KR20230149751A - Information processing device for operating in conjunction with job and department recommendation programs using arificial intelligence and method for operating the same - Google Patents

Abstract

The present invention relates to an information processing device and its operation method that operate in conjunction with a job and department recommendation program using artificial intelligence. The problem to be solved is a career aptitude diagnostic test and career/advanced education by type for a large number of learners. Through the process of implementing content services, we utilize the learner data accumulated to analyze the type of learner, minimize the learner's adverse reactions and atypical reactions that occur during the career aptitude diagnostic test, and identify jobs with low preference by the learner. By going through the job selection process to minimize each job and clustering the data classified according to type and series based on the learner's selection rate, it is possible to recommend jobs by type with high satisfaction, and by classifying and clustering jobs based on the learner's preference by type. The aim is to prevent unnecessarily extensive job recommendations through a simple intersection of types by recommending jobs and going through job selection to minimize jobs with low learner preferences.
For example, data on occupations of interest selected by the learner among job data temporarily recommended through career aptitude diagnostic tests for each type, and codebook data containing keywords entered by the learner in the process of providing career/advanced education content, It relates to an information processing device that operates in conjunction with a program that recommends occupations and departments. It constructs a weight DB for each occupation by assigning weights to the occupation data of interest based on the learner's occupation preference, and stores the weight DB for each occupation. An occupation cluster DB construction unit that constructs an occupation cluster DB by clustering the data based on the learner's occupation preference; Input keywords included in the codebook data are collected by predefined career aptitude types, clustering is performed based on the frequency of the collected input keywords to construct a keyword cluster DB, and the clusters stored in the keyword cluster DB are classified according to the frequency. A weighted keyword DB construction unit that constructs a weighted keyword DB by assigning weights; A learner data analysis unit that collects input keywords that match keywords included in the weighted keyword DB among the input keywords included in the codebook data of individual learners, and derives career aptitude type data of the learner according to the frequency of the collected input keywords; And an occupation and department recommendation unit that recommends job clusters containing the career aptitude type data among the job clusters stored in the job cluster DB, and finally recommends job and department data belonging to the job cluster with the highest weight among the recommended job clusters. Disclosed is an information processing device comprising:

Description

Information processing device and method of operation operating in conjunction with a job and department recommendation program using artificial intelligence

Embodiments of the present invention relate to an information processing device that operates in conjunction with a job and department recommendation program using artificial intelligence and a method of operating the same.

Generally, students or job seekers worry a lot about their career path or job. For example, you worry about which department or job is right for you and which department you should choose to get the job that suits you.

In most cases like this, there is a problem that cannot be found with a clear answer, although we vaguely talk about it with school teachers, parents, and friends. Therefore, most students or job seekers often make decisions arbitrarily, considering their grades, school preferences, and aptitude.

To overcome these problems, there is an increasing need to automatically recommend jobs that suit students based on student record information and the department of the university they wish to attend when selecting that job.

In this regard, the conventional job recommendation service analyzes the learner's type through a diagnostic test and provides a service that recommends a job group that is a simple union of jobs according to type.

However, this conventional job recommendation service has the problem of recommending an unnecessarily large number of jobs by providing a simple union of job data such as multiple intelligences, interests, values, and personality.

In addition, the conventional job recommendation service has low reliability for the test results of learners with low career maturity, so the accuracy of job recommendations is inevitably lowered, and subjects who answer without properly reading the contents of the questions or respond randomly tend to have abnormal reactions. Alternatively, there are limitations in detecting atypical response tendencies.

Public Patent Publication No. 10-2016-0025486 (Publication date: March 8, 2016) Public Patent Publication No. 10-2021-0153953 (Publication date: December 20, 2021) Public Patent Publication No. 10-2015-0112596 (Publication date: October 7, 2015) Public Patent Publication No. 10-2018-0090624 (Publication Date: August 13, 2018)

An embodiment of the present invention analyzes the type of learner by utilizing learner data accumulated through the process of conducting career aptitude diagnostic tests and career/advanced education content services by type for a large number of learners, and when performing a career aptitude diagnostic test. Satisfaction satisfaction by going through the task of minimizing the learner's adverse reactions and atypical reactions that occur and the job selection task of minimizing jobs with low learner preference, and clustering the data classified according to type and series based on the learner's selection rate. It is possible to recommend jobs by high type, recommend jobs by classifying and clustering jobs based on the learner's preference by type, and go through a job selection process to minimize jobs with low learner preference, thereby recommending an unnecessarily large number of jobs by simple intersection by type. Provides an information processing device and method of operation that operate in conjunction with a job and department recommendation program using artificial intelligence that can prevent.

An information processing device that operates in conjunction with a job and department recommendation program using artificial intelligence according to an embodiment of the present invention includes job data of interest selected by the learner among job data temporarily recommended through a career aptitude diagnostic test for each type, and career path. /This relates to an information processing device that operates in conjunction with a program that recommends an individual learner's occupation and department based on code book data containing keywords entered by the learner in the process of providing educational content, and the learner's occupation data of interest An occupation cluster DB construction unit that constructs an occupation-specific weight DB by assigning weights based on the occupation preference, and constructs an occupation cluster DB by clustering the data stored in the occupation-specific weight DB based on the learner's occupation preference; Input keywords included in the codebook data are collected by predefined career aptitude types, clustering is performed based on the frequency of the collected input keywords to construct a keyword cluster DB, and the clusters stored in the keyword cluster DB are classified according to the frequency. A weighted keyword DB construction unit that constructs a weighted keyword DB by assigning weights; A learner data analysis unit that collects input keywords that match keywords included in the weighted keyword DB among the input keywords included in the codebook data of individual learners, and derives career aptitude type data of the learner according to the frequency of the collected input keywords; And an occupation and department recommendation unit that recommends job clusters containing the career aptitude type data among the job clusters stored in the job cluster DB, and finally recommends job and department data belonging to the job cluster with the highest weight among the recommended job clusters. Includes.

In addition, the occupation cluster DB construction unit collects occupation data by performing web crawling based on the occupation data of interest, considers the number of collected occupation data as the learner's occupation preference, and determines the learner's interest according to the occupation preference. A weight generation unit for each occupation that assigns weights for each occupation to build a weight DB for each occupation; and an occupation cluster generator that constructs the occupation cluster DB by classifying the data stored in the occupation-specific weight DB into groups with similar occupation preferences through K-means clustering based on the learner's occupation preference.

In addition, the weighted keyword DB construction unit includes a first input keyword collection unit that collects input keywords including nouns and verbs included in the codebook data by predefined career aptitude types; An input keyword cluster performing unit that determines the frequency of input keywords collected by career aptitude type through the first input keyword collection unit and constructs a keyword cluster DB by performing clustering on the input keywords according to the identified frequency; and an input keyword weight generator that constructs a weighted keyword DB by assigning a weight to each keyword cluster according to the frequency of the keyword cluster stored in the keyword cluster DB.

In addition, the learner data analysis unit includes a second input keyword collection unit that collects input keywords that match keywords included in the weighted keyword DB among input keywords including nouns and verbs included in the individual learner's codebook data; And it may include a learner type analysis unit that derives career aptitude type data of the learner according to the frequency of the input keywords collected through the second input keyword collection unit.

In addition, the occupation and department recommendation unit includes a conflicting interest type removal unit that extracts an occupation of a conflicting interest type from the career aptitude type data derived through the learner data analysis unit and removes the extracted conflicting interest type occupation data; a cluster-based job recommendation unit that recommends a job cluster including career aptitude type data from which the conflicting interest type job data has been removed among job clusters stored in the job cluster DB; And when an occupation of interest is selected by an individual learner for an occupation cluster recommended through the cluster-based job recommendation unit, a preset weight is given to the occupation selected as the occupation of interest, and then the occupation and department data with the highest weight among the occupation clusters are collected. If you are not selected for the final recommendation or job of interest, or if you take the career aptitude test again and the results are different from the previous career aptitude type data, the job and department that belong to the job cluster with the highest weight among the recommended job clusters. After final recommendation of the data, the career aptitude diagnostic test is performed again, and the result is the same as the previous career aptitude type data, it includes a final job and department recommendation section that finally recommends the job and department data belonging to the job cluster with the next best weight. can do.

A method of operating an information processing device linked to a job and department recommendation program using artificial intelligence according to another embodiment of the present invention includes job data of interest selected by the learner among job data temporarily recommended through a career aptitude diagnostic test for each type, and It relates to a method of operating an information processing device that links with a program that recommends an individual learner's occupation and department based on code book data containing keywords entered by the learner in the process of providing career/advanced education content, and the occupation data of interest Construct an occupational cluster DB by assigning weights based on the learner's occupational preference to construct an occupation-specific weight DB, and clustering the data stored in the occupation-specific weight DB based on the learner's occupational preference to construct an occupation cluster DB. step; Input keywords included in the codebook data are collected by predefined career aptitude types, clustering is performed based on the frequency of the collected input keywords to construct a keyword cluster DB, and the clusters stored in the keyword cluster DB are classified according to the frequency. A weighted keyword DB construction step of constructing a weighted keyword DB by assigning weights; A learner data analysis step of collecting input keywords that match keywords included in the weighted keyword DB among the input keywords included in the individual learner's codebook data, and deriving the learner's career aptitude type data according to the frequency of the collected input keywords; And a job and department recommendation step that recommends job clusters containing the career aptitude type data among job clusters stored in the job cluster DB, but finally recommends job and department data belonging to the job cluster with the highest weight among the recommended job clusters. Includes.

In addition, in the occupation cluster DB construction step, occupation data is collected by performing web crawling based on the occupation data of interest, the number of collected occupation data is regarded as the learner's occupation preference, and the learner's occupation preference is determined according to the occupation preference. A weight generation step for each occupation of constructing a weight DB for each occupation by assigning weights to each occupation of interest; And it may include an occupation cluster creation step of constructing the occupation cluster DB by classifying the data stored in the occupation-specific weight DB into groups with similar occupation preferences through K-means clustering based on the learner's occupation preference.

In addition, the weighted keyword DB construction step includes: a first input keyword collection step of collecting input keywords including nouns and verbs included in the codebook data for each predefined career aptitude type; An input keyword clustering step of determining the frequency of input keywords collected by career aptitude type through the first input keyword collection step and constructing a keyword cluster DB by performing clustering on the input keywords according to the identified frequency; And it may include an input keyword weight generation step of constructing a weighted keyword DB by assigning a weight to each keyword cluster according to the frequency of the keyword cluster stored in the keyword cluster DB.

In addition, the learner data analysis step includes a second input keyword collection step of collecting input keywords that match keywords included in the weighted keyword DB among input keywords including nouns and verbs included in the individual learner's codebook data; And it may include a learner type analysis step of deriving the learner's career aptitude type data according to the frequency of the input keywords collected through the second input keyword collection step.

In addition, the occupation and department recommendation step includes: a conflicting interest type removal step of extracting a conflicting interest type occupation from the career aptitude type data derived through the learner data analysis step and removing the extracted conflicting interest type occupation data; A cluster-based job recommendation step of recommending a job cluster including career aptitude type data from which the conflicting interest type job data has been removed among job clusters stored in the job cluster DB; And when an occupation of interest is selected from an individual learner for an occupation cluster recommended through the cluster-based job recommendation step, a preset weight is given to the occupation selected as the occupation of interest, and then the occupation and department data with the highest weight among the occupation clusters are collected. If the final recommendation is made or if the job of interest is not selected, the career aptitude test is re-administered and the results are different from the previous career aptitude type data. Among the recommended job clusters, the job belonging to the job cluster with the highest weight and Final job and department recommendation stage in which the department data is finally recommended, the career aptitude diagnostic test is performed again, and if the results are the same as the previous career aptitude type data, the job and department data belonging to the job cluster with the next best weight are finally recommended. may include.

According to the present invention, the type of learner is analyzed using learner data accumulated through the process of conducting career aptitude diagnostic tests and career/advanced education content services by type for a large number of learners, and the types of learners generated during the career aptitude diagnostic test are analyzed. After going through the task of minimizing the learner's adverse reactions and atypical reactions and the job selection task of minimizing jobs with low preference by the learner, data classified according to type and series are clustered based on the learner's selection rate to classify the types with high satisfaction. Job recommendation is possible, and jobs are recommended by classifying and clustering jobs based on the learner's preference for each type, and go through a job selection process to minimize jobs with low learner preference, thereby preventing unnecessarily extensive job recommendations through simple intersection by type. An information processing device and its operation method that operate in conjunction with a job and department recommendation program using artificial intelligence can be provided.

Figure 1 is a block diagram showing the configuration of a job and department recommendation device using artificial intelligence according to an embodiment of the present invention.
Figure 2 is a diagram showing an example of a learner's codebook data and the analysis type analyzed by the learner data analysis unit according to an embodiment of the present invention.
Figure 3 is a diagram showing an example of keyword configuration by type stored in the keyword DB according to an embodiment of the present invention.
Figure 4 is a diagram illustrating in detail the direct recommendation method of the job recommendation unit according to an embodiment of the present invention.
Figure 5 is a flowchart showing the configuration of a job and department recommendation method using artificial intelligence according to another embodiment of the present invention.

The terms used in this specification will be briefly explained, and the present invention will be described in detail.

The terms used in the present invention are general terms that are currently widely used as much as possible while considering the function in the present invention, but this may vary depending on the intention or precedent of a person working in the art, the emergence of new technology, etc. In addition, in certain cases, there are terms arbitrarily selected by the applicant, and in this case, the meaning will be described in detail in the description of the relevant invention. Therefore, the terms used in the present invention should be defined based on the meaning of the term and the overall content of the present invention, rather than simply the name of the term.

When it is said that a part "includes" a certain element throughout the specification, this means that, unless specifically stated to the contrary, it does not exclude other elements but may further include other elements. In addition, terms such as "... unit" and "module" used in the specification refer to a unit that processes at least one function or operation, which may be implemented as hardware or software, or as a combination of hardware and software. .

Below, with reference to the attached drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily implement the present invention. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein. In order to clearly explain the present invention in the drawings, parts that are not related to the description are omitted, and similar parts are given similar reference numerals throughout the specification.

The job and department recommendation device using artificial intelligence according to an embodiment of the present invention provides job data of interest selected by the learner among job data temporarily recommended through a career aptitude diagnostic test for each type, and career/advanced education content in the process of providing. It can be comprised of an information processing device that operates in conjunction with a program that recommends an individual learner's occupation and department based on code book data containing keywords entered by the learner.

In this way, prior to an embodiment of the present invention, a 'career aptitude 3-factor test' and a 'codebook' can be conducted. The career aptitude 3-factor test is a type-specific career aptitude test conducted on a large number of learners (all users of the platform). As a diagnostic test, it consists of an occupational interest test, multiple intelligence test, and occupational values test. It can test career aptitude for each type. Occupational data is temporarily recommended for each learner through the diagnostic test data, and the recommended direct learners are You can provide each job of interest data selected by you.

In addition, Codebook is a career and advancement education content service that provides career and advancement related content and can obtain codebook data containing the learner's response (input keyword) to the inquiry.

In one embodiment of the present invention, individual learner customization with higher accuracy is performed along with machine learning algorithms among artificial intelligence technologies based on learner data including the above-mentioned diagnostic test data (including job interest data) and codebook data (including learner input keywords). Job and department (career) recommendation services can be provided.

Figure 1 is a block diagram showing the configuration of a job and department recommendation device using artificial intelligence according to an embodiment of the present invention, and Figure 2 is an example of a learner's codebook data and analysis of learner data according to an embodiment of the present invention. It is a diagram showing the analysis type analyzed in the unit, Figure 3 is a diagram showing an example of keyword configuration by type stored in the keyword DB according to an embodiment of the present invention, and Figure 4 is a direct diagram of the job recommendation unit according to an embodiment of the present invention. This diagram is shown to explain the recommendation method in detail.

Referring to FIG. 1, the occupation and department recommendation device 1000 using artificial intelligence according to an embodiment of the present invention includes an occupation cluster DB construction unit 100, a weighted keyword DB construction unit 200, and a learner data analysis unit. It may include at least one of (300) and a job and department recommendation section (400).

The occupation cluster DB construction unit 100 constructs an occupation-specific weight DB (3) by assigning weights to the occupation data of interest based on the learner's occupation preference, and transfers the data stored in the occupation-specific weight DB (3) to the learner's occupation data. An occupational cluster DB (4) can be constructed by performing clustering based on occupational preference.

To this end, the occupation cluster DB construction unit 100 may include at least one of an occupation-specific weight generator 110 and an occupation cluster generator 120.

The occupation-specific weight generator 110 performs web crawling based on the occupation data of interest stored in the learner's interest occupation DB (1), collects occupation data in the occupation DB (2), and calculates the number of collected occupation data. It is regarded as the learner's job preference, and a weight DB (3) for each job can be constructed by assigning weights to each job of interest of the learner according to the job preference.

The learner's career interest DB (1) contains the career data selected by the learner as the career of interest among the career data recommended to each learner after the three-factor career aptitude test, and data obtained by running Codebook, a career and advancement education content service (codebook data). Occupation data (occupation data of interest) selected by the learner as an occupation of interest derived from analysis may be stored.

The occupation-specific weight generator 110 collects various occupation data related to the learner's interest occupation data stored in the learner's interest occupation DB (1) through web crawling and accumulates it in the occupation DB (2), Classify and analyze the job data accumulated in the job DB (2) by type and series, set weights for the job data according to the learner's preference (occupation of interest), and create job data with weights reflected based on the learner's preference. It can be secured.

For example, as a basis for generating weights for each occupation, a weight of 1 point can be generated when the occupation of interest is accumulated 200 times. When targeting approximately 1,000 initial learners, the total number of career selections of interest is 3,900 times, the maximum number of specific career selections is 1,300 times, and the standard for generating a weight of 1 point can be set based on 200 times, which is about 15% of the 1,300 times. (The maximum number of times a specific occupation is selected as an occupation of interest is 1,300 times). Here, if a standard exceeding 15% is set, the initial barrier to creating weights is high, so it is not easy to secure weighted data, and if a standard less than 15% is set, weights are created for multiple occupations, making it difficult to obtain weighted data. The discriminatory power of star weights may be lowered.

The occupation cluster generator 120 is a means of classifying similar groups based on the learner's preference and clustering them into clusters, and divides the data stored in the occupation-specific weight DB 3 into K- based on the learner's occupation preference. An occupational cluster DB (4) can be constructed based on data obtained by classifying groups with similar occupational preferences through K-means clustering.

The occupation cluster generator 120 is a module that classifies data stored in the occupation-specific weight DB into similar groups based on preference and clusters them into clusters. Each cluster may correspond to a group of occupations by type with similar preferences. This occupation cluster generator 120 can classify the data in the occupation-specific weight DB 3 using the K-means clustering algorithm, which corresponds to a data object according to the K-means clustering algorithm. n occupation data can be divided into k groups less than or equal to n to form an occupation cluster.

For example, as a criterion for creating clusters by occupation, it can be clustered into five occupation clusters based on the weight of each occupation.

In this way, the process of dividing groups by the K-means clustering algorithm can be accomplished by minimizing the cost function, and in this process, the preference-based similarity between job data within the same group increases and the similarity with data objects in other groups increases. It decreases.

In addition, the K-means clustering algorithm can perform clustering by setting the sum of the squares of the distance between the center of each group and the data objects within the group as a cost function and updating the group membership of each data object in the direction of minimizing the value of this function. there is.

The weighted keyword DB construction unit 200 collects the input keywords included in the codebook data by predefined career aptitude type and constructs the keyword cluster DB (5) by performing clustering based on the frequency of the collected input keywords. In addition, a weighted keyword DB (6) can be constructed by assigning weights according to frequency to the clusters stored in the keyword cluster DB (5).

To this end, the weight keyword DB construction unit 200 may include at least one of a first input keyword collection unit 210, an input keyword cluster execution unit 220, and an input keyword weight generation unit 230.

The first input keyword collection unit 210 is a module that collects learner data (diagnostic test data, codebook data), and as shown in Figure 2, a keyword DB among input keywords including nouns and verbs included in the codebook data. Input keywords that match the data stored in (7) can be collected and classified by the collected career aptitude types (occupational interest type, multiple intelligence type, occupational value type).

More specifically, through the three-factor career aptitude test, codebook data of learners classified by type can be secured and learners' input keywords can be collected according to frequency. For example, self-understanding capabilities among the codebook data of social learners. Words with high frequency in questions can be accumulated as social keywords.

The input keyword cluster performing unit 220 determines the type and frequency of input keywords collected by career aptitude type through the first input keyword collection unit 210, and clusters the input keywords according to the identified type and frequency. You can build a keyword cluster DB (5) by performing .

The input keyword weight generator 230 may construct the weighted keyword DB 6 by assigning a weight to each keyword cluster according to the frequency of the keyword cluster stored in the keyword cluster DB 5. In other words, weights can be assigned to keywords formed in high-frequency clusters. For example, weights of 1 to 10 can be created, and weights of 10 to 1 can be assigned in the order of clusters of high-frequency attributes. there is.

The learner data analysis unit 300 collects input keywords that match keywords included in the weighted keyword DB (6) (or keyword DB (7)) among the input keywords included in the individual learner's codebook data, and collects the collected input keywords. Depending on the frequency of input keywords, the learner's career aptitude type data can be derived.

Here, the keyword DB (7) is a DB that classifies key keywords according to each type derived as a result of the career aptitude three-factor test (occupational interest, multiple intelligence, and occupational values test) as shown in Figure 3, and is classified into occupation and department. Data classified by type can be stored, including nouns and verbs.

This learner data analysis unit 300 may include at least one of a second input keyword collection unit 310 and a learner type analysis unit 320.

The second input keyword collection unit 310 inputs input that matches the keyword included in the weighted keyword DB (6) (or keyword DB (7)) among input keywords including nouns and verbs included in the individual learner's codebook data. You can collect keywords.

The learner type analysis unit 320 may derive the learner's career aptitude type data according to the frequency of the input keywords collected through the second input keyword collection unit 310. In other words, the learner's job interest, multiple intelligence, and job value type (learner's career aptitude type data) can be derived based on the frequency of the collected input keywords.

The job and department recommendation unit 400 recommends job clusters containing career aptitude type data among job clusters stored in the job cluster DB 4, and includes jobs belonging to the job cluster with the highest weight among the recommended job clusters. Department data can be finally recommended.

To this end, the job and department recommendation unit 400 may include at least one of a conflicting interest type removal unit 410, a cluster-based job recommendation unit 420, and a final job and department recommendation unit 430.

The conflicting interest type removal unit 410 extracts the occupation of the conflicting interest type from the career aptitude type data (learner's occupational interest, multiple intelligence, and occupational value type data) derived through the learner data analysis unit 300, The extracted conflicting interest type occupation data can be removed.

The conflicting interest type removal unit 410 can increase job recommendation accuracy by removing jobs of conflicting interest types that conflict with the learner's interest type (can be confirmed from career aptitude type data), and remove jobs by removing conflicting interest types. This can prevent the accuracy of recommendations from being reduced or from recommending a large number of meaningless jobs.

For example, if the learner's interest type is field-type, social-type jobs, which are the interest type with the lowest consistency with field-type, can be excluded. In this embodiment, the conflicting interest type refers to an interest type with a low level of consistency among Holland's occupational interest types.

The cluster-based job recommendation unit 420 may recommend an job cluster including career aptitude type data from which conflicting interest type job data has been removed among job clusters stored in the job cluster DB 4. In other words, the intersection of occupations included in the learner's type cluster (occupational interest type cluster, multiple intelligence type cluster, occupational value type cluster) in the occupational cluster DB (4) can be recommended.

The final job and department recommendation unit 430 is a job cluster (job interest type cluster, multiple intelligence type cluster, job value type cluster) recommended through the cluster-based job recommendation unit 420, as shown in FIG. When an occupation of interest (or series of interest) is selected by an individual learner, a preset weight is given to the occupation selected as the occupation of interest (or series of interest), and then the occupations belonging to the intersection of the direct cluster with the highest weight among the occupation clusters are selected. Department data can be finally recommended.

In addition, the final occupation and department recommendation unit 430, as shown in FIG. 4, if the occupation of interest is not selected, performs the career aptitude diagnostic test again and if the result is different from the previous career aptitude type data , among the recommended occupation clusters, the occupation and department data belonging to the intersection of the occupation cluster with the highest weight can be finally recommended.

In addition, the final occupation and department recommendation unit 430, as shown in FIG. 4, if the occupation of interest is not selected, performs the career aptitude diagnostic test again and if the result is the same as the previous career aptitude type data , the occupation and department data belonging to the intersection of occupation clusters with suboptimal weights can be finally recommended.

Figure 5 is a flowchart showing the configuration of a job and department recommendation method using artificial intelligence according to another embodiment of the present invention.

Referring to Figure 5, the job and department recommendation method using artificial intelligence (S1000) according to another embodiment of the present invention includes an occupation cluster DB construction step (S100), a weighted keyword DB construction step (S200), and a learner data analysis step. It may include at least one of (S300) and the job and department recommendation step (S400).

In the occupation cluster DB construction step (S100), a weight DB (3) for each occupation is constructed by assigning weights to the occupation data of interest based on the learner's occupation preference, and the data stored in the occupation weight DB (3) is used to determine the learner's An occupational cluster DB (4) can be constructed by performing clustering based on occupational preference.

To this end, the occupation cluster DB construction step (S100) may include at least one of an occupation-specific weight generation step (S110) and an occupation cluster creation step (S120).

In the occupation-specific weight generation step (S110), web crawling is performed based on the occupation data of interest stored in the learner's interest occupation DB (1) to collect occupation data in the occupation DB (2), and the number of collected occupation data is calculated. It is regarded as the learner's job preference, and a weight DB (3) for each job can be constructed by assigning weights to each job of interest of the learner according to the job preference.

The learner's career interest DB (1) contains the career data selected by the learner as the career of interest among the career data recommended to each learner after the three-factor career aptitude test, and data obtained by running Codebook, a career and advancement education content service (codebook data). Occupation data (occupation data of interest) selected by the learner as an occupation of interest derived from analysis may be stored.

In the occupation-specific weight generation step (S110), various occupation data related to the learner's interest occupation data stored in the learner's interest occupation DB (1) is collected through web crawling and accumulated into the occupation DB (2), Classify and analyze the job data accumulated in the job DB (2) by type and series, set weights for the job data according to the learner's preference (occupation of interest), and create job data with weights reflected based on the learner's preference. It can be secured.

For example, as a basis for generating weights for each occupation, a weight of 1 point can be generated when the occupation of interest is accumulated 200 times. When targeting approximately 1,000 initial learners, the total number of career selections of interest is 3,900 times, the maximum number of specific career selections is 1,300 times, and the standard for generating a weight of 1 point can be set based on 200 times, which is about 15% of the 1,300 times. (The maximum number of times a specific occupation is selected as an occupation of interest is 1,300 times). Here, if a standard exceeding 15% is set, the initial barrier to creating weights is high, so it is not easy to secure weighted data, and if a standard less than 15% is set, weights are created for multiple occupations, making it difficult to obtain weighted data. The discriminatory power of star weights may be lowered.

In the occupation cluster creation step (S120), as a means of classifying similar groups based on the learner's preference and clustering them into clusters, the data stored in the occupation-specific weight DB (3) is divided into K- An occupational cluster DB (4) can be constructed based on data obtained by classifying groups with similar occupational preferences through K-means clustering.

In the occupation cluster creation step (S120), it is a module that classifies data stored in the occupation-specific weight DB into similar groups based on preference and clusters them into clusters. Each cluster may correspond to a group of occupations by type with similar preferences. In this occupation cluster creation step (S120), the data in the occupation-specific weight DB (3) can be classified using the K-means clustering algorithm, which corresponds to a data object according to the K-means clustering algorithm. n occupation data can be divided into k groups less than or equal to n to form an occupation cluster.

For example, as a criterion for creating clusters by occupation, it can be clustered into five occupation clusters based on the weight of each occupation.

In this way, the process of dividing groups by the K-means clustering algorithm can be accomplished by minimizing the cost function, and in this process, the preference-based similarity between job data within the same group increases and the similarity with data objects in other groups increases. It decreases.

In addition, the K-means clustering algorithm can perform clustering by setting the sum of the squares of the distance between the center of each group and the data objects within the group as a cost function and updating the group membership of each data object in the direction of minimizing the value of this function. there is.

In the weighted keyword DB construction step (S200), input keywords included in the codebook data are collected by predefined career aptitude types, and clustering is performed based on the frequency of the collected input keywords to construct a keyword cluster DB (5). In addition, a weighted keyword DB (6) can be constructed by assigning weights according to frequency to the clusters stored in the keyword cluster DB (5).

To this end, the weighted keyword DB construction step (S200) may include at least one of a first input keyword collection step (S210), an input keyword cluster performing step (S220), and an input keyword weight generation step (S230).

In the first input keyword collection step (S210), as a module for collecting learner data (diagnostic test data, codebook data), as shown in FIG. 2, a keyword DB among input keywords including nouns and verbs included in the codebook data Input keywords that match the data stored in (7) can be collected and classified by the collected career aptitude types (occupational interest type, multiple intelligence type, occupational value type).

More specifically, through the three-factor career aptitude test, codebook data of learners classified by type can be secured and learners' input keywords can be collected according to frequency. For example, self-understanding capabilities among the codebook data of social learners. Words with high frequency in questions can be accumulated as social keywords.

In the input keyword cluster performance step (S220), the type and frequency of input keywords collected by career aptitude type through the first input keyword collection step (S210) are identified, and the input keywords are clustered according to the identified type and frequency. You can build a keyword cluster DB (5) by performing .

In the input keyword weight generation step (S230), the weighted keyword DB (6) can be constructed by assigning a weight to each keyword cluster according to the frequency of the keyword cluster stored in the keyword cluster DB (5). In other words, weights can be assigned to keywords formed in high-frequency clusters. For example, weights of 1 to 10 can be created, and weights of 10 to 1 can be assigned in the order of clusters of high-frequency attributes. there is.

In the learner data analysis step (S300), among the input keywords included in the individual learner's codebook data, input keywords that match the keywords included in the weighted keyword DB (6) (or keyword DB (7)) are collected, and the collected Depending on the frequency of input keywords, the learner's career aptitude type data can be derived.

Here, the keyword DB (7) is a DB that classifies key keywords according to each type derived as a result of the career aptitude three-factor test (occupational interest, multiple intelligence, and occupational values test) as shown in Figure 3, and is classified into occupation and department. Data classified by type can be stored, including nouns and verbs.

This learner data analysis step (S300) may include at least one of a second input keyword collection step (S310) and a learner type analysis step (S320).

In the second input keyword collection step (S310), input matching the keyword included in the weighted keyword DB (6) (or keyword DB (7)) among input keywords including nouns and verbs included in the individual learner's codebook data. You can collect keywords.

In the learner type analysis step (S320), the learner's career aptitude type data can be derived according to the frequency of the input keywords collected through the second input keyword collection step (S310). In other words, the learner's job interest, multiple intelligence, and job value type (learner's career aptitude type data) can be derived based on the frequency of the collected input keywords.

In the job and department recommendation step (S400), job clusters containing career aptitude type data are recommended among job clusters stored in the job cluster DB (4), and jobs belonging to the job cluster with the highest weight among the recommended job clusters are recommended. Department data can be finally recommended.

To this end, the job and department recommendation step (S400) may include at least one of a conflicting interest type removal step (S410), a cluster-based job recommendation step (S420), and a final job and department recommendation step (S430).

In the step of removing conflicting interest types (S410), jobs of conflicting interest types are extracted from the career aptitude type data (learner's job interest, multiple intelligence, job value type data) derived through the learner data analysis step (S300), The extracted conflicting interest type occupation data can be removed.

In the step of removing conflicting interest types (S410), job recommendation accuracy can be increased by removing jobs with conflicting interest types that conflict with the learner's interest type (can be confirmed in career aptitude type data), and job recommendation accuracy can be increased by removing conflicting interest types. This can prevent the accuracy of recommendations from being reduced or from recommending a large number of meaningless jobs.

For example, if the learner's interest type is field-type, the social-type occupation, which is the interest type with the lowest consistency with field-type, can be excluded. In this embodiment, the conflicting interest type refers to an interest type with a low level of consistency among Holland's occupational interest types.

In the cluster-based job recommendation step (S420), among the job clusters stored in the job cluster DB 4, an job cluster including career aptitude type data from which conflicting interest type job data has been removed may be recommended. In other words, the intersection of occupations included in the learner's type cluster (occupational interest type cluster, multiple intelligence type cluster, occupational value type cluster) in the occupational cluster DB (4) can be recommended.

In the final job and department recommendation step (S430), as shown in FIG. 4, job clusters (occupational interest type cluster, multiple intelligence type cluster, occupational value type cluster) recommended through the cluster-based job recommendation step (S420) are selected. When an occupation of interest (or series of interest) is selected by an individual learner, a preset weight is given to the occupation selected as the occupation of interest (or series of interest), and then the occupations belonging to the intersection of the direct cluster with the highest weight among the occupation clusters are selected. Department data can be finally recommended.

In addition, in the final job and department recommendation step (S430), if the job of interest is not selected as shown in Figure 4, the career aptitude diagnostic test is performed again, and if the results are different from the previous career aptitude type data, , among the recommended occupation clusters, the occupation and department data belonging to the intersection of the occupation cluster with the highest weight can be finally recommended.

In addition, in the final job and department recommendation step (S430), if the job of interest is not selected as shown in FIG. 4, the career aptitude diagnostic test is performed again, and if the result is the same as the previous career aptitude type data, , the occupation and department data belonging to the intersection of occupation clusters with suboptimal weights can be finally recommended.

What has been described above is only one embodiment for implementing the information processing device and its operating method that operates in conjunction with the job and department recommendation program using artificial intelligence according to the present invention, and the present invention is limited to the above embodiment. Without departing from the gist of the invention, as claimed in the following patent claims, anyone with ordinary knowledge in the field to which the invention pertains believes that the technical spirit of the present invention is applicable to the extent that various modifications can be made. something to do.

1000: Job and department recommendation device using artificial intelligence
100: Job cluster DB construction department
110: Weight generation unit for each occupation
120: Job cluster creation unit
200: Weighted keyword DB construction unit
210: First input keyword collection unit
220: Input keyword cluster execution unit
230: Input keyword weight generation unit
300: Learner data analysis unit
310: Second input keyword collection unit
320: Learner type analysis unit
400: Job and Department Recommendation Department
410: Conflict interest type removal unit
420: Cluster-based job recommendation unit
430: Final job and department recommendation section
S1000: Job and department recommendation method using artificial intelligence
S100: Job cluster DB construction stage
S110: Weight generation step for each occupation
S120: Job cluster creation step
S200: Weighted keyword DB construction step
S210: First input keyword collection step
S220: Input keyword cluster execution step
S230: Input keyword weight generation step
S300: Learner data analysis step
S310: Second input keyword collection step
S320: Learner type analysis step
S400: Job and Department Recommendation Stage
S410: Opposition interest type removal step
S420: Cluster-based job recommendation step
S430: Final job and department recommendation stage
1: Student interest occupation database
2: Occupation DB
3: Weight DB for each occupation
4: Occupation cluster DB
5: Keyword Cluster DB
6: Weighted keyword DB
7: Keyword DB

Claims (10)

Individual learner's occupation and department based on codebook data including job interest data selected by the learner among job data temporarily recommended through career aptitude diagnostic tests by type, and keywords entered by the learner in the process of providing career/advanced education content. It relates to an information processing device that operates in conjunction with a program that recommends,
Constructing an occupation-specific weight DB by assigning weights to the occupation data of interest based on the learner's occupation preference, and constructing an occupation cluster DB by clustering the data stored in the occupation-specific weight DB based on the learner's occupation preference. Occupational Cluster DB Construction Department;
Input keywords included in the codebook data are collected by predefined career aptitude types, clustering is performed based on the frequency of the collected input keywords to construct a keyword cluster DB, and the clusters stored in the keyword cluster DB are classified according to the frequency. A weighted keyword DB construction unit that constructs a weighted keyword DB by assigning weights;
A learner data analysis unit that collects input keywords that match keywords included in the weighted keyword DB among the input keywords included in the codebook data of individual learners, and derives career aptitude type data of the learner according to the frequency of the collected input keywords; and
It recommends job clusters containing the career aptitude type data among the job clusters stored in the job cluster DB, and includes a job and department recommendation section that finally recommends job and department data belonging to the job cluster with the highest weight among the recommended job clusters. An information processing device that operates in conjunction with a job and department recommendation program using artificial intelligence.
According to claim 1,
The occupational cluster DB construction department,
Collect job data by performing web crawling based on the job data of interest, consider the number of collected job data as the learner's job preference, and assign weights to each job of interest of the learner according to the job preference. Weight generation unit for each occupation that builds a weight DB; and
An artificial job cluster generator, which constructs the job cluster DB by classifying the data stored in the job-specific weight DB into groups with similar job preferences through K-means clustering based on the learner's job preference. An information processing device that operates in conjunction with job and department recommendation programs that utilize intelligence.
According to claim 1,
The weighted keyword DB construction unit,
A first input keyword collection unit that collects input keywords including nouns and verbs included in the codebook data by predefined career aptitude types;
An input keyword cluster performing unit that determines the frequency of input keywords collected by career aptitude type through the first input keyword collection unit and constructs a keyword cluster DB by performing clustering on the input keywords according to the identified frequency; and
In conjunction with a job and department recommendation program using artificial intelligence, it includes an input keyword weight generator that builds a weighted keyword DB by assigning a weight to each keyword cluster according to the frequency of the keyword cluster stored in the keyword cluster DB. A working information processing device.
According to claim 1,
The learner data analysis unit,
a second input keyword collection unit that collects input keywords that match keywords included in the weighted keyword DB among input keywords including nouns and verbs included in the individual learner's codebook data; and
It operates in conjunction with a job and department recommendation program using artificial intelligence, comprising a learner type analysis unit that derives career aptitude type data of the learner according to the frequency of the input keywords collected through the second input keyword collection unit. Information processing device.
According to claim 1,
The above job and department recommendation department is,
a conflicting interest type removal unit that extracts an occupation of a conflicting interest type from the career aptitude type data derived through the learner data analysis unit and removes the extracted conflicting interest type occupation data;
a cluster-based job recommendation unit that recommends a job cluster including career aptitude type data from which the conflicting interest type job data has been removed among job clusters stored in the job cluster DB; and
When an individual learner selects an occupation of interest for an occupation cluster recommended through the cluster-based job recommendation unit, a preset weight is given to the occupation selected as the occupation of interest, and then the occupation and department data with the highest weight among the occupation clusters are finalized. If you do not receive a recommendation or selection of an occupation of interest, or if the career aptitude test is retaken and the results are different from the previous career aptitude type data, the occupation and department data belonging to the occupation cluster with the highest weight among the recommended job clusters If the result is the same as the previous career aptitude type data after performing the career aptitude diagnostic test again, a final job and department recommendation section that finally recommends the job and department data belonging to the job cluster with the second best weight is included. An information processing device that operates in conjunction with a job and department recommendation program using artificial intelligence.
Individual learner's occupation and department based on codebook data including job interest data selected by the learner among job data temporarily recommended through career aptitude diagnostic tests by type, and keywords entered by the learner in the process of providing career/advanced education content. It relates to a method of operating an information processing device that links with a recommended program,
Constructing an occupation-specific weight DB by assigning weights to the occupation data of interest based on the learner's occupation preference, and constructing an occupation cluster DB by clustering the data stored in the occupation-specific weight DB based on the learner's occupation preference. Job cluster DB construction stage;
Input keywords included in the codebook data are collected by predefined career aptitude types, clustering is performed based on the frequency of the collected input keywords to construct a keyword cluster DB, and the clusters stored in the keyword cluster DB are classified according to the frequency. A weighted keyword DB construction step of constructing a weighted keyword DB by assigning weights;
A learner data analysis step of collecting input keywords that match keywords included in the weighted keyword DB among the input keywords included in the individual learner's codebook data, and deriving the learner's career aptitude type data according to the frequency of the collected input keywords; and
Among the job clusters stored in the job cluster DB, the job cluster that includes the career aptitude type data is recommended, and the job and department data belonging to the job cluster with the highest weight among the recommended job clusters are finally recommended. A method of operating an information processing device linked to a job and department recommendation program using artificial intelligence, comprising:
According to clause 6,
The job cluster DB construction step is,
Collect job data by performing web crawling based on the job data of interest, consider the number of collected job data as the learner's job preference, and assign weights to each job of interest of the learner according to the job preference. Weight generation step for each occupation to build a weight DB; and
Characterized by an occupation cluster creation step of constructing the occupation cluster DB by classifying the data stored in the occupation-specific weight DB into groups with similar occupation preferences through K-means clustering based on the learner's occupation preference. Method of operating an information processing device that links with a job and department recommendation program using artificial intelligence.
According to clause 6,
The weighted keyword DB construction step is,
A first input keyword collection step of collecting input keywords including nouns and verbs included in the codebook data for each predefined career aptitude type;
An input keyword clustering step of determining the frequency of input keywords collected by career aptitude type through the first input keyword collection step and constructing a keyword cluster DB by performing clustering on the input keywords according to the identified frequency; and
Linked with a job and department recommendation program using artificial intelligence, comprising an input keyword weight generation step of constructing a weighted keyword DB by assigning a weight to each keyword cluster according to the frequency of the keyword cluster stored in the keyword cluster DB. A method of operating an information processing device.
According to clause 6,
The learner data analysis step is,
A second input keyword collection step of collecting input keywords that match keywords included in the weighted keyword DB among input keywords including nouns and verbs included in the individual learner's codebook data; and
Linked with a job and department recommendation program using artificial intelligence, comprising a learner type analysis step of deriving career aptitude type data of the learner according to the frequency of the input keywords collected through the second input keyword collection step. How an information processing device operates.
According to clause 6,
The job and department recommendation steps are:
A conflicting interest type removal step of extracting an occupation of a conflicting interest type from the career aptitude type data derived through the learner data analysis step and removing the extracted conflicting interest type occupation data;
A cluster-based job recommendation step of recommending a job cluster including career aptitude type data from which the conflicting interest type job data has been removed among job clusters stored in the job cluster DB; and
When an occupation of interest is selected by an individual learner for an occupation cluster recommended through the cluster-based job recommendation step, a preset weight is given to the occupation selected as the occupation of interest, and then the occupation and department data with the highest weight among the occupation clusters are selected. If you are not selected for the final recommendation or job of interest, or if you take the career aptitude test again and the results are different from the previous career aptitude type data, the job and department that belong to the job cluster with the highest weight among the recommended job clusters. After the data is finally recommended, the career aptitude diagnostic test is performed again, and the results are the same as the previous career aptitude type data, the final job and department recommendation step is performed to finally recommend the job and department data belonging to the job cluster with the next best weight. A method of operating an information processing device linked to a job and department recommendation program using artificial intelligence, comprising: