Post screening intelligent management system based on AI big data

Abstract

The invention relates to the technical field of data processing, in particular to an intelligent post screening management system based on AI big data, which is provided with a history determination module, an event analysis module, a preliminary analysis module and a retrieval module. Extracting a key event according to an event response standard, and recording event information corresponding to the key event; according to the condition association probability of the keywords corresponding to the resume information and the job application information in the event information, determining the association relationship of the keywords, and constructing an association relationship database; analyzing the resume information uploaded by the user side, identifying the distribution characteristics of each associated keyword, calculating an associated dense representation parameter according to the distribution characteristics, and setting a distribution label for the resume information; according to the method, the job application information is adaptively retrieved according to the distribution labels, so that on the premise that the data reliability is guaranteed, the screening and matching data volume is reduced, and the screening and matching efficiency and accuracy are improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an intelligent management system for job screening based on AI big data. Background Art

[0002] There are a large number of job seekers in the labor market, especially in popular industries and fields, where competition for positions is fierce. Job seekers hope to find positions that match their abilities and career plans more efficiently. Therefore, an intelligent and comprehensive screening system is needed to assist decision-making.

[0003] Chinese patent application publication number: CN118798836A, discloses a precise matching recruitment system based on big data analysis, including: loading description documents of all recruitment positions of an enterprise, using the TF-IDF method to filter out common word sets and unique word sets of each recruitment position from the description documents; filtering out a target position set from all recruitment positions based on the common word sets; retrieving the job seeker's resume to be matched and the keyword set in the resume to be matched, and extracting the common word set and unique word set of the target position, inputting the keyword set, the common word set and the unique word set into a convolutional neural network pre-trained for judging whether the person-job match is consistent, and obtaining a judgment result; summarizing all the positions that are compatible with the job seeker from the target position set according to the judgment result; the present invention is conducive to avoiding repeated resume screening and interviews, and improving the accuracy and efficiency of enterprise recruitment.

[0004] However, there are still the following problems in the prior art:

[0005] It is easy to overlook the different contributions of the corresponding keywords in resume information and job application information to reaching an interview agreement, which makes the screening and matching of job application information unclear, the screening and matching goals are vague, and reduces the accuracy of screening and matching. Summary of the invention

[0006] To this end, the present invention provides an AI big data-based job screening intelligent management system to overcome the problem in the prior art that it is easy to ignore the different contributions of the corresponding keywords of resume information and job search information to reaching an interview agreement, resulting in unclear screening and matching of job search information, vague screening and matching targets, and reduced accuracy of screening and matching.

[0007] To achieve the above objectives, the present invention provides an AI big data-based job screening intelligent management system, which includes:

[0008] A history determination module is used to obtain resume information uploaded by each user and job application information uploaded by the enterprise, extract key events according to event response standards, and record event information corresponding to the key events;

[0009] An event analysis module, which is connected to the history determination module, is used to call the event information of the key event, determine the association relationship of each keyword according to the conditional association probability of the keywords corresponding to the resume information and the job search information in the event information, and construct an association relationship database;

[0010] A preliminary analysis module, which is connected to the event analysis module and responds to the resume information uploaded by the user end to parse the resume information, identify the distribution characteristics of each related keyword, calculate the related intensive representation parameter according to the distribution characteristics, and set a distribution label for the resume information;

[0011] A retrieval module, which is connected to the preliminary analysis module and includes a deep matching unit and a conventional matching unit;

[0012] The deep matching unit is used to remove keywords from the resume information to obtain reserved keywords, determine a push sequence for job search information based on a conditional association probability between the reserved keywords and keywords corresponding to each job search information, match job search information based on push sequence segments corresponding to each keyword, and determine a push priority for each job search information based on the sequence segments to push job search information;

[0013] The conventional matching unit is used to determine the semantic relevance between the resume information and the keywords corresponding to the job search information, and generate a job search information push list according to the semantic relevance to push the job search information;

[0014] The distribution characteristics include the number of associated keywords and the mean value of the conditional association probability corresponding to each of the associated keywords.

[0015] Furthermore, the history determination module is used to extract key events according to the event response standard, including:

[0016] To identify the events that meet the event response criteria as the critical events;

[0017] Among them, the event response standard includes that the user end and the enterprise end reach an interview agreement; the event includes interactive operations between the user end and the enterprise end.

[0018] Furthermore, the event analysis module is used to determine the association relationship of each keyword including:

[0019] If the conditional association probability of the keywords corresponding to the resume information and the job application information is greater than or equal to the conditional association probability threshold, it is determined that there is an association relationship between the keywords;

[0020] The conditional association probability is the probability that the keyword appears in both the resume information and the job search information.

[0021] Furthermore, the preliminary analysis module is used to calculate the associated dense characterization parameters according to the distribution characteristics, including:

[0022] The ratio of the number of related keywords to the number threshold is used as the first related dense feature;

[0023] The ratio of the mean value of the conditional association probability corresponding to each of the associated keywords to the mean value of the conditional association probability threshold is used as a second association intensive feature;

[0024] Using the first association intensive feature and the second association intensive feature to perform a weighted sum as the association intensive representation parameter;

[0025] The associated keywords are keywords with an associated relationship.

[0026] Furthermore, the preliminary analysis module is used to set distribution labels for the resume information, including:

[0027] If the associated dense representation parameter is greater than or equal to the associated dense representation parameter threshold, setting the label of the resume information to dense distribution;

[0028] If the association intensive representation parameter is less than the association intensive representation parameter threshold, the label of the resume information is set to dispersed distribution.

[0029] Furthermore, the retrieval module is used to call the deep matching unit or the conventional matching unit according to the distribution tag, including:

[0030] If the labels set in the resume information are densely distributed, the deep matching unit is called;

[0031] If the label set for the resume information is dispersed distribution, the regular matching unit is called.

[0032] Furthermore, the deep matching unit is used to remove keywords including:

[0033] Used to identify keywords in resume information;

[0034] It is used to eliminate keywords that have no correlation among the keywords.

[0035] Furthermore, the deep matching unit is used to determine the push sequence for job search information based on the conditional association probability between the reserved keywords and the keywords corresponding to each job search information, including:

[0036] to respectively determine the conditional association probabilities between the reserved keywords and the keywords corresponding to each job search information;

[0037] The job search information is arranged in descending order according to the condition association probability to obtain the push sequence.

[0038] Furthermore, the deep matching module is used to match job search information with each keyword based on the push sequence segment, and the push priority of each job search information based on the sequence segment includes:

[0039] Used to divide the push sequence into a plurality of push sequence segments;

[0040] To match each keyword in the sequence segment with each keyword in the job search information, and determine the job search information that meets the matching conditions;

[0041] To give priority to pushing job search information corresponding to the sequence segments with the highest ranking;

[0042] The matching condition is that the semantic relevance between each keyword and the keyword in the job-seeking information is greater than a predetermined semantic relevance threshold.

[0043] Furthermore, the conventional matching unit is used to generate a job search information push list according to the semantic relevance, including:

[0044] If the semantic relevance is greater than or equal to the semantic relevance threshold, the corresponding job search information is placed in the job search information push list.

[0045] Compared with the prior art, the present invention is provided with a history determination module, an event analysis module, a preliminary analysis module and a retrieval module. By acquiring the resume information uploaded by each user end and the job-seeking information uploaded by the enterprise end, key events are extracted according to event response standards, and event information corresponding to the key events is recorded; the association relationship of each keyword is determined according to the conditional association probability of the keywords corresponding to the resume information and the job-seeking information in the event information, and an association relationship database is constructed; the resume information uploaded by the user end is parsed, the distribution characteristics of each associated keyword are identified, the association intensive representation parameters are calculated according to the distribution characteristics, and distribution labels are set for the resume information; the job-seeking information is adaptively retrieved according to the distribution labels, and thus the present invention reduces the amount of data for screening and matching while ensuring data reliability, and improves the efficiency and accuracy of screening and matching.

[0046] In particular, the present invention sets up an event analysis module, which determines the correlation between resume information and job application information corresponding to the interview agreement by analyzing pre-recorded key events. In actual situations, job seekers and employers will upload the corresponding resume information and job application information to the recruitment platform, and interact on the platform to facilitate the interview agreement. Based on this, the present invention analyzes a large number of events before the interview agreement is reached to identify the similarities in such events. For example, the correlation between the resume information and the job application information is analyzed by the degree of matching of the corresponding keywords of the resume information and the job application information. The correlation relationship is used to represent the high contribution of the corresponding keywords contained in the resume information to the interview agreement reached by both parties, and at the same time reflects that the keyword has high availability and strong data representation when matching the corresponding position.

[0047] In particular, the present invention sets up a preliminary analysis module, which preliminarily analyzes the distribution of keywords associated with the job search information in the resume information, based on the basic conditions between the pre-analyzed resume information and the corresponding keywords of the job search information that can facilitate the job seeker and the employer to reach an interview agreement. That is, the probability of occurrence of such keywords in the resume information is analyzed. In actual situations, the proportion and accuracy of the keywords contained in the resume information will affect the accuracy of the screening and matching of the job search information. Therefore, the present invention calculates the association density characterization parameter by associating the keywords and the mean of the conditional association probabilities of each associated keyword and the corresponding keywords of the job search information, so as to characterize the distribution density of the keywords corresponding to the resume information uploaded by the user and the corresponding job search information, and then characterizes the matching degree of the resume information and the job search information. Under the premise of ensuring data reliability, the present invention reduces the amount of data for screening and matching, and improves the efficiency and accuracy of screening and matching.

[0048] In particular, the present invention sets up a deep matching unit. Since the resume information in which the labels are set is densely distributed and contains more keywords that facilitate the two parties to reach an interview agreement, the present invention takes these keywords as the key consideration basis, pre-eliminates keywords that have no correlation with the corresponding job search information, and reduces the amount of data for screening and matching and related interference words that affect the screening and matching while ensuring the accuracy of screening. A deep analysis is performed based on the retained keywords after the elimination operation combined with the conditional association probability corresponding to the corresponding retained keywords to determine the push sequence of the job search information, determine the push priority, and push the job search information in the obtained sequence segment, thereby improving the accuracy of screening and matching. Therefore, the present invention reduces the amount of data for screening and matching while ensuring data reliability, and improves the efficiency and accuracy of screening and matching. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is a functional module diagram of an AI big data job screening intelligent management system based on an embodiment of the invention;

[0050] Figure 2 A logical decision diagram for setting distribution labels for resume information in an embodiment of the invention;

[0051] Figure 3 A logic decision diagram for calling a deep matching unit or a conventional matching unit according to a distribution tag according to an embodiment of the invention;

[0052] Figure 4 This is a logical decision diagram for generating a job search information push list according to the semantic association according to an embodiment of the invention. DETAILED DESCRIPTION

[0053] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0054] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0055] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the term "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0056] See also Figure 1 As shown, Figure 1 This is a functional module diagram of an AI big data job screening intelligent management system according to an embodiment of the present invention. The AI ​​big data job screening intelligent management system according to an embodiment of the present invention includes:

[0057] A history determination module is used to obtain resume information uploaded by each user and job application information uploaded by the enterprise, extract key events according to event response standards, and record event information corresponding to the key events;

[0058] An event analysis module, which is connected to the history determination module, is used to call the event information of the key event, determine the association relationship of each keyword according to the conditional association probability of the keywords corresponding to the resume information and the job search information in the event information, and construct an association relationship database;

[0059] A preliminary analysis module, which is connected to the event analysis module and responds to the resume information uploaded by the user end to parse the resume information, identify the distribution characteristics of each related keyword, calculate the related intensive representation parameter according to the distribution characteristics, and set a distribution label for the resume information;

[0060] A retrieval module, which is connected to the preliminary analysis module and includes a deep matching unit and a conventional matching unit;

[0061] The deep matching unit is used to remove keywords from the resume information to obtain reserved keywords, determine a push sequence for job search information based on a conditional association probability between the reserved keywords and keywords corresponding to each job search information, match job search information based on push sequence segments corresponding to each keyword, and determine a push priority for each job search information based on the sequence segments to push job search information;

[0062] The conventional matching unit is used to determine the semantic relevance between the resume information and the keywords corresponding to the job search information, and generate a job search information push list according to the semantic relevance to push the job search information;

[0063] The distribution characteristics include the number of associated keywords and the mean value of the conditional association probability corresponding to each of the associated keywords.

[0064] Specifically, the user end refers to job seekers, and the enterprise end refers to employers. Both parties can facilitate the conclusion of an interview agreement through relevant interactions on the recruitment platform. At the same time, based on the relevant rules of the recruitment platform, the resume information uploaded by job seekers on the recruitment platform and the job search information uploaded by employers will be stored in the background of the recruitment platform. Therefore, the present invention gradually analyzes the information recorded in the background and adaptively pushes the job search information to the user end.

[0065] Specifically, before obtaining resume information and job application information, you need to obtain authorization from the corresponding terminal, which will not be repeated here.

[0066] Specifically, there is no limitation on the specific structures of the history determination module, event analysis module, preliminary analysis module and retrieval module, and they themselves or each unit therein can be composed of logical components or a combination of logical components, and the logical components include field programmable processors, computers or microprocessors in computers.

[0067] Specifically, the history determination module is used to extract key events according to the event response standard, including:

[0068] To identify the events that meet the event response criteria as the critical events;

[0069] Among them, the event response standard includes that the user end and the enterprise end reach an interview agreement; the event includes interactive operations between the user end and the enterprise end.

[0070] It can be understood that the interactive operation refers to the relevant operations performed by the user side and the enterprise side to facilitate the interview agreement, including the interview invitation pop-up window sent by the enterprise side to the user side and the user side clicking the button in response to the interview invitation pop-up window and the interaction between the user side and the enterprise side in the interactive interface, etc., which will not be elaborated here.

[0071] Specifically, the event analysis module is used to determine the association relationship of each keyword, including:

[0072] If the conditional association probability of the keywords corresponding to the resume information and the job application information is greater than or equal to the conditional association probability threshold, it is determined that there is an association relationship between the keywords;

[0073] The conditional association probability is the probability that the keyword appears in both the resume information and the job search information.

[0074] It is understandable that the degree of matching between the resume information obtained to reach an interview agreement and the keywords corresponding to the job application information can reflect the degree to which both parties are inclined to reach an interview agreement. Therefore, in this embodiment, keywords with an associated relationship are determined as associated keywords, which will not be repeated here.

[0075] In implementation, the conditional association probability is determined by calculating the co-occurrence frequency of keywords corresponding to the resume information and the job application information. The co-occurrence frequency can be determined by constructing a bag-of-words model. Of course, other methods can also be used, which will not be repeated here.

[0076] Specifically, event information refers to the resume information uploaded by the corresponding user side and the job search information uploaded by the enterprise side after the key event is achieved.

[0077] Specifically, when calculating the conditional association probability, the recorded time information can be stored in a database, and the probability of the keywords appearing in both the resume information and the job search information can be calculated based on the information in the database to obtain the conditional association probability, which will not be repeated here.

[0078] In implementation, the purpose of setting the conditional association probability threshold of the keywords corresponding to the resume information and the job application information is to characterize the situation where the co-occurrence frequency and matching degree of the keywords in the resume information and the job application information are high;

[0079] Among them, the conditional association probability threshold is determined based on the mean value of the conditional association probabilities of the keywords corresponding to the resume information and the job application information. By obtaining the event information corresponding to several key events recorded in the background for reaching an interview agreement, the conditional association probabilities of the keywords corresponding to the resume information and the job application information are called, and the mean value of the conditional association probability is solved. For the purpose of setting the conditional association probability threshold, the conditional association probability threshold is determined between 1.15 times and 1.35 times the mean value of the conditional association probability.

[0080] Specifically, the present invention sets up an event analysis module, which determines the correlation between resume information and job application information corresponding to the interview agreement by analyzing pre-recorded key events. In actual situations, job seekers and employers will upload the corresponding resume information and job application information to the recruitment platform, and interact on the platform to facilitate the interview agreement. Based on this, the present invention analyzes a large number of events before the interview agreement is reached to identify the similarities in such events. For example, the correlation between the resume information and the job application information is analyzed by the degree of matching of the corresponding keywords. The correlation relationship is used to represent the high contribution of the corresponding keywords contained in the resume information to the interview agreement between the two parties, and at the same time reflects that the keyword has high availability and strong data representation when matching the corresponding position.

[0081] Specifically, the preliminary analysis module is used to calculate the associated dense representation parameters according to the distribution characteristics, including:

[0082] The ratio of the number of related keywords to the number threshold is used as the first related dense feature;

[0083] The ratio of the mean value of the conditional association probability corresponding to each of the associated keywords to the mean value of the conditional association probability threshold is used as a second association intensive feature;

[0084] Using the first association intensive feature and the second association intensive feature to perform a weighted sum as the association intensive representation parameter;

[0085] The associated keywords are keywords with an associated relationship.

[0086] In this implementation, when performing weighted summation, the weight of the first associated dense feature is set to 0.45, and the weight of the second associated dense feature is set to 0.55;

[0087] Regarding the determination of the number threshold of associated keywords and the mean threshold of the conditional association probability between each associated keyword and the keywords corresponding to the job search information, during implementation, the event information corresponding to several key events for reaching interview agreement recorded in the background is obtained, the number data of associated keywords and the mean data of the conditional association probability between each associated keyword and the keywords corresponding to the job search information are called, and the average value of the number of associated keywords and the average value of the conditional association probability mean are solved. Since the purpose of setting the number threshold of associated keywords and the mean threshold of the conditional association probability between each associated keyword and the keywords corresponding to the job search information is to represent the high degree of match between the keywords in the resume information and the job search information and the dense distribution of the associated keywords, the number threshold of associated keywords is determined between 1.1 times and 1.02 times the average value of the number of associated keywords, and the conditional association probability mean threshold is determined between 1.15 times and 1.2 times the average value of the conditional association probability mean.

[0088] Specifically, see Figure 2 As shown, it is a logical decision diagram for setting distribution labels for resume information in an embodiment of the present invention. The preliminary analysis module is used to set distribution labels for the resume information, including:

[0089] If the associated dense representation parameter is greater than or equal to the associated dense representation parameter threshold, setting the label of the resume information to dense distribution;

[0090] If the association intensive representation parameter is less than the association intensive representation parameter threshold, the label of the resume information is set to dispersed distribution.

[0091] The threshold of the association dense representation parameter is selected in the interval [1.25, 1.34].

[0092] Specifically, the present invention sets up a preliminary analysis module, which preliminarily analyzes the distribution of keywords associated with the job search information in the resume information, based on the basic conditions between the pre-analyzed resume information and the corresponding keywords of the job search information that can facilitate the job seeker and the employer to reach an interview agreement. That is, the probability of occurrence of such keywords in the resume information is analyzed. In actual situations, the proportion and accuracy of the keywords contained in the resume information will affect the accuracy of the screening and matching of the job search information. Therefore, the present invention calculates the association density characterization parameter by associating the keywords and the mean of the conditional association probabilities of each associated keyword and the corresponding keywords of the job search information, so as to characterize the distribution density of the keywords corresponding to the resume information uploaded by the user and the corresponding job search information, and then characterizes the matching degree of the resume information and the job search information. Under the premise of ensuring data reliability, the present invention reduces the amount of data for screening and matching, and improves the efficiency and accuracy of screening and matching.

[0093] Specifically, see Figure 3 As shown, it is a logical decision diagram for calling a deep matching unit or a conventional matching unit according to a distribution tag in an embodiment of the present invention. The retrieval module is used to call a deep matching unit or a conventional matching unit according to the distribution tag, and includes:

[0094] If the labels set in the resume information are densely distributed, the deep matching unit is called;

[0095] If the label set for the resume information is dispersed distribution, the regular matching unit is called.

[0096] Specifically, the deep matching unit is used to remove keywords including:

[0097] Used to identify keywords in resume information;

[0098] It is used to eliminate keywords that have no correlation among the keywords.

[0099] If the conditional association probability of the keywords corresponding to the resume information and the job application information is less than the conditional association probability threshold, it is determined that there is no association relationship between the keywords.

[0100] Specifically, the deep matching unit is used to determine the push sequence for job search information based on the conditional association probability between the reserved keywords and the keywords corresponding to each job search information, including:

[0101] to respectively determine the conditional association probabilities between the reserved keywords and the keywords corresponding to each job search information;

[0102] The job search information is arranged in descending order according to the condition association probability to obtain the push sequence.

[0103] It can be understood that the reserved keywords are all keywords with associated relationships, and the conditional association probability is used when determining the corresponding keywords, so the conditional association probability of the reserved keywords can be determined.

[0104] Specifically, the deep matching module is used to match job search information with each keyword based on the push sequence segment, and the push priority of each job search information based on the sequence segment includes:

[0105] Used to divide the push sequence into a plurality of push sequence segments;

[0106] To match each keyword in the sequence segment with each keyword in the job search information, and determine the job search information that meets the matching conditions;

[0107] It is used to give priority to pushing job search information corresponding to the sequence segments with higher rankings. It can be understood that the larger the mean value of the conditional association probability corresponding to the keywords in the sequence segment is, the higher the corresponding ranking is.

[0108] The matching condition is that the semantic relevance between each keyword and the keyword in the job-seeking information is greater than a predetermined semantic relevance threshold.

[0109] Specifically, there is no limitation on the division method of the push sequence segments, which can be determined based on the ratio of the total number of keywords, and the keywords corresponding to each sequence segment have the same proportion.

[0110] Specifically, there is no limitation on the method of determining the semantic association between keywords corresponding to resume information and job application information. In some possible implementations, a word vector model, such as Word2Vec or FastText, is used to map related words into a vector space, and cosine similarity is used to represent the semantic association.

[0111] The semantic relevance threshold is predetermined, and pre-recorded key events are analyzed to determine event information, and the mean semantic relevance between keywords in job search information and resume information is solved, and the semantic relevance threshold is set to between 0.75 and 0.85 times the mean semantic relevance.

[0112] Specifically, the present invention sets up a deep matching unit. Since the resume information in which the set labels are densely distributed contains more keywords that facilitate the two parties to reach an interview agreement, the present invention takes these keywords as the key consideration basis, pre-eliminates keywords that have no correlation with the corresponding job search information, and reduces the amount of data for screening and matching and related interference words that affect the screening and matching while ensuring the accuracy of screening. A deep analysis is performed based on the retained keywords after the elimination operation and the conditional association probability corresponding to the corresponding retained keywords to determine the push sequence of the job search information, determine the push priority, and push the job search information in the obtained sequence segment, thereby improving the accuracy of screening and matching. Therefore, the present invention reduces the amount of data for screening and matching while ensuring data reliability, and improves the efficiency and accuracy of screening and matching.

[0113] Specifically, see Figure 4 As shown, it is a logical decision diagram for generating a job search information push list according to the semantic association according to an embodiment of the present invention. The conventional matching unit is used to generate a job search information push list according to the semantic association, including:

[0114] If the semantic relevance is greater than or equal to the semantic relevance threshold, the corresponding job search information is placed in the job search information push list.

[0115] It is understandable that after the job search information is placed in the push list, it is pushed according to the list order, which will not be repeated here.

[0116] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. An intelligent management system for job screening based on AI big data, characterized in that: include: A history determination module is used to obtain resume information uploaded by each user and job application information uploaded by the enterprise, extract key events according to event response standards, and record event information corresponding to the key events; An event analysis module, which is connected to the history determination module, is used to call the event information of the key event, determine the association relationship of each keyword according to the conditional association probability of the keywords corresponding to the resume information and the job search information in the event information, and construct an association relationship database; A preliminary analysis module, which is connected to the event analysis module and responds to the resume information uploaded by the user end to parse the resume information, identify the distribution characteristics of each related keyword, calculate the related intensive representation parameter according to the distribution characteristics, and set a distribution label for the resume information; A retrieval module, which is connected to the preliminary analysis module and includes a deep matching unit and a conventional matching unit; The deep matching unit is used to remove keywords from the resume information to obtain reserved keywords, determine a push sequence for job search information based on a conditional association probability between the reserved keywords and keywords corresponding to each job search information, match job search information based on push sequence segments corresponding to each keyword, and determine a push priority for each job search information based on the sequence segments to push job search information; The conventional matching unit is used to determine the semantic relevance between the resume information and the keywords corresponding to the job search information, and generate a job search information push list according to the semantic relevance to push the job search information; The distribution characteristics include the number of associated keywords and the mean value of the conditional association probability corresponding to each of the associated keywords.

2. The AI ​​big data-based job screening intelligent management system according to claim 1 is characterized in that: The history determination module is used to extract key events according to the event response standard, including: To identify the events that meet the event response criteria as the critical events; Among them, the event response standard includes that the user end and the enterprise end reach an interview agreement; the event includes interactive operations between the user end and the enterprise end.

3. The AI ​​big data-based job screening intelligent management system according to claim 1 is characterized in that: The event analysis module is used to determine the association relationship of each keyword, including: If the conditional association probability of the keywords corresponding to the resume information and the job application information is greater than or equal to the conditional association probability threshold, it is determined that there is an association relationship between the keywords; The conditional association probability is the probability that the keyword appears in both the resume information and the job search information.

4. The AI ​​big data-based job screening intelligent management system according to claim 1 is characterized in that: The preliminary analysis module is used to calculate the associated dense characterization parameters according to the distribution characteristics, including: The ratio of the number of related keywords to the number threshold is used as the first related dense feature; The ratio of the mean value of the conditional association probability corresponding to each of the associated keywords to the mean value of the conditional association probability threshold is used as a second association intensive feature; Using the first association intensive feature and the second association intensive feature to perform a weighted sum as the association intensive representation parameter; The associated keywords are keywords with an associated relationship.

5. The AI ​​big data-based job screening intelligent management system according to claim 1 is characterized in that: The preliminary analysis module is used to set distribution labels for the resume information, including: If the associated dense representation parameter is greater than or equal to the associated dense representation parameter threshold, setting the label of the resume information to dense distribution; If the association intensive representation parameter is less than the association intensive representation parameter threshold, the label of the resume information is set to dispersed distribution.

6. The AI ​​big data-based job screening intelligent management system according to claim 1 is characterized in that: The retrieval module is used to call the deep matching unit or the conventional matching unit according to the distribution tag, including: If the labels set in the resume information are densely distributed, the deep matching unit is called; If the label set for the resume information is dispersed distribution, the regular matching unit is called.

7. The AI ​​big data-based job screening intelligent management system according to claim 1 is characterized in that: The deep matching unit is used to remove keywords including: Used to identify keywords in resume information; It is used to eliminate keywords that have no correlation among the keywords.

8. The AI ​​big data-based job screening intelligent management system according to claim 1 is characterized in that: The deep matching unit is used to determine the push sequence for job search information based on the conditional association probability between the reserved keywords and the keywords corresponding to each job search information, including: to respectively determine the conditional association probabilities between the reserved keywords and the keywords corresponding to each job search information; The job search information is arranged in descending order according to the condition association probability to obtain the push sequence.

9. The AI ​​big data-based job screening intelligent management system according to claim 8 is characterized in that: The deep matching module is used to match job search information with each keyword based on the push sequence segment, and the push priority of each job search information is determined based on the sequence segment, including: Used to divide the push sequence into a plurality of push sequence segments; To match each keyword in the sequence segment with each keyword in the job search information to determine the job search information that meets the matching conditions; To give priority to pushing job search information corresponding to the sequence segments with the highest ranking; The matching condition is that the semantic relevance between each keyword and the keyword in the job-seeking information is greater than a predetermined semantic relevance threshold.

10. The AI ​​big data-based job screening intelligent management system according to claim 1 is characterized in that: The conventional matching unit is used to generate a job search information push list according to the semantic relevance, including: If the semantic relevance is greater than or equal to the semantic relevance threshold, the corresponding job search information is placed in the job search information push list.

Images