KR20180041478A - Human resouce matching system based on wallet system, and method thereof - Google Patents

Abstract

According to an aspect of the present invention, a human resource matching server comprises: a job searching module which classifies information of a job seeker into quantitative information able to be digitized and qualitative information unable to be digitized, determines a job appropriate for the job seeker by a first machine learning method by using the quantitative information, extracts characteristic words by a data mining method by using the qualitative information, calculates a degree of suitability for each job of the job seeker by a second machine learning method by using the extracted words, and calculates the final degree of suitability for each job by using the job, determined by using the quantitative information, and the degree of suitability for each job by using the qualitative information; and a job matching module which determines a recommended job seeker and a recommended job by using job information for each job of a company and the final degree of suitability for each job of the job seeker.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a wallet system based on a human resource matching system,

The present invention relates to a talent matching system, and more particularly, to a talent matching system that quantifies job applicants' fitness by job type using information of a job seeker entered in a wallet system and recommends a talent suitable for a company based on the fitness.

In the existing recruitment system, if a job offerer advertises a job through various media such as internet newspapers or makes a referral to a job placement agency, the job seeker submits necessary documents such as an application form and a resume to the relevant company, It was a way of reconsidering and recruiting necessary personnel.

In this way, job seekers have to repeat the process of applying for jobs to companies that want to work, submitting necessary documents, interviewing them, and so on until the job search is completed, wasting time.

In addition, companies are required to judge job seekers based on abstract data such as job applicants' report cards, English scores, and self-introduction statements, so it is difficult to find talents that meet the needs of the company. Also, It is necessary to carry out an operation of sorting and evaluating the time required for the operation.

The development of the Internet has made it easier for employers to submit job applications and submitting documents online, but in the case of companies, it is still necessary to classify volunteers by hand by the staff, and procedures and costs are incurred to verify the reliability of submitted documents There is still a problem.

SUMMARY OF THE INVENTION The present invention has been made in view of the technical background as described above, and provides a smartphone matching system based on a Wallet system that provides an artificial intelligent talent matching platform based on data mining and text mining technology through big data secured by a university and a machine learning module The purpose of that is to do.

The objects of the present invention are not limited to the above-mentioned objects, and other objects not mentioned can be clearly understood by those skilled in the art from the following description.

According to an aspect of the present invention, there is provided a human resource matching server for classifying a job seeker's information into quantitative information that can be quantified and qualitative information that can not be quantified, A method for determining a job suitable for the job seeker, extracting feature words by data mining using the qualitative information, calculating fitness for each job by the second machine learning method using the extracted words, A job search module for calculating the fitness according to the final job using the job determined using the quantitative information and the fitness calculated according to the job using the qualitative information; And a job matching module for determining a recommendation job seeker and a recommendation job by using the candidate job information for each job of the company and the fitness degree of each job candidate according to the final job.

According to another aspect of the present invention, there is provided a human resource matching method comprising: receiving information of a job seeker; Classifying the information of the applicant into quantitative information that can be digitized and qualitative information that can not be digitized; Determining a job suitable for the job seeker by a first machine learning method using the quantitative information; Extracting feature words by data mining using the qualitative information; Calculating a goodness-of-fit of the job seeker by a second machine learning method using extracted words; Calculating fitness for each job by using the job determined using the quantitative information and the fitness calculated for each job using the qualitative information; And determining a recommendation job seeker and a recommendation job by using the job offer information of each job by the company and the fitness degree of each job candidate by the final job.

According to the present invention, human resource matching and recommendation algorithms are provided using quantitative and qualitative information of job seekers to secure manpower and manpower management by a systematic system, thereby reducing the management cost of the enterprise and providing a satisfactory job for the job seeker There is an effect.

1 is a structural view of a human resource matching system according to an embodiment of the present invention;
2 is a structural diagram of a talent matching server according to an embodiment of the present invention;
FIG. 3 is an exemplary diagram for determining a job of a job seeker by a decision tree according to an embodiment of the present invention; FIG.
FIG. 4 is a diagram illustrating a result of calculating job fitness according to qualitative information of a job seeker according to an embodiment of the present invention; FIG.
FIG. 5 is an exemplary diagram illustrating a result of a job matching method according to an embodiment of the present invention; FIG.
6 is a flowchart of a talent matching method according to another embodiment of the present invention.
Figure 7 is an illustration of a computer system in which a job matching method according to one embodiment of the present invention is performed.

BRIEF DESCRIPTION OF THE DRAWINGS The advantages and features of the present invention, and the manner of achieving them, will be apparent from and elucidated with reference to the embodiments described hereinafter in conjunction with the accompanying drawings. The present invention may, however, be embodied in many different forms and should not be construed as being limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Is provided to fully convey the scope of the invention to those skilled in the art, and the invention is only defined by the scope of the claims. It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. In the present specification, the singular form includes plural forms unless otherwise specified in the specification. As used herein, the terms " comprises, " and / or "comprising" refer to the presence or absence of one or more other components, steps, operations, and / Or additions.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 shows an overall structure of a talent matching system based on a wallet system according to an embodiment of the present invention.

The Wallet system-based talent matching system 10 includes a wallet system 100, a corporate mailbox 200, and a talent matching server 300.

The talent matching system 10 predicts the job suitability and job performance of job seekers for talent matching and uses a decision tree method among the data mining techniques as a method of categorizing data into categories for this purpose. If multinomial logistic analysis such as type group and gender type is required, Multinomial Logistic Regression method can be used to supplement it.

When the output data of the neural network circuit based on the input data is different from the actual result, it is fed back to obtain the individual input data The neural network circuit weights are updated to obtain optimal job suitability and job performance predictions.

Hereinafter, each component of the talent matching system 10 will be described in detail.

The Wallet system 100 is a system for managing information of job seekers utilizing university-specific big data.

The WALLET SYSTEM 100 is a personalized database that stores information on school attendance, graduation, and credit, etc., verified by the school server for each school, and personal information such as an individual's self-introduction letter or career certificate, DB.

Since the information stored in the DB is already verified by the school server, additional verification is unnecessary and thus the reliability of the DB information can be secured. In addition, the information stored in the database can be prevented from being modified by the user, so that the accuracy of the information can be maintained.

The personal DB stores information for each individual, unlike the DB. Since it is personal information, it is possible to modify information such as addition or editing by the user.

The information stored in the personal DB is stored with overseas training information, foreign language score, career information, etc. verified by the school server, and self-introduction sheets, recommendation letters, and school recommendation letters written by users are also stored.

Even if the information stored in the personal DB is not verified by the user, the information of the overseas training information, the foreign language score, the career information, etc., verified by the school server can be prevented from being corrected.

The Wallet system 100 thus manages the information of the applicant and provides this information to the talent matching server 300 as needed.

The company mailbox 200 stores information such as job information and job conditions of a corporation and information of a job seeker.

The company mailbox 200 stores job information and job conditions set by the manager or the recruitment manager of the company and provides the job information to the talent matching server 300 so that the talent matching server 300 can use the matching information to match the qualified job seeker .

Among the information in the company mailbox 200, the information on the job hunting and the like can be accessed by a job seeker so as to retrieve information on the hiring, entitlement, etc. However, the information of the hiring information may be received from the wallet system 100, It may not be open to the public.

The information of the job seeker may include information of all the job seekers and may include only information of the job seekers filtered by the talent matching server 300 among the information in the wallet system 100. [

The enterprise recruiter can confirm the information of the job seeker in the company mailbox 200 through the filtering system or the browsing system.

The information of a job seeker can also be created by a job seeker directly connecting to the company mailbox 200 without depending on the Wallet system 100 or the talent matching server 300. [

The job seeker can access the corporate mailbox 200 through a smart phone application or a web application (Web App.), Access the corporate mailbox 200 through a personal PC or the like to create information of a job seeker, You can also use the native app environment on a web environment such as a PC or a smartphone using an operating system such as Android or iOS.

When the job seeker directly accesses and creates the information of the job seeker, the information stored in the company mailbox 200 is additionally required to be authenticated.

The job seeker can directly access the company mailbox 200 or submit the necessary documents to the company mailbox 200 through the wallet system 100 paperlessly so that the documents can be output for each company supported by the job seeker to be visited or mailed It is possible to avoid the hassle of accepting the document.

In addition, corporate recruiters can receive paperless materials (diploma certificate, transcript, TOEIC report, MOS transcript, etc.) validated through Wallet System 100, And the cost can be reduced.

The authentication process may be performed by connecting to the wallet system 100 and collectively processing, or by separately accessing and authenticating the issuing organizations of the respective documents.

The talent matching server 300 recommends talent suitable for the company by using the information of the job seeker provided from the wallet system 100 and the job information and the job condition provided in the company mailbox 200. [

The talent matching server 300 searches for suitable job for a job seeker by using qualitative information such as a job seeker's credit score, English score and personal information such as personality test and counseling result, And performs optimal matching.

For this, the talent matching server 300 analyzes the quantitative information and the qualitative information based on the big data analysis technique, and obtains the numerical quantified values by analyzing the results through the machine learning based quantification algorithm, You can recommend the talent you want.

2 shows a specific configuration of the talent matching server 300. As shown in FIG.

The talent matching server 300 includes a job search module 310 and a job matching module 320.

The job search module 310 receives the information of the job seeker from the wallet system 100 and calculates the fitness of the job seeker according to the job as quantified data.

NCS (National Competency Standards) classification can be used to classify jobs of job seekers.

The NCS classification is a national incompetence standard, which is a systematized state of knowledge, skills and attitudes required to perform tasks in the industrial field. Since NCS classification is classified into about 887 detailed jobs, it can be used as a basis for job seekers' job search. Specifically, they are classified in the order of 'major division (24) → middle division (80) → minor division (238) → sub division (887)'.

The NCS classification has the effect of determining the specific job of the job seeker compared to the conventional technology because the job is classified in detail.

Table 1 shows an example of the NCS classification.

Job classification can be classified not only by NCS job classification, but also by Job Title itself, and Job Classification can be provided by Company Mailbox 200 when the job classification of the company itself is used.

The job search module 310 divides the information of the job seeker provided in the Wallet system 100 into quantitative information and qualitative information. Job seeker information includes personal information, academic background, career experience, overseas experience, foreign language ability, special features, volunteer activities, self introduction, and other matters.

The job search module 310 first determines a job suitable for a job seeker using quantitative information.

FIG. 3 shows a method for determining a job based on a decision tree algorithm based on quantitative information.

The information of candidate applicant A is obtained by using the TOEIC score, the volunteer activity time, and the like acquired from the wallet system 100, and then, in each node of the decision tree (decision tree learning method) As a result of repeating the selection at each node, it is possible to obtain the result that NCS classification of overseas business is most suitable for applicant A's job. In addition to the decision tree algorithm, a variety of machine learning algorithms can be used for quantitative analysis.

Selection criteria at each node can be updated with optimized criteria through machine learning and can be updated with new criteria through evaluation of employees employed by the talent matching system 10. [

After the job is determined by quantitative information, the fitness of each job is calculated using qualitative information.

FIG. 4 is a diagram illustrating an example of a method of calculating fitness by job using text mining using qualitative information.

When the text mining technology is used, it is possible to optimize prediction of job suitability and job performance by extracting formalized features based on individual linguistic statistical data using the big data collected through the individual wallet system 100.

The predicted result is processed again on the basis of the machine learning to form the semi-structured / non-structured individual data, thereby providing the data of the job seeker matched with the optimal talent meeting the conditions required by the recruiting enterprise .

In the qualitative information such as the self-introduction letter or the counseling result of the employment center, the words corresponding to the job seeker are extracted using the text mining technique.

Based on the extracted words, the weight can be determined using machine learning, and the fitness of each job can be evaluated quantitatively.

Fit-specific fitness values are further refined by iterative machine learning and updated continuously through feedback based on company requirements or talent recruitment results, so that more accurate job-specific fit can be calculated.

It is possible to derive the final fitness ranking of each job by synthesizing the fitness according to job by qualitative information and quantitative information. In this case, the accuracy of the result can be further improved by multiplying the result of the qualitative information and the result of the quantitative information by the weights, respectively.

Qualitative data such as self-introduction, recommendation letters, and motivation motives are analyzed from the linguistic aspect through text mining techniques, and the fitness and job performance required by job type And by using it to supplement deficient parts of the analyzed data of quantitative data, it is possible to recommend a job seeker to the optimal job through the talent matching system 10.

The job matching module 320 performs matching on the talents required by the company using the job relevance result of the job seeker by the job search module 310.

FIG. 5 shows an example of a method of recommending a talent using job-specific fitness of each job.

It identifies the personnel per job required by the company, and matches the job applicants' job satisfaction by using the quantitative and qualitative information, and performs talent matching in a way that recommends suitable talents.

The talent matching server 300 can improve the system by utilizing the information of the existing staff as well as the information of the job seeker.

By analyzing quantitative and qualitative information of existing employees and comparing the information of existing employees' performance, personnel reports and years of service, it is possible to analyze quantitative and qualitative information and correlations with the talents the company desires, The performance of the system 10 can be improved.

In addition, the talent matching server 300 continuously feedbacks the result values after the employment of the talent, such as the type of job announcement by company, job suitability, and job performance satisfaction, and uses it for repetitive machine learning of the artificial intelligence system. .

6 shows a flowchart of a talent matching method according to another embodiment of the present invention.

In the talent matching method, the talent matching server 300 receives information from the wallet system 100 and the company mailbox 200, and matches the optimal job seekers required for the company by job.

First, the quantitative information of the job seeker is inputted (S610).

The quantitative information of a job seeker can be input information that can be quantitatively evaluated by numerical values such as a credit or a language score among the information of a job seeker stored in the Wallet system 100. [

A job seeker can directly input his / her information through a native application or a web app installed on a smart phone or a tablet PC without going through the wallet system 100.

In the case where a job seeker inputs information directly, rather than the wallet system 100, the talent matching server 300 accesses the wallet system 100 for authenticating the quantitative information and confirms whether the data is authentic or not, Additional authentication may be required, such as verifying the original.

After inputting the quantitative information, the jobs of the job seeker are decided by the job selection algorithm based on the decision tree based on the quantitative information.

Decision tree The tree-based job selection algorithm can determine the job that is suitable for the job seeker by repeating the selection by the quantitative information of the job seeker at each node of the decision tree as described above.

The job determined in this way is not the final but the job fitness of the job seeker is calculated considering the job fit obtained by the qualitative information.

Upon completion of the analysis of the quantitative information, information such as a self introduction sheet which can not be quantified is input (S630).

Because qualitative information can not use decision tree based selection method like quantitative information, it uses machine learning based job fitness calculation method. To do this, we use text mining technique in quantitative information, (S640).

Various techniques such as Support Vector Machine (SVM), Principal Component Analysis (PCA), Random Forest, and Neural Network can be used as data mining techniques for extracting text.

After extracting the words from the qualitative information, the job fitness is calculated using the extracted words, and the final job fitness is calculated considering the job determined by the quantitative information (S650).

As shown in FIG. 4, the degree of fitness by machine learning is calculated by quantifying the degree suitable for each job by weight, as shown in FIG. 4, by calculating the fitness using machine learning using the weight for each extracted word.

As the machine learning method, various algorithms such as artificial neural network, association rule learning method, genetic programming method, inductive logic programming method, clustering, and Bayesian network can be used.

When the computation of fitness by job by qualitative information is completed, the final job fitness is calculated by considering the job fitness obtained by quantitative information together.

The final job fit can be obtained by correcting the job fit obtained by the qualitative information by adding or multiplying weights by the job information obtained by the quantitative information.

For example, as shown in FIG. 4, if the job satisfaction based on qualitative information is 85.6% for overseas business and 47.2% for technology development and the job based on quantitative information is for overseas business, the final result is a 10% , 95.6% and technology development 47.2%, and the overseas business result is multiplied by the weight K, multiplied by 1.1, for example, and 94.16% can be calculated as the final job fitness for overseas business.

This weight A or weight K can be updated to obtain more accurate results by repeated machine learning and can be updated to find the desired talent through analysis of the company's existing staff.

Finally, when the calculation of job satisfaction of the job seeker is completed, the result is used to match the talent in consideration of the number of people employed by the job (S660).

It is also possible to finally filter job seekers at the bottom of the job fit in the final talent matching step. For example, even if the job seeker has the highest fitness for a specific job, if the job applicant does not fit 80% of the job, it does not match the talent.

By setting the absolute filtering criterion, it is possible to further improve the accuracy and reliability of human matching.

In the job matching stage, it is possible to match both the tacit preference of the jobseeker of the company and the mutual benefit of the jobseeker and the company together.

For job seekers and firms with high preference, weights are given so that they can reflect preferences in the job fitness score.

When all the steps of analyzing the information of the job seeker and calculating the job fitness and matching the talent are completed, the matching result is feedbacked to improve the system (S670).

Decision tree algorithms, text mining, and machine learning based quantification all have an effect of improving algorithms by machine learning based on big data. Therefore, it is possible to improve the accuracy of each step by reflecting the result of job fitness calculation to machine learning.

By connecting the job seeker and the company by matching the talent with consideration of the quantitative and qualitative information based on the above-mentioned Wallet system, not only the job search activity of the company or the job search time of the individual in the conventional manual work can be shortened, Matching algorithms can provide more accurate human matching.

Meanwhile, the talent matching method according to an embodiment of the present invention can be implemented in a computer system or recorded on a recording medium. 7, the computer system includes at least one processor 721, a memory 723, a user input device 726, a data communication bus 722, a user output device 727, And a storage 728. Each of the above-described components performs data communication via a data communication bus 722. [

The computer system may further include a network interface 129 coupled to the network. The processor 721 may be a central processing unit (CPU) or a semiconductor device that processes instructions stored in the memory 723 and / or the storage 728.

The memory 723 and the storage 728 may include various forms of volatile or non-volatile storage media. For example, the memory 723 may include a ROM 724 and a RAM 725.

Accordingly, the talent matching method according to the embodiment of the present invention can be implemented in a computer-executable method. When the talent matching method according to an embodiment of the present invention is performed in a computer device, computer-readable instructions can perform the recognition method according to the present invention.

Meanwhile, the above-described human resource matching method according to the present invention can be implemented as a computer-readable code on a computer-readable recording medium. The computer-readable recording medium includes all kinds of recording media storing data that can be decoded by a computer system. For example, there may be a ROM (Read Only Memory), a RAM (Random Access Memory), a magnetic tape, a magnetic disk, a flash memory, an optical data storage device and the like. The computer-readable recording medium may also be distributed and executed in a computer system connected to a computer network and stored and executed as a code that can be read in a distributed manner.

While the present invention has been described in detail with reference to the accompanying drawings, it is to be understood that the invention is not limited to the above-described embodiments. Those skilled in the art will appreciate that various modifications, Of course, this is possible. Accordingly, the scope of protection of the present invention should not be limited to the above-described embodiments, but should be determined by the description of the following claims.

10: Talent Matching System
100: Wallet system 200: Corporate mailbox
300: Talent matching server
310: job search module 320: job matching module

Claims (10)

The information of the job seeker is classified into quantitative information which can be quantified and qualitative information which can not be quantified,
Determining a job that is suitable for the job seeker by the first machine learning method using the quantitative information,
Extracting characteristic words by data mining using the qualitative information, calculating fitness of the job seekers according to the job by a second machine learning method using the extracted words,
A job search module for calculating a fitness score for each job by using the job determined using the quantitative information and the fitness score calculated using the qualitative information; And
A job matching module for determining a recommended job seeker and recommendation job by using job information for each job of the company and the relevance of the job seeker according to the final job;
Wherein the server comprises:
2. The method of claim 1, wherein the first machine learning method comprises:
Using decision tree learning methods
In-Person Matching Server.
The method according to claim 1,
Wherein the first machine learning method or the second machine learning method is performed by analyzing information of a job seeker whose recommendation has been determined by the job matching module or an employee of the company
In-Person Matching Server.
The method according to claim 1,
The classification of the job is classified using NCS (National Competency Standards) classification
In-Person Matching Server.
The method of claim 1, wherein the information of the applicant
And is provided from a wallet system for storing authenticated information by verifying whether or not the applicant information is authentic
In-Person Matching Server.
A talent matching method performed by a talent matching server comprising one or more processors,
Receiving information of a job seeker;
Classifying the information of the applicant into quantitative information that can be digitized and qualitative information that can not be digitized;
Determining a job suitable for the job seeker by a first machine learning method using the quantitative information;
Extracting feature words by data mining using the qualitative information;
Calculating a goodness-of-fit of the job seeker by a second machine learning method using extracted words;
Calculating fitness for each job by using the job determined using the quantitative information and the fitness calculated for each job using the qualitative information; And
Determining a recommended job seeker and recommendation job by using the job offer information of each job according to the job and the fitness degree of the job candidate according to the final job;
Wherein the method comprises the steps of:
7. The method of claim 6, wherein the step of determining the recommender job seeker
The preference of the enterprise to the job seeker or the preference of the job seeker to the enterprise is weighted
In-person talent matching method.
7. The method of claim 6, further comprising: after the step of determining the recommender job seeker and recommendation job
And a feedback step of updating the first machine learning method or the second machine learning method using the determined recommended job seeker and recommendation job results
In-person talent matching method.
7. The method of claim 6, wherein the information of the applicant
And is provided from a wallet system that stores information authenticated as authentic by verifying authenticity of the applicant information
In-person talent matching method.
A wallet system including a medical database storing clinical information verified by a school server and a personal database storing personal option information of a job seeker;
Corporate mailboxes that store job hiring or job terms; And
The information of the job seeker provided from the wallet system is classified into quantitative information that can be quantified and qualitative information that can not be quantified,
Determining a job that is suitable for the job seeker by the first machine learning method using the quantitative information,
Extracting characteristic words by data mining using the qualitative information, calculating fitness of the job seekers according to the job by a second machine learning method using the extracted words,
The fitness of each job is calculated by using the job determined using the quantitative information and the fitness according to job calculated using the qualitative information,
A talent matching server for determining a recommended job seeker and a recommendation job by using the job candidate information of the firm and the fitness degree of each job candidate according to the final job;
And a human resource matching system.

Images