TYPE Original Research

PUBLISHED 22 July 2022

DOI 10.3389/fpsyg.2022.962291

Artificial intelligence-based
OPEN ACCESS organizational human resource
management and operation
EDITED BY
Sang-Bing Tsai,
Wuyi University, China

REVIEWED BY
Xuetao Li,
system
Hubei University of Automotive
Technology, China
Jinyan Lin,
Yang Yang *
Yango University,
School of Management and Economics, North China University of Water Resources and Electric
China
Power, Zhengzhou, China
*CORRESPONDENCE
Yang Yang
[email protected] The trend of globalization, marketization, and informatization continues to
SPECIALTY SECTION strengthen, in today’s development environment, how to seize the opportunity
This article was submitted to and obtain a competitive advantage in human resources is an important issue
Organizational Psychology,
a section of the journal that needs to be explored. Human resource management refers to the effective
Frontiers in Psychology use of relevant human resources inside and outside the organization through
RECEIVED 06 June 2022 management forms under the guidance of economics and humanistic thinking.
ACCEPTED 05 July 2022 It is a general term for a series of activities that ensure the achievement of
PUBLISHED 22 July 2022
organizational goals and the maximization of member development. With
CITATION
Yang Y (2022) Artificial intelligence-based
the rapid development of society and economy, the competition between
organizational human resource enterprises has intensified. If an enterprise wants to adapt to social development,
management and operation system. it is necessary to strengthen the internal management of the organization. The
Front. Psychol. 13:962291.
doi: 10.3389/fpsyg.2022.962291 internal management also needs to rely on human resource management. The
COPYRIGHT
purpose of this paper is to study an organization’s human resource management
© 2022 Yang. This is an open-access article and operation system based on artificial intelligence. It expects to use artificial
distributed under the terms of the Creative intelligence technology to design the human resource management system
Commons Attribution License (CC BY). The
use, distribution or reproduction in other and to improve the quality of employees to make the enterprise develop
forums is permitted, provided the original toward a more scientific and reasonable method. It uses artificial intelligence
author(s) and the copyright owner(s) are
technology to mine the relevant data of enterprises, understand the situation
credited and that the original publication in
this journal is cited, in accordance with of enterprises in a timely manner, and adjust unreasonable rules. This paper
accepted academic practice. No use, establishes a dynamic capability evaluation model and an early warning model
distribution or reproduction is permitted
which does not comply with these terms. for human resource management and further studies the improvement
approach based on human resource management. This paper analyzes the
application, feasibility, and practical significance of data mining technology
in human resource management systems. It focuses on the commonly
used algorithms in the field of data mining and proposes specific algorithm
application scenarios and implementation ideas combined with the needs of
human resource management practices. The experimental results of this paper
show that the average working life of incumbent employees is 3.5years, the
average length of employees who leave the company is 5years, and some
employees are 5–6years old. From this data, it can be seen that the average
number of years of on-the-job employees is short, and the work experience has
yet to be accumulated.

KEYWORDS

human resources, artificial intelligence, operational systems, management

operations, data mining

Frontiers in Psychology 01 frontiersin.org

Yang 10.3389/fpsyg.2022.962291

Introduction than those implemented separately. HRM bundles associated

with merit control and appraisal, learning, exchange, and
The rapid development of knowledge economy has broken the decision-making, involvement, and delegation were all found
barriers of information transmission and brought new business to be actively related to creativity in both firms in the
ideas and methods at the same time, and the change of thinking manufacturing and service sectors. In conclusion, HRM
mode has made people more and more aware of the importance practices are important for company innovation when applied
of human resources. It has been proved by practice that the together rather than in isolation (Crowley and Bourke, 2017).
production of human resources plays a decisive role in the Sandra drew on a resource-based perspective and used
development of production and guarantees the implementation of suppositions about a company’s CSP strengths from an objective
the organization’s business strategy. With the continuous progress third-party database. The study results show that human
of society, human resource management faces many problems, resource management and creativity are key capabilities because
including the challenges of structural diversification and they enable the creation and enhancement of other capabilities
competition diversification. In the era of knowledge economy, the (Rothenberg et al., 2017). Organizational learning is becoming
advanced western management thoughts have a strong impact on increasingly important for strategic updating. Agile
the traditional management knowledge system, and the organizations are especially successful in the current
importance of organizational human resources has made its status environment, where companies need to be efficient and adapt
in organizational management increasingly promoted. These are to change. Organizational Learning Organizational learning
the reasons for adopting new methods. Studying the dynamic refers to the various actions that organizations take around
capabilities of human resources can provide an effective method information and knowledge skills to achieve development goals
for human resource management to judge the effectiveness of and improve core competitiveness. Diaz-Fernandez took a
organizational management behavior and job suitability standards, constructive view. He emphasized the distinction in terms of
and it provides intellectual support for enterprises. The use of AI what is learned between sales and delivery units in the case of
is a way to help companies dig deeper into the treasures of HRM different HRM measures of risk and human energy profiles. The
information and improve their efficiency and core competitiveness. experimental results point to a mediating role of human capital
Modern management puts forward higher requirements for between HRM practices and learning (Diaz-Fernandez et al.,
enterprise human resource management mode, so it is very 2017). Although these theories discuss human resources, they
suitable to implement computerization of enterprise human do not use artificial intelligence technology.
resource information processing. AI is a hot technology nowadays, and there are many studies
The tremendous advances in information and on it. Burton wanted to develop a curriculum to introduce
communication technology (ICT) and the emergence of the AI. He offered specific suggestions for integrating AI into the
knowledge economy pose great opportunities for companies curriculum and how to teach a stand-alone course on the ethics
to meet their business objectives. Majumder focused on of AI (Burton et al., 2017). Bin provides a conceptual framework
an in-depth understanding of current human resource for integrating statistical ideas with human input into AI products
management practices and their impact on employee and research. These ideas include the principles of experimental
satisfaction. For the study, he selected 100 bank employees, of design for randomization and local control. He discussed the
which 88 employees responded correctly, with a response rate connection of these principles in self-driving cars and automated
of 88%, and the questionnaire was compiled using a point scale. medical diagnostics (Bin and Kumbier, 2018). Polina outlined the
Point scale (point scale), also known as “point scale,” is a test next generation of artificial intelligence and blockchain
scale with points as the scoring unit. Its advantages are that the technologies. It enables patients to control and monetize their
scoring is accurate and simple. Research shows that employees personal data through new tools, as well as incentives for
are not equally satisfied across all human resource management continuous health monitoring (Polina et al., 2018). Organizational
dimensions. Most employees were dissatisfied with Learning Organizational learning refers to the various actions that
compensation packages, followed by rewards and incentives, organizations take around information and knowledge skills to
career development, training and development, management achieve development goals and improve core competitiveness.
style, job design, and responsibilities. Therefore, the quality of Although these theories have expounded artificial intelligence,
these human resource management dimensions should their integration with human resource management is
be improved (Majumder, 2017). Guest proposed an approach to less practical.
human resource management that prioritizes well-being and With the progress of the times, people pay increasingly
positivity. He provided evidence to support practice choices and attention to resources. In the current fierce competition
argued that these practices also have the potential to improve environment, companies must strengthen the management of
individual and organizational performance (Guest, 2017). employees within the organization if they want to stay at the
Crowley investigated to see whether HRM practices forefront of the industry. Experiments show that the incumbency
implemented as complementary ‘bundles’ or ‘systems’ in Irish rate of management positions is as high as 80%, the turnover rate
manufacturing and service sector companies are more effective is only 20%, and the incumbency rate of market positions is 71%.

Frontiers in Psychology 02 frontiersin.org

Yang 10.3389/fpsyg.2022.962291

It shows that the turnover of different positions is different; the performance appraisal, performance communication and
incumbency rate of low salary is 67%, the incumbency rate of feedback, and performance improvement.
general salary is 79%, and the incumbency rate of high salary is The property owner is separated from the laborer, and the
88%. It shows that salary is an important factor affecting remuneration paid by the owner to the laborer is called salary, and
employees, and retaining employees can start from salary. The we now call it the value created (Shivam et al., 2020). With the
turnover rate without promotion within 3 years is 25%. The progress of the times, people are increasingly concerned about
turnover rate of job promotion is 10%, indicating that long-term things other than external compensation, that is, inner feelings,
lack of promotion is an important factor leading to which we call intrinsic compensation. This is an important reason
employee turnover. for the internal imbalance of employees (Belizon et al., 2017).
External remuneration refers to salary bonuses, allowances and
subsidies, insurance benefits, etc., including monetary and
Human resource management nonmonetary returns. We can think of this kind of compensation
operation system method as human needs. According to Maslow’s Hierarchy of Needs,
we can divide it into different categories, and its distribution
Human resource management pattern is shown in Figure 1.
The continuous development of enterprises is inseparable
With the development of the times, the connotation of from the progress of employees. Compared with recruiting new
human resources continues to expand. However, “human employees, training existing employees in line with the
resources” was first proposed in the middle of the last century. company’s development strategy is one of the lowest-cost and
At that time, scholars believed that human resources were a fastest effective human resource management measures.
special kind of resources, which must be stimulated by relevant According to the actual situation, it is also verified that training
incentive mechanisms, thus bringing their due value (Grzonka and development is an important means to improve the
et al., 2017; Lee et al., 2017a). Enterprises believe that human comprehensive quality of employees. Corporate training enables
resource management refers to the deployment requirements employees to quickly understand the corporate culture and have
and macro directions according to the enterprise development a general understanding of the company’s work. In this process,
strategy. It fully mobilizes the enthusiasm of employees to work, the company can also conduct a comprehensive evaluation of
stimulates the potential of employees, and finally achieves the employees and match employees to jobs. Conventional
strategic goals of enterprise development. People are the active enterprise training management includes a four-step process of
force of social development. If people cannot develop their training needs analysis, training plan formulation, organization
knowledge and skills, the development of all new things will and implementation, training effect evaluation, and feedback.
be hindered. Human resource management has the ability to Development is not the same as training. Development needs to
predict the human resources of the organization screen improve the creativity and skills of employees, coordinate the
personnel according to the predicted conditions, integrate personal goals of employees and corporate goals, and enable
compensation, and optimize the construction of the entire employees to develop healthily and quickly in the enterprise.
organization (Lee et al., 2017b; Youssef et al., 2017). Due to the hysteresis of the training effect and the differences
Organization optimization is the planner of the overall layout and personalization of individual employees in their work,
of the enterprise; organization optimization is the founder of scientific and objective evaluation of the training effect is the
enterprise system; organization optimization is the promoter of most difficult part of the training management process (Cooke,
enterprise reform; organization optimization is the cultivator of 2017; Ones et al., 2017). Figure 2 shows the enterprise talent
enterprise cultural heritage. development planning model.
There is a lot of content in human resource management, and
here we briefly describe performance compensation and training
(Labovitz et al., 2017). Performance refers to the stage results Self-
achieved by employees according to their duties and the behavioral realization Spirituality
performance in the process of achieving the stage results. What
Respect
we call performance management is a series of measures taken to
achieve performance (Agrawal et al., 2017; Price and Flach, 2017). Social Matter
Performance management forms an objective and fair closed-loop
management mechanism to motivate employees to make work Security
behaviors that are beneficial to the development of the enterprise,
to achieve organizational goals and achieve closer integration of Physiology

organizational and employee performance (Ghoson and Faris, FIGURE 1

2021). The performance management system mainly includes Distribution of Maslow’s hierarchical model of needs.
performance planning, including performance implementation,

Frontiers in Psychology 03 frontiersin.org

Yang 10.3389/fpsyg.2022.962291

Brief introduction of artificial intelligence operation function used to measure the difference between the
predicted value f(x) of the model and the real value Y. It is a
Data mining is a continuous process. When using this nonnegative real-valued function, usually represented by L [Y,
method, it is first necessary to determine the mining objects and f(x)]. The loss function, the smaller it is, the more robust the
targets. In the process of data mining, we need to classify the model is.
objects according to the rules. The theme of this paper is human Combined with human resource management, we can use the
resource management, so we can subdivide the goals into linear regression method to establish a human resource demand
employee-related elements such as performance and turnover in forecast model. The regression algorithm has many data features,
the data mining process, and then select the appropriate target set and it is an algorithm used to evaluate the relationship between
according to the relevant target groups. Figure 3 shows the the dependent variable and multiple independent variables. In a
relevant flow chart of data mining. human resource management system, the emergence of a decision
The multiple regression algorithm belongs to the primary will affect multiple factors. However, the algorithm needs to
artificial intelligence algorithm, and its principle is very simple, so determine whether there is a certain linear correlation between
it is very widely used in practical applications. In regression the influencing factors of human resources and the prediction
analysis, if there are two or more independent variables, it is called target in the actual application process.
multiple regression. In fact, a phenomenon is often associated From an enterprise’s point of view, human resource
with multiple factors, so it is more effective to predict or estimate management is mainly determined through experience and needs,
with multiple independent variables than with only one but its basis is a relatively subjective element, and there are no
independent variable. We briefly introduce it. specific rules to follow. Through the linear regression algorithm,
the relevant factors that affect the human resource demand of the
gε ( a ) = ε 0a0 + ε1a1 + ε 2 a2 +  + ε k ak (1) enterprise can be accurately found, to help the enterprise manager
to more clearly define the amount of human resources required by
the enterprise, make a more scientific and reasonable allocation
In formula (1), gε ( a ) represents the multiple regression method, and lay a good human resource for the development of
value, k represents its dimension, and ε k ak represents the the enterprise. Base.
sample set. K-means algorithm belongs to the clustering algorithm and is
widely used in the field of data mining. We can also use it in the
gε ( A ) = ε ∗ A (2) human resource management module to establish a performance
appraisal model. It helps enterprises to construct the correct
decision-making method, realize the objectivity of assessment,
In formula (2), A represents a dimension matrix. and realize scientific management.

Y ( ε 0 ,ε1,ε 2 ,,ε k ) = ∑ ( gε ( a0 ,a1,a2 ,,ak ) − b p ) (3) ϑ (a ∪ b)

 Support = ( a,b ) = Q ( a,b ) = (4)
k

In formula (3), we take the mean difference as the loss

ϑ (a ∪ b)
function and then fit it by minimizing it. The loss function is an Confidence = ( a,b ) = Q ( a,b a ) = (5)
ϑ (a )

In the above function expression, a , b represents two

different events, formula (4) represents support, and formula (5)
Demand
Management Plan represents confidence.
Strategy
Usually, we apply the K-means algorithm to the employee
Blending
performance appraisal management module of the human
Planning
resource management system. Realization needs to select
Quality indicators related to employee performance appraisal
Talent Improvement
Environment according to the actual situation of the enterprise. Such
indicators as innovation and team capability are then assigned
points based on their importance. The total score of the
Dismissal
indicator represents the overall performance level of the
Supply
$ $
employee. The construction of this employee performance
FIGURE 2
appraisal model can help enterprises to abandon the traditional
Enterprise talent development planning model. subjective decision-making method and realize the objectivity
of employee appraisal and evaluation. It is an important

Frontiers in Psychology 04 frontiersin.org

Yang 10.3389/fpsyg.2022.962291

preproce model
collect application
ssing selection Evaluate
FIGURE 3
Relevant flow chart of data mining.

decision-making model. Through the algorithms, we can mine the

matching degree between the current position of the company’s
0.5 employees and other positions, so as to provide scientific matching
information when the company’s personnel changes, and its
model structure is shown in Figure 5.

0.3
Human resource management
operation system experiment
System structure design

Database design is an important part of enterprise human

0 resource probabilistic system design. In the database design, it is
FIGURE 4
necessary to carefully study the specification documents and
Common models of conditional probability. design documents, clarify the content of the database design, and
then select the appropriate database management system
according to the content. E-R is a diagram used to model data and
identify the relationship between functions and data. Figure 6 is
auxiliary tool for modern enterprises to achieve the E-R diagram of the human resource management system.
scientific management. Employee information needs to be recorded in the system.
The Naive Bayes algorithm is a classification method based on According to the system settings in Table 1, the employee
the Yebes theorem and the assumption of independence of feature number is generally set to a char type with a length of 10 digits.
conditions, and it also belongs to a class of data mining algorithms. This number will be given when the employee joins the job. This
number is unique and cannot be changed later. The name is set
F ( R )( a,b ) > F ( t )( a,b ) (6) to a variable-length char type, and the length can be up to 20
characters. Gender settings can input 2-bit length char types,
In the above function expressions, F ( R ) and F ( t ) are not and some information settings can input 12-bit length char
the probability functions we often say, they refer to Bayesian types. The first and second digits are the number of the
probability, which is what we often say the posterior probability. department, the third digit is the incremental post code of the
Figure 4 shows a common model of conditional probability. department, and the fourth digit is the number of employees in
In practical applications, for a given training set, we need to this position. In fact, the employee number is the most
independently assume the joint distribution probability according important among these information, and the employee
to the feature conditions, and then calculate the posterior information can be searched according to the number in the
probability according to the model. When managing the human system. Therefore, in the data registration form, except for the
resource structure based on the employee personal information in employee number, which is not allowed to be empty, all other
the human resource management system, we first need to select information can be empty.
representative employee characteristics, set the minimum support
and confidence for mining, and discover the implicit strong
correlation in the data. For example, dig out the unreasonable Database logic design
aspects of the degree structure and professional title structure of
the enterprise, determine some problems in the current workforce This article creates each table in the database according to the
structure, and realize the optimization of the workforce. In designed entity E-R diagram. It establishes different data tables
addition, we can make promotion decisions based on this according to the relevant information required by the enterprise

Frontiers in Psychology 05 frontiersin.org

Yang 10.3389/fpsyg.2022.962291

employee
information
send

Promotion manager
information algorithm Analysis

FIGURE 5
Employee promotion phase algorithm-related model.

According to the E-R diagram, we also divide it into three

categories for introduction in the attendance table. The first is the
sign in
ID of the employee. The ID is generally set to a Bigint type with
a length of 8 bits. The ID here is equivalent to positioning. When
employees use the system to punch in, they will automatically
locate, so the ID will change, which is also a way for companies
Attendance
to check employees’ attendance. The employee number is
generally set to a 5-digit VarChat type, which is unique to each
employee and cannot be changed by anyone. The most important
thing about attendance is to clock in and clock out. They generally
Salary set a VarChat type with a length of 5 digits. After using the system
Employee
to punch in, relevant information will be automatically generated
score to help the personnel department understand the attendance of
employees. An important part of this structure is the remarks,
assessment
this part is to solve the problem of other attendance situations,
proportional field work, or the problem of not being able to clock
FIGURE 6
E-R Diagram of human resource management system. in because of enterprise network problems. These situations can
be remarked to address attendance issues.
According to the data in Table 4, we have systematically
to investigate and analyze the relevant situation of employees. The designed the salary situation of enterprise employees. According
more important data table structure is given below. to the E-R diagram, time is an essential part of salary design. The
According to the data in Table 2, we briefly introduce some month is generally set as a VarChat type with a length of 20 digits,
relevant information and structures. According to the E-R diagram, which represents the salary of the employee in the month. If there
the department number is generally set to a varChat type with a are other related salaries that were not paid in full last month, they
length of 5 digits. Because the enterprise has different departments, can be remarked in this section. Next is the number and name of
a fixed number is given when the department information enters the employee. This part is the same as the introduction above, and
the system, and the number is fixed and no one has the right to the employee number is unique. There are allowances and other
change it. The name of the department is generally set to a varChat types in the employee’s salary component, including leave, which
type with a length of 10 digits. The name and number are different. involves bonuses and deductions other than the basic salary, and
The number cannot be changed, but the name can be changed. As these contents are recorded in the corresponding parts of the
long as it is within a reasonable character, it can be changed. The system. These contents and the final paid salary are generally set
manager-related information of the department is generally set to to a 5-digit money type, and all information can be searched by
a varChat type with a length of 20 digits, and the manager’s area of raising the employee number in the salary part of the system.
responsibility can be briefly described in the system. The address
of the department is generally set to a varChat type with a length
of 20 digits, and the telephone of the department is generally set to Human resource management
a varChat type with a length of 50 digits. According to the above operation system
department information table, the number of the department is the
most important, and the relevant information of the department Employee indicator data
can be obtained in the system according to the department number.
According to the data in Table 3, we have systematically In enterprise management, the management needs to analyze
designed the attendance of employees of the enterprise. the basic situation of employees. In the process of analysis, it is

Frontiers in Psychology 06 frontiersin.org

Yang 10.3389/fpsyg.2022.962291

TABLE 1 Basic information of employees.

Category Employee Employee name Gender Department Phone number Remarks

number
Type Int Char Char Char Char varChar
Size 8 10 2 12 12 50
Properties No Yes Yes yes yes yes

TABLE 2 Departmental table related structure. the perspective of employee turnover, we conducted a survey
on the company’s turnover within a week. According to the
Name Type Size
survey data, the number of departures on Saturday is the
Number varChat 5 largest, and the number of departures on Wednesday is the
Name varChat 10 least. However, employee turnover occurs every day, and most
Manager varChat 20 on weekends. This may be related to the reality that most
Address varChat 20 employees choose to leave on the weekend may have the basic
Phone number varChat 50 work at hand completed. And it is free to schedule an interview
next week.
Judging from the number of incumbent employees and new
TABLE 3 Attendance sheet related structure.
employees in the company, the number of employees in the first
Name Type Size 2 days of the week is relatively small, but the overall number
ID Bigint 8 fluctuates little, which belongs to the normal flow of personnel.
Number varChat 5 As the number of interviewers increases, the total number of
Name varChat 10 people in the company will continue to increase. According to
Commuting status varChat 5 the survey data, the company has 330 employees on Saturday, the
Off duty status Datetime 5 largest number in a week. Judging from the number of new
Date Datetime 8 employees, the number of new employees on Friday was the
Remarks varChat 10 least, with only 19, and the most on Monday was 35. This
situation is basically consistent with the actual situation,
indicating that the records of the human resource management
TABLE 4 Structure related to payroll. system are consistent with the facts, and we can use the data
provided by the system to analyze the situation of the employees
Name Type Size
of the enterprise.
Month varChat 20 Employees are the most important resource of an enterprise.
Number varChat 5 How to retain employees is essential to improve the development
Name varChat 10 of the company. To this end, we analyzed the relationship between
Basic salary Money 5 employee turnover and working years. According to the data in
Allowance Money 5 Figure 8, we have investigated the working years of the current
Leave of absence Money 5 and former employees of the company. First of all, we observe the
Deductions Money 5 on-the-job employees. The average working life of on-the-job
Actual salary Money 5 employees is 3.5 years, and half of them have less than 2 years.
From this data, it can be seen that the average number of years of
on-the-job employees is short, and the work experience has yet to
mainly to investigate the relevant data indicators of employees. be accumulated.
Enterprises include different departments, and employees may Judging from the situation of resigned employees, the
leave or newly join at different stages. We can intuitively show average age of resigned employees reached 5 years. Some of the
them according to the relevant functions of the system. The employees are 4–5 years old, and some employees are 5–6 years
management can also analyze the management situation of the old. According to this data, we know that most of the employees
current department based on these data, so that the enterprise can who leave are old employees. The “runaway” of these people is
develop in a more scientific direction. a large loss for the company. According to this situation, at
We briefly introduced the relevant data of employees above. present, the company’s labor turnover is relatively large, the
In Figure 7, we took Company A as an example to investigate working years of new employees are not enough, and the
the employee turnover of the formula. According to the data in number of old employees is decreasing. Management needs to
the data classification, we divided it into resigned employees, adjust company policies under objective conditions to retain
incumbent employees, and new employees. First of all, from employees. The situation of employees can be displayed very

Frontiers in Psychology 07 frontiersin.org

Yang 10.3389/fpsyg.2022.962291

Number of people
30

25

20

15

10

0
Monday Tuesday Wed Thursday Friday Saturday Sunday

340 Number of people available Number of additions 40

330 35
320 30
310
25

Value
300
Value

20
290
280 15
270 10
260 5
250 0
Monday Tuesday Wed Thursday Friday Saturday Sunday
Group
FIGURE 7
Description of employee resignation and job hunting.

8
5
6

3.5
4
2

2
1 1
FIGURE 8
Employee working year analysis.

intuitively in the human resource management system. The Staff positions

management needs to analyze the relevant data in time, find
countermeasures, and ensure that the formula develops in the As a relatively sound enterprise, the company will set up
direction of health science. different positions to cooperate with the work. However, according

Frontiers in Psychology 08 frontiersin.org

Yang 10.3389/fpsyg.2022.962291

to the situation of the century, the treatment of different positions positions and sales positions, and marketing positions have the
is different, and the resignation of employees in different positions lowest incumbency rate at 71%. From the overall situation, the
is also different. To investigate the resignation situation of incumbency rate of each position is relatively high, indicating that
employees in different positions, we used the human resource there is no shortage of employees in Company A. However, this
management system to summarize the in-service and resignation data does not represent the quality of employees. It can be seen
situations of employees in different positions. The details are from the above that a large number of “runaways” of old employees
as follows. is the most deadly.
Different positions have different attractiveness to employees. From the perspective of salary, we conducted surveys from three
Some employees may like this job, and some people are looking dimensions: low salary, general salary, and high salary. According to
forward to this salary. According to the data in Figure 9, the data survey, the incumbency rate of low salary is 67%, and the
we conducted a survey from the aspects of position and salary. turnover rate is 33%; the incumbency rate of general salary is 79%,
We first analyze the situation from the position. According to the and the turnover rate is 21%; the incumbency rate of high salary is
survey, the incumbency rate of management positions is self-high, 88%, and the turnover rate is 12%. It can be seen that the turnover
reaching 80%, and the turnover rate is only 20%. It can be seen rate of low salary is the highest, and the turnover rate of high salary
that the management positions of the company A are the most is the lowest, indicating that salary is an important factor affecting
popular positions among all positions. It is followed by R&D employees, and retaining employees can start from salary.

On duty Separation
90
80
70
60
VALUE

50
40
30
20
10
0
IN

N
IL
FI

IN

O
M

RI

TI
TI

SK

LL
D

U
U
N
A

SE

IB
CT
IE

TR
SC

FA

IS
U

D
N
A
M

GROUP

On duty Separation
100

80

60
VALUE

40

20

0
L
R

ER
A
O

PP
RM
PO

U
O
N

GROUP
FIGURE 9
Analysis of employee positions.

Frontiers in Psychology 09 frontiersin.org

Yang 10.3389/fpsyg.2022.962291

On duty
120
Separation
100

80
Value

60

40

20

0
1 2 3 4 5 6 7
Group
On duty
100
Separation
80

60
Value

40

20

0
Not promoted promoted
Group
FIGURE 10
Analysis of employee promotions.

Employee promotion situation the turnover rate is 93%. It can be seen from this that employee
treatment is not the only factor that affects employees leaving
Employees need to seek progress in the enterprise, and if the job. It may be due to external competition leading to brain
employees have been unable to get promoted, they may leave. drain. In response to this situation, companies need to take
Because the inability to be promoted may lead to the inability early countermeasures to reduce the occurrence of
to raise wages, which will dampen the enthusiasm of employees this situation.
and be detrimental to the normal development of the company. From the perspective of employee promotion, according to
Therefore, enterprises need to formulate a scientific and the basic laws of the market, we conducted a survey on employee
reasonable job promotion system to stimulate the enthusiasm turnover in the experiment using one promotion every 3 years.
of employees. To investigate the promotion situation of According to the data, the incumbency rate of those who have not
employees, we conducted a survey on the operation projects been promoted within 3 years is 75%, and the turnover rate is
and promotion situation of employees, and the details are 25%; the incumbency rate of those who have been promoted
as follows. within 3 years is 90%, and the turnover rate is 10%. According to
Projects and promotions of employees were investigated. the data, employees who have not been promoted within 3 years
According to the data in Figure 10, first of all, looking at the have a higher chance of leaving, which also shows that long-term
projects handled by employees, the incumbency rate of lack of promotion is an important factor leading to
employees who only handle one project is 20%, and the employee turnover.
turnover rate is 80%. 65% of employees who have handled the
two projects have an on-the-job rate and a 35% turnover rate.
The employees who have handled the three projects have an Conclusion
on-the-job rate of 95% and a turnover rate of 5%. From this
situation, it can be seen that in the number of projects 1–3, the With the continuous impact of the economy and the
in-service rate of employees is increasing. In the number of environment, the competition of enterprises is becoming
projects 1–3, the on-the-job rate of employees is declining, increasingly fierce. If we want to maximize the benefits of
and the turnover rate of employees is rising. Especially in the enterprises by relying on high-tech means, we must closely follow
7 projects handled, the on-the-job rate of employees is 7%, and the needs of technological change and establish a new management

Frontiers in Psychology 10 frontiersin.org

Yang 10.3389/fpsyg.2022.962291

model. Therefore, this paper aims to study the organizational Author contributions
human resource management and operation system based on
artificial intelligence. It expects to use artificial intelligence YY has participated in conception and design, analysis and
technology to design human resource management systems. It interpretation of the data, drafting the article, and revising it
improves the quality of employees so that the enterprise develops critically for important intellectual content.
toward a more scientific and reasonable method. Experimental
investigations have found that compensation, performance, and
promotion are effective ways to manage employees. Especially in Funding
terms of salary management, it needs to be adjusted in time
according to the actual situation, and at the same time, it is Supported by the high level talent scientific research startup
necessary to formulate employee promotion channels to meet the foundation of North China University of Water Resources and
development needs of employees. Enterprises can use the human Electric Power (No. 201801004).
resource management system to understand the situation of
employees in a timely manner and adjust the management policies
of the enterprise in time. By continuously improving the strength Conflict of interest
of the workforce, they work together to forge ahead with the
company’s development goals. Although this paper has some The author declares that the research was conducted in the
conclusions in the process of exploration, there are still absence of any commercial or financial relationships that could
shortcomings: the operating efficiency of the system needs to be construed as a potential conflict of interest.
be improved, and the system is not efficient enough when dealing
with a large amount of data.
Publisher’s note
Data availability statement All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their affiliated organizations,
The original contributions presented in the study are included or those of the publisher, the editors and the reviewers. Any product
in the article/supplementary material; further inquiries can that may be evaluated in this article, or claim that may be made by its
be directed to the corresponding author. manufacturer, is not guaranteed or endorsed by the publisher.

References
Agrawal, A., Gans, J. S., and Goldfarb, A. (2017). What to expect from artificial Labovitz, D. L., Shafner, L., and Gil, M. R. (2017). Using artificial intelligence to
intelligence. MIT Sloan Manag. Rev. 58, 23–26. reduce the risk of nonadherence in patients on anticoagulation therapy. Stroke 48,
1416–1419. doi: 10.1161/STROKEAHA.116.016281
Belizon, M. J., Morley, M. J., and Gunnigle, P. (2017). Modes of integration of
human resource management practices in multinationals. Pers. Rev. 45, 539–556. Lee, E. J., Kim, Y. H., and Kim, N. (2017b). Deep into the brain: artificial
doi: 10.1108/PR-09-2014-0207 intelligence in stroke imaging. J. Stroke 19, 277–285. doi: 10.5853/jos.2017.02054
Bin, Y. U., and Kumbier, K. (2018). Artificial intelligence and statistics. Front. Info. Lee, H., Troschel, F. M., and Tajmir, S. (2017a). Pixel-level deep segmentation:
Tech. Electronic Engineering 19, 6–9. doi: 10.1631/FITEE.1700813 artificial intelligence quantifies muscle on computed tomography for body
morphometric analysis. J. Digit. Imaging 30, 487–498. doi: 10.1007/s10278-017-9988-z
Burton, E., Goldsmith, J., and Koenig, S. (2017). Ethical considerations in artificial
intelligence courses. AI Mag. 38, 22–34. doi: 10.1609/aimag.v38i2.2731 Majumder, M. (2017). Human resource management practices and employees’
satisfaction towards private banking sector in Bangladesh. Int. Rev. Manag. Mark. 2,
Cooke, F. L. (2017). Concepts, contexts, and mindsets: putting human resource
52–58.
management research in perspectives. Hum. Resour. Manag. J. 28, 1–13. doi:
10.1111/1748-8583.12163 Ones, D. S., Viswesvaran, C., and Schmidt, F. L. (2017). Realizing the full potential
of psychometric meta-analysis for a cumulative science and practice of human
Crowley, F., and Bourke, J. (2017). The influence of human resource
resource management. Hum. Resour. Manag. Rev. 27, 201–215. doi: 10.1016/j.
management systems on innovation: evidence from Irish manufacturing and
hrmr.2016.09.011
service firms. Int. J. Innov. Manag. 21, 1750003–1750536. doi: 10.1142/
S1363919617500037 Polina, M., Lucy, O., and Yury, Y. (2018). Converging blockchain and next-
generation artificial intelligence technologies to decentralize and accelerate
Diaz-Fernandez, M., Pasamar-Reyes, S., and Valle-Cabrera, R. (2017). Human
biomedical research and healthcare. Oncotarget 9, 5665–5690. doi: 10.18632/
capital and human resource management to achieve ambidextrous learning: A
oncotarget.22345
structural perspective. BRQ Bus. Res. Q. 20, 63–77. doi: 10.1016/j.brq.2016.
03.002 Price, S., and Flach, P. A. (2017). Computational support for academic peer
review: a perspective from artificial intelligence. Commun. ACM 60, 70–79. doi:
Ghoson, T., and Faris, A. (2021). Effects of knowledge management on
10.1145/2979672
organizational performance. American J. Business and Operations Res. 5, 61–84. doi:
10.54216/AJBOR.050201 Rothenberg, S., Hull, C. E., and Tang, Z. (2017). The impact of human resource
management on corporate social performance strengths and concerns. Bus. Soc. 56,
Grzonka, D., Jakóbik, A., and Koodziej, J. (2017). Using a multi-agent system
391–418. doi: 10.1177/0007650315586594
and artificial intelligence for monitoring and improving the cloud performance
and security. Futur. Gener. Comput. Syst. 86, 1106–1117. doi: 10.1016/j. Shivam, G., Kshitij, S., and Vanita, J. (2020). Egocentric Performance Capture: A
future.2017.05.046 Review. Fusion: Practice and Applic. 2, 64–73. doi: 10.54216/FPA.020204
Guest, D. E. (2017). Human resource management and employee well-being: Youssef, A., El-Telbany, M., and Zekry, A. (2017). The role of artificial intelligence
towards a new analytic framework. Hum. Resour. Manag. J. 27, 22–38. doi: in photo-voltaic systems design and control: A review. Renew. Sust. Energ. Rev. 78,
10.1111/1748-8583.12139 72–79. doi: 10.1016/j.rser.2017.04.046

Frontiers in Psychology 11 frontiersin.org