[0039] In this embodiment, a causal directed graph is a graphical model where nodes represent variables and edges represent causal relationships between variables. The direction of each edge indicates the direction of the causal relationship, that is, which variable is the cause and which is the result target; the causal directed graph is used to characterize the causal relationships between various resume features in the target structured dataset; the resume features include applicant features and job matching features; among them, applicant features refer to relevant features used to characterize the personal information and personal capabilities of the applicant, which may include age, education level, working years, etc.; job matching features refer to relevant features used to characterize the matching degree between the applicant's resume and the target job, which may include job matching degree, salary expectation, whether the interview is passed, job competency score, etc. The bias index set refers to a set including at least one bias index of applicant features; the bias index refers to the degree of deviation of data or a model during the decision-making process, which reflects whether there is a certain unfair tendency or deviation in the decision or result; it is used to measure and correct possible biases in the screening process to ensure that the screening process is more fair and reasonable.
[0040] Optionally, the method for determining the target causal directed graph may be to determine at least one causal path according to the causal relationships between various resume features in the target structured dataset, and combine at least one causal path to obtain a candidate causal directed graph; based on the Bayesian principle, perform causal path screening on the candidate causal directed graph to obtain the target causal directed graph.
[0041] In this embodiment, the causal relationship describes the causal influence between variables (resume features), where one variable (the cause) directly or indirectly causes a change in another variable (the result). A causal path refers to a path composed of a series of causal relationships in the causal network, representing the process by which a variable affects another variable through a series of intermediate variables. A candidate causal directed graph refers to a possible causal relationship graph proposed through data mining or expert knowledge under preset conditions. Causal path screening refers to selecting the causal path that best conforms to the data from multiple candidate causal paths based on certain criteria (such as Bayesian inference, maximum likelihood estimation, etc.).
[0042] Furthermore, based on the Bayesian principle, performing causal path screening on the candidate causal directed graph to obtain the target causal directed graph may be to, for each causal path of the candidate causal directed graph, based on the Bayesian principle, determine the path conditional probability of this causal path according to the historical resume screening data; if the path conditional probability meets the path generation condition, then determine this causal path as the path of the final target causal directed graph.
[0043] In this embodiment, the path generation condition is set through a large number of experiments or artificially according to the actual situation or empirical values. For example, this condition may be that the path conditional probability is greater than or equal to the conditional probability threshold.
[0044] Exemplarily, taking the causal path X→Z→Y as an example, the path conditional probability P(X→Z→Y) of this path can be determined by the following formula:
[0045] P(X→Z→Y) = P(Z|X) × P(Y|Z) × β k ;
[0046] Wherein, P(Z|X) refers to the conditional probability from X to Z, which can be obtained by statistical analysis of historical resume screening data. P(Y|Z) refers to the conditional probability from Z to Y, which can be obtained by statistical analysis of historical resume screening data. β refers to the path credibility attenuation coefficient, generally 0 < β < 1. k is the number of variables in the causal path.
[0047] It can be understood that by introducing the path credibility attenuation coefficient, this formula reasonably attenuates the probability of a long causal path, avoids misjudgment of causal relationships caused by overly long paths, thereby accurately determining the causal paths in the data, effectively identifying and eliminating false correlations; for example, when analyzing the relationship between educational background, internship experience and work ability, it can accurately determine which causal paths are real and effective and which are false.
[0048] S130. Based on the target causal directed graph, determine the set of causal relationship strengths between each job feature in the target structured data set according to the target structured data set.
[0049] In this embodiment, the causal relationship strength is used to quantify the degree of influence of one feature on another; it is usually represented by a numerical value, such as using a correlation coefficient, a regression coefficient or other statistical indicators; the causal relationship strength reflects the strength of the relationship between variables. The set of causal relationship strengths refers to the set of the causal relationship strengths between all features in the data set; the causal relationship strength between each pair of features is quantified and organized into a set or list.
[0050] Optionally, the set of causal relationship strengths can be obtained by inputting the target structured data set into a causal model, and the set of causal relationship strengths between each job feature in the target structured data set is obtained. Wherein, the causal model is constructed according to the historical resume screening data.
[0051] In this embodiment, the causal model is a mathematical or graphical model used to represent and analyze the causal relationships between different features; the causal model can help study and quantify the causal influence between features, thereby revealing how different features interact with each other. The historical resume screening data refers to the resume data collected during the recruitment process, which usually includes the personal information, educational background, work experience, skills, etc. of job seekers, as well as the results of whether these resumes pass the screening.
[0052] Exemplarily, the causal model can be constructed by the following formula:
[0053]
[0054] Among them, C(X, Y) refers to the strength of the causal relationship between resume feature X and resume feature Y. x i refers to the i-th sample value of resume feature X. yi refers to the i-th sample value of resume feature Y. refers to the mean value of all sample values of resume feature X. refers to the mean value of all sample values of resume feature Y. a refers to the time impact coefficient used to measure the degree to which the causal relationship between variables is affected by the time interval Δt(X, Y); for example, when analyzing work experience and work performance, as the time interval between the work experience accumulation time and the performance evaluation time increases, the strength of the causal relationship may weaken.
[0055] It can be understood that through this formula, by comprehensively considering the statistical correlation between variables and the impact of time factors on the causal relationship, a causal model that better fits the actual business scenario is constructed, providing a key basis for subsequent precise processing.
[0056] S140. Determine the structured data to be matched for at least one resume to be screened and the target resume screening rule for the target position according to the bias index set, the causal relationship strength set, the preset resume screening rule, and the target structured data of at least one resume to be screened.
[0057] In this embodiment, the structured data to be matched refers to the corrected target structured data, which can truly reflect the ability level of the applicant. The target resume screening rule refers to the corrected resume screening rule.
[0058] Optionally, according to the bias index set, correct the eigenvalue and eigenweight of at least one applicant feature in the target structured data of at least one resume to be screened to obtain the structured data to be matched for at least one resume to be screened; according to the bias index set and the causal relationship strength set, correct the eigenweight of at least one resume feature in the preset resume screening rule to obtain the target resume screening rule.
[0059] Exemplarily, taking gender bias as an example, for the eigenvalue of the applicant feature underestimated for female applicants, the following correction formula is used:
[0060]
[0061] Among them, refers to the corrected eigenvalue of the applicant feature of female applicants. s f refers to the eigenvalue of the applicant feature of female applicants before correction. refers to the mean value of the eigenvalues of the applicant features of male applicants before correction. It refers to the mean value of the eigenvalue of the applicant characteristics of female applicants before correction. γ refers to the correction coefficient, which is generally greater than 0 and less than 1.
[0062] It can be understood that through this formula, the underestimated skill level data of female applicants can be appropriately adjusted to truly reflect their ability levels.
[0063] Exemplarily, the correction of the characteristic weight of applicant characteristics can be achieved through the following formula:
[0064]
[0065] Among them, is the corrected characteristic weight. w A is the characteristic weight before correction. ε is the weight correction factor, which is generally greater than 0. BI(A) refers to the bias index of applicant characteristics.
[0066] Furthermore, the determination method of the target resume screening rule can be that for each resume characteristic in the preset resume screening rule, if the bias index of this resume characteristic in the bias index set meets the weight correction condition, then according to the strength of the causal relationship, the characteristic weight of this resume characteristic is corrected to obtain the corrected characteristic weight; the corrected characteristic weights of at least one resume characteristic are updated to the preset resume rule to obtain the target resume screening rule.
[0067] In this embodiment, the weight correction condition is preset artificially through a large number of experiments and according to the actual situation or empirical values. For example, this condition can be that the bias index is greater than or equal to the bias index threshold.
[0068] Exemplarily, taking the example that the weight of the graduation school characteristic in the resume characteristic is too high and the weight of the actual work project experience characteristic is too low, the correction of the characteristic weight of the resume characteristic in the preset resume screening rule can be achieved through the following formula:
[0069]
[0070] Among them, refers to the corrected characteristic weight of the graduation school characteristic. w1 refers to the characteristic weight of the graduation school characteristic before correction. δ is the weight correction coefficient, which is generally greater than 0. C(graduation school, work performance) refers to the strength of the causal relationship between the graduation school and work performance. refers to the corrected characteristic weight of the actual work project experience characteristic. w2 refers to the characteristic weight of the actual work project experience characteristic before correction. C(actual work project experience, work performance) refers to the strength of the causal relationship between the actual work project experience and work performance.
[0071] It is understandable that through this formula, the feature weights are reasonably adjusted according to the causal relationship strength to ensure equal opportunities for different groups in the decision-making process.
[0072] S150. According to the target resume screening rules, screen the candidate resume data of at least one resume to be screened to obtain the target resume.
[0073] Exemplarily, according to the adjusted rules and data, determine whether the candidate meets the job requirements and whether to enter the next round of interviews, etc., so as to screen out the target resume from the resumes to be screened.
[0074] In the embodiment of the present application, by performing feature screening on the candidate structured data of at least one resume to be screened according to the preset resume screening rules of the target position, a target structured data set is obtained; wherein, the target structured data set includes the target structured data of at least one resume to be screened; the candidate structured data refers to the structured data obtained after preprocessing the resume data of the resume to be screened; according to the target structured data set, a target causal directed graph and a set of bias indices of the applicant characteristics in the target structured data set are determined; wherein, the target causal directed graph is used to represent the causal relationship between the resume characteristics in the target structured data set; the resume characteristics include applicant characteristics and job matching characteristics; based on the target causal directed graph, according to the target structured data set, a set of causal relationship strengths between the job characteristics in the target structured data set is determined; according to the set of bias indices, the set of causal relationship strengths, the preset resume screening rules, and the target structured data of at least one resume to be screened, the candidate structured data to be matched of at least one resume to be screened and the target resume screening rules of the target position are determined; according to the target resume screening rules, screen the candidate resume data of at least one resume to be screened to obtain the target resume. The above technical solution, by sorting out the causal relationship between each resume characteristic and combining the feature bias index, corrects the resume data of the user resume and the screening rules of the target position, realizes accurate identification and elimination of the false correlation in the data, comprehensively eliminates data bias, and helps to improve the accuracy and reliability of resume screening.
[0075] Embodiment 2
[0076] Figure 2It is a flowchart of a resume screening method integrating causal reasoning according to Embodiment 2 of the present application. On the basis of the technical solutions of the above embodiments, this embodiment refines "based on the target causal directed graph, and according to the target structured data set, determining the causal relationship strength between each job feature in the target structured data set" into "based on the causal path in the target causal directed graph, screening out at least one pair of resume feature groups from the resume features of the target structured data set; wherein, the resume feature group refers to two resume features with a causal relationship; for each pair of resume feature groups, according to the first feature value set of the first resume feature and the second feature value set of the second resume feature in the resume feature group, determining the causal relationship strength between the first resume feature and the second resume feature". It should be noted that for the parts not detailed in the embodiments of the present application, reference can be made to the relevant descriptions of other embodiments. As Figure 2 shown, the method includes:
[0077] S210. According to the preset resume screening rules of the target position, perform feature screening on the candidate structured data of at least one resume to be screened to obtain a target structured data set.
[0078] S220. According to the target structured data set, determine a target causal directed graph and a bias index set of the applicant features in the target structured data set.
[0079] Exemplarily, the bias index of the applicant feature can be determined by the following formula:
[0080]
[0081] wherein, BI(A) refers to the bias index of the applicant feature. a i refers to the value of the applicant feature A in the i-th sample (for example, 0 represents female and 1 represents male). refers to the mean value of all values of the applicant feature A. pi is the value of the job matching feature related to the decision (such as the selection probability, credit score, etc.) in the i-th sample. is the mean value of all values of the job matching feature related to the decision.
[0082] S230. Based on the causal path in the target causal directed graph, screen out at least one pair of resume feature groups from the resume features of the target structured data set.
[0083] In this embodiment, the resume feature group refers to two resume features with a causal relationship in the target causal directed graph.
[0084] S240. For each pair of resume feature groups, determine the causal relationship strength between the first resume feature and the second resume feature according to the first eigenvalue set of the first resume feature and the second eigenvalue set of the second resume feature in the resume feature group.
[0085] In this embodiment, the first resume feature and the second resume feature refer to two features with a causal relationship. The first eigenvalue set refers to all possible values of a certain resume feature in the dataset; for example, for the feature of education level, the first eigenvalue set may include different values such as "bachelor's degree" and "master's degree". The second eigenvalue set refers to all possible values of another resume feature in the dataset; for instance, assuming the second resume feature is "number of years of work experience", its second eigenvalue set may include "1 year", "2 years", "3 years", etc. The causal relationship strength represents the strength of the causal relationship between two features and is usually quantified using a numerical value.
[0086] Optionally, calculate the mean of the first eigenvalue set of the first resume feature and the second eigenvalue set of the second resume feature in the resume feature group respectively to obtain the first mean of the first resume feature and the second mean of the second resume feature; subtract the mean from each eigenvalue in the first eigenvalue set, and subtract the mean from each eigenvalue in the second eigenvalue set to obtain the first difference set of eigenvalues and the second difference set of eigenvalues; perform a causal relationship strength calculation on the first difference set of eigenvalues and the second difference set of eigenvalues to obtain the causal relationship strength between the first resume feature and the second resume feature.
[0087] In this embodiment, the feature mean refers to the arithmetic mean of all eigenvalues in an eigenvalue set. For example, if the first eigenvalue set is the number of years of work experience of "1 year, 2 years, 3 years", the first mean is (1 + 2 + 3) / 3 = 2 years. The difference set of eigenvalues refers to the difference between each eigenvalue and the feature mean. For example, assuming the mean of the first eigenvalue set is 2 years, then the difference of "1 year" is (1 - 2) = -1 year, the difference of "2 years" is (2 - 2) = 0 year, and so on; this difference set represents the deviation degree of each sample data.
[0088] Exemplarily, the determination of the causal relationship strength can be achieved through the following formula:
[0089]
[0090] Among them, C(X, Y) refers to the causal relationship strength between resume feature X and resume feature Y. xi i refers to the i-th sample value of resume feature X. yi refers to the i-th sample value of resume feature Y. refers to the mean of all sample values of resume feature X. It refers to the mean of all sample values of resume feature Y. a refers to the time impact coefficient, which is used to measure the degree to which the causal relationship between variables is affected by the time interval Δt(X,Y); for example, when analyzing work experience and work performance, as the time interval between the work experience accumulation time and the performance evaluation time increases, the strength of the causal relationship may weaken.
[0091] S250. Determine the structured data to be matched for at least one resume to be screened and the target resume screening rule for the target position according to the prejudice index set, the causal relationship strength set, the preset resume screening rule, and the target structured data of at least one resume to be screened.
[0092] S260. Screen the structured data to be matched of at least one resume to be screened according to the target resume screening rule to obtain the target resume.
[0093] In the embodiment of the present application, by performing feature screening on the candidate structured data of at least one resume to be screened according to the preset resume screening rule of the target position, a target structured data set is obtained; wherein, the target structured data set includes the target structured data of at least one resume to be screened; the candidate structured data refers to the structured data obtained after preprocessing the resume data of the resume to be screened; according to the target structured data set, a target causal directed graph and a prejudice index set of the applicant features in the target structured data set are determined; wherein, the target causal directed graph is used to represent the causal relationship between each resume feature in the target structured data set; the resume features include applicant features and position matching features; based on the causal path in the target causal directed graph, at least one pair of resume feature groups is screened out from the resume features of the target structured data set; wherein, the resume feature group refers to two resume features with a causal relationship; for each pair of resume feature groups, according to the first feature value set of the first resume feature and the second feature value set of the second resume feature in the resume feature group, the causal relationship strength between the first resume feature and the second resume feature is determined; according to the prejudice index set, the causal relationship strength set, the preset resume screening rule, and the target structured data of at least one resume to be screened, the structured data to be matched for at least one resume to be screened and the target resume screening rule for the target position are determined; according to the target resume screening rule, the structured data to be matched of at least one resume to be screened is screened to obtain the target resume. The above technical solution helps to improve the accuracy and reliability of resume screening by sorting out the causal relationship between each resume feature, combining the feature prejudice index, and correcting the resume data of the user's resume and the screening rule of the target position, so as to accurately identify and eliminate the false correlation in the data and comprehensively eliminate data prejudice.
[0094] Embodiment III
[0095] Figure 3FIG. 0 is a schematic structural diagram of a resume screening device integrating causal reasoning, which is applicable to the situation of uniformly screening resumes after offline tests for all applicants. The resume screening device integrating causal reasoning can be implemented in the form of hardware and / or software, and can be configured in a computer device, such as a server. As Figure 3 shown, the device includes:
[0096] A feature screening module 310, configured to perform feature screening on candidate structured data of at least one resume to be screened according to a preset resume screening rule of a target position, so as to obtain a target structured data set; wherein, the target structured data set includes target structured data of at least one resume to be screened; the candidate structured data refers to structured data obtained by preprocessing the resume data of the resume to be screened;
[0097] A directed graph determination module 320, configured to determine a target causal directed graph and a bias index set of applicant characteristics in the target structured data set according to the target structured data set; wherein, the target causal directed graph is used to represent the causal relationship between resume characteristics in the target structured data set; resume characteristics include applicant characteristics and position matching characteristics;
[0098] A strength set determination module 330, configured to determine a causal relationship strength set between each position characteristic in the target structured data set based on the target causal directed graph according to the target structured data set;
[0099] A data correction module 340, configured to determine the structured data to be matched of at least one resume to be screened and the target resume screening rule of the target position according to the bias index set, the causal relationship strength set, the preset resume screening rule and the target structured data of at least one resume to be screened;
[0100] A resume screening module 350, configured to screen the resume data to be matched of at least one resume to be screened according to the target resume screening rule to obtain a target resume.
[0101] In the embodiment of the present application, by screening features of candidate structured data of at least one resume to be screened according to a preset resume screening rule of a target position, a target structured data set is obtained; wherein, the target structured data set includes target structured data of at least one resume to be screened; candidate structured data refers to structured data obtained after preprocessing the resume data of the resume to be screened; according to the target structured data set, a target causal directed graph and a bias index set of applicant features in the target structured data set are determined; wherein, the target causal directed graph is used to represent the causal relationship between resume features in the target structured data set; resume features include applicant features and position matching features; based on the target causal directed graph, according to the target structured data set, a causal relationship strength set between position features in the target structured data set is determined; according to the bias index set, the causal relationship strength set, the preset resume screening rule and the target structured data of at least one resume to be screened, the structured data to be matched of at least one resume to be screened and the target resume screening rule of the target position are determined; according to the target resume screening rule, the resume data to be matched of at least one resume to be screened is screened to obtain a target resume. The above technical solution, by sorting out the causal relationship between resume features and combining the feature bias index, corrects the resume data of the user resume and the screening rule of the target position, realizes accurate identification and elimination of false correlations in the data, and comprehensively eliminates data bias, which helps to improve the accuracy and reliability of resume screening.
[0102] Optionally, the strength set determination module 330 includes:
[0103] A feature group screening unit, configured to screen out at least one pair of resume feature groups from the resume features of the target structured data set based on the causal path in the target causal directed graph; wherein, a resume feature group refers to two resume features with a causal relationship;
[0104] A relationship strength determination unit, configured to, for each pair of resume feature groups, determine the causal relationship strength between the first resume feature and the second resume feature according to the first feature value set of the first resume feature and the second feature value set of the second resume feature in the resume feature group.
[0105] Optionally, the relationship strength determination unit is specifically configured to:
[0106] Respectively calculate the mean values of the first feature value set of the first resume feature and the second feature value set of the second resume feature in the resume feature group to obtain the first mean value of the first resume feature and the second mean value of the second resume feature;
[0107] Subtract the mean value from each feature value in the first feature value set, and subtract the mean value from each feature value in the second feature value set, to obtain a first feature difference set and a second feature difference set;
[0108] Calculate the causal relationship strength for the first feature difference set and the second feature difference set to obtain the causal relationship strength between the first resume feature and the second resume feature.
[0109] Optionally, the directed graph determination module 320 is specifically configured to:
[0110] Determine at least one causal path according to the causal relationship between resume features in the target structured dataset, and combine the at least one causal path to obtain a candidate causal directed graph;
[0111] Based on the Bayesian principle, perform causal path screening on the candidate causal directed graph to obtain the target causal directed graph.
[0112] Optionally, the data correction module 340 includes:
[0113] The resume data correction unit is configured to correct the feature values and feature weights of at least one applicant feature in the target structured data of at least one resume to be screened according to the bias index set, and obtain the structured data to be matched of at least one resume to be screened;
[0114] The screening rule correction unit is configured to correct the feature weights of at least one resume feature in the preset resume screening rule according to the bias index set and the causal relationship strength set, and obtain the target resume screening rule.
[0115] Optionally, the screening rule correction unit is specifically configured to:
[0116] For each resume feature in the preset resume screening rule, if the bias index of the resume feature in the bias index set meets the weight correction condition, then correct the feature weight of the resume feature according to the causal relationship strength to obtain the corrected feature weight;
[0117] Update the corrected feature weights of at least one resume feature to the preset resume rule to obtain the target resume screening rule.
[0118] The resume screening device integrating causal reasoning provided by the embodiments of the present application can execute the resume screening method integrating causal reasoning provided by any embodiment of the present application, and has the corresponding functional modules and beneficial effects for executing each resume screening method integrating causal reasoning.
[0119] According to the embodiments of the present application, the present application also provides an electronic device, a readable storage medium, and a computer program product.
[0120] Embodiment Four
[0121] Figure 4FIG. 0 is a schematic structural diagram of an electronic device 410 for implementing the resume screening method for integrated causal reasoning according to an embodiment of the present application. The electronic device is intended to represent various forms of digital computers, such as, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, personal digital processors, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present application described and / or claimed herein.
[0122] As Figure 4 shown, the electronic device 410 includes at least one processor 411, and a memory communicatively connected to the at least one processor 411, such as a read-only memory (ROM) 412, a random access memory (RAM) 413, etc. The memory stores a computer program executable by the at least one processor. The processor 411 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 412 or the computer program loaded from the storage unit 418 into the random access memory (RAM) 413. In the RAM 413, various programs and data required for the operation of the electronic device 410 can also be stored. The processor 411, the ROM 412, and the RAM 413 are connected to each other through a bus 414. An input / output (I / O) interface 415 is also connected to the bus 414.
[0123] Multiple components in the electronic device 410 are connected to the I / O interface 415, including: an input unit 416, such as a keyboard, a mouse, etc.; an output unit 417, such as various types of displays, speakers, etc.; a storage unit 418, such as a magnetic disk, an optical disc, etc.; and a communication unit 419, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 419 allows the electronic device 410 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.
[0124] The processor 411 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 411 include but are not limited to a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The processor 411 executes the various methods and processes described above, such as the resume screening method for integrated causal reasoning.
[0125] In some embodiments, the resume screening method integrating causal reasoning can be implemented as a computer program tangibly embodied in a computer-readable storage medium, such as storage unit 418. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 410 via ROM 412 and / or communication unit 419. When the computer program is loaded into RAM 413 and executed by the processor 411, one or more steps of the resume screening method integrating causal reasoning described above can be performed. Alternatively, in other embodiments, the processor 411 can be configured for the resume screening method integrating causal reasoning by any other suitable means (e.g., by means of firmware).
[0126] The various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuitry, integrated circuit systems, field-programmable gate arrays (FPGA), application-specific integrated circuits (ASIC), application-specific standard products (ASSP), systems-on-a-chip (SOC), complex programmable logic devices (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special-purpose or general-purpose programmable processor that receives data and instructions from a storage system, at least one input device, and at least one output device, and transmits the data and instructions to the storage system, the at least one input device, and the at least one output device.
[0127] The computer programs for implementing the methods of this application can be written in any combination of one or more programming languages. These computer programs can be provided to the processor of a general-purpose computer, a special-purpose computer, or other programmable resume screening devices integrating causal reasoning, such that when the computer programs are executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer programs can be executed entirely on the machine, partially on the machine, as a stand-alone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.
[0128] In the context of this application, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer-readable storage medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, the computer-readable storage medium can be a machine-readable signal medium. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0129] For providing interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, speech input, or tactile input).
[0130] The systems and techniques described herein can be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of the communication network include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.
[0131] A computing system may include a client and a server. The client and the server are generally far from each other and usually interact via a communication network. The client-server relationship is created by computer programs running on respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system, solving the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services.
[0132] It should be understood that various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps recited in this application can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of this application can be achieved, and no limitation is made herein.
[0133] The above specific embodiments do not constitute a limitation on the protection scope of this application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application shall be included within the protection scope of this application.
Claims
1. A resume screening method integrating causal reasoning, characterized in that, Including: Performing feature screening on the candidate structured data of at least one resume to be screened according to the preset resume screening rules of the target position, to obtain a target structured data set; wherein, the target structured data set includes the target structured data of the at least one resume to be screened; the candidate structured data refers to the structured data obtained after preprocessing the resume data of the resume to be screened; According to the target structured data set, determining a target causal directed graph and a bias index set of applicant characteristics in the target structured data set; wherein, the target causal directed graph is used to represent the causal relationship between each resume feature in the target structured data set; the resume features include applicant characteristics and position matching characteristics; Based on the target causal directed graph, according to the target structured data set, determining a causal relationship intensity set between each position feature in the target structured data set; According to the bias index set, the causal relationship intensity set, the preset resume screening rules and the target structured data of the at least one resume to be screened, determining the structured data to be matched of the at least one resume to be screened and the target resume screening rules of the target position; According to the target resume screening rules, screening the resume data to be matched of the at least one resume to be screened to obtain a target resume.
2. The method according to claim 1, wherein The determining the causal relationship intensity between each position feature in the target structured data set based on the target causal directed graph and according to the target structured data set includes: Based on the causal paths in the target causal directed graph, screening out at least one pair of resume feature groups from the resume features in the target structured data set; wherein, the resume feature group refers to two resume features with a causal relationship; For each pair of resume feature groups, determining the causal relationship intensity between the first resume feature and the second resume feature according to the first feature value set of the first resume feature and the second feature value set of the second resume feature in the resume feature group.
3. The method according to claim 2, wherein Determining the causal relationship intensity between the first resume feature and the second resume feature according to the first feature value set of the first resume feature and the second feature value set of the second resume feature in the resume feature group includes: Respectively calculating the mean values of the first feature value set of the first resume feature and the second feature value set of the second resume feature in the resume feature group to obtain the first mean value of the first resume feature and the second mean value of the second resume feature; Subtracting each feature value in the first feature value set from the first mean value, and subtracting each feature value in the second feature value set from the second mean value to obtain a first feature difference set and a second feature difference set; Performing causal relationship intensity calculation on the first feature difference set and the second feature difference set to obtain the causal relationship intensity between the first resume feature and the second resume feature.
4. The method according to claim 1, wherein Determining a target causal directed graph according to the target structured data set includes: According to the causal relationship between each resume feature in the target structured data set, determining at least one causal path, and combining the at least one causal path to obtain a candidate causal directed graph; Based on the Bayesian principle, causal path screening is performed on the candidate causal directed graph to obtain the target causal directed graph.
5. The method according to claim 1, wherein The determination of the to-be-matched structured data of the at least one resume to be screened and the target resume screening rule of the target position according to the bias index set, the causal relationship strength set, the preset resume screening rule, and the target structured data of the at least one resume to be screened includes: According to the bias index set, the eigenvalue and eigenweight of at least one applicant feature in the target structured data of the at least one resume to be screened are corrected to obtain the to-be-matched structured data of the at least one resume to be screened; According to the bias index set and the causal relationship strength set, the eigenweight of at least one resume feature in the preset resume screening rule is corrected to obtain the target resume screening rule.
6. The method according to claim 5, wherein The correction of the eigenweight of at least one resume feature in the preset resume screening rule according to the bias index set and the causal relationship strength to obtain the target resume screening rule includes: For each resume feature in the preset resume screening rule, if the bias index of the resume feature in the bias index set meets the weight correction condition, the eigenweight of the resume feature is corrected according to the causal relationship strength to obtain the corrected eigenweight; The corrected eigenweights of the at least one resume feature are updated to the preset resume rule to obtain the target resume screening rule.
7. A resume screening device integrating causal reasoning, characterized in that, It includes: A feature screening module, configured to perform feature screening on the candidate structured data of at least one resume to be screened according to the preset resume screening rule of the target position to obtain a target structured data set; wherein, the target structured data set includes the target structured data of the at least one resume to be screened; the candidate structured data refers to the structured data obtained after preprocessing the resume data of the resume to be screened; A directed graph determination module, configured to determine a target causal directed graph and the bias index set of applicant features in the target structured data set according to the target structured data set; wherein, the target causal directed graph is used to represent the causal relationship between each resume feature in the target structured data set; the resume features include applicant features and position matching features; A strength set determination module, configured to determine the causal relationship strength set between each position feature in the target structured data set based on the target causal directed graph according to the target structured data set; A data correction module, configured to determine the to-be-matched structured data of the at least one resume to be screened and the target resume screening rule of the target position according to the bias index set, the causal relationship strength set, the preset resume screening rule, and the target structured data of the at least one resume to be screened; A resume screening module, configured to screen the to-be-matched resume data of the at least one resume to be screened according to the target resume screening rule to obtain the target resume.
8. An electronic device, characterized in that, It includes: One or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the resume screening method for integrated causal reasoning according to any one of claims 1-6.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the resume screening method for integrated causal reasoning according to any one of claims 1-6.
10. A computer program product, comprising a computer program which, when executed by a processor, implements the resume screening method for integrated causal reasoning according to any one of claims 1-6.