Research on the Operation System of Chinese Performance Market

Based on Data Mining Technology

Yong Wang and Nam-gyuCho

Department of Performance Art Management, Sangmyung University, Seoul 03015, Republic of Korea, Korea

Keywords: Operation System, Chinese Performance Market, Data Mining.

Abstract: The rapid development of computer technology makes it possible to process data, which promotes the great

development of database technology. Data mining is a data processing technology developed to meet this

need. In order to solve the problems faced by China's performance market at present, the author explores and

analyzes the operation mode of the performance market. According to the present situation of China's

performance market operation, a performance market operation system based on data mining technology is

composition of the system are discussed in depth. The system runs stably and the data statistics are processed

accurately, which plays a positive role in promoting the performance industry to broaden new business

markets.

1 INTRODUCTION

As an important part of cultural industry, the market

maturity of performance industry is still not high. At

present, in China, except for Beijing and Shanghai,

where the performance industry is relatively

developed and forms a relatively mature performance

market, the performance industry in other regions is

almost in a deep sleep state (Ebrahimi, Asemi,

Nezarat, et al. 2021). More and more countries realize

the great influence and restriction of culture on

contemporary social and economic life. In recent

years, with the continuous improvement of China's

overall economic level, people's awareness of

cultural consumption is also constantly improving.

Cultural industry, especially mass cultural industry,

has been paid more and more attention by the

government and all sectors of society (Klepac, Kopal,

Mrsic 2019). There are some defects in the business

scope of the original performance market, which are

mainly due to the lack of market development, market

consumption and the high price of performance

tickets. However, through the improvement at the

present stage, it has improved the operation scope of

the original market and promoted the domestic

market economy.

Data mining is a very broad interdisciplinary

subject, which brings together different technologies,

especially information technology and statistical

analysis technology. The main purpose of analyzing

data is to provide real and valuable information for

business decision-making, so as to obtain profits.

Driven by modern science and technology, marketing

is developing towards automation, digitalization and

informationization. We need to extract hidden

information from a large amount of data that can

support decision-making, analyze the comprehensive

impact of business operation on society, economy

and environment, and predict the future development

trend of enterprises (Cui, An, Zhang 2021, Cui 2021).

Combined with the actual situation of the operation

system construction in China's performance market,

this project provides an advanced and practical

solution for the integration of operation system based

on data mining technology. This project is highly

targeted and has certain practical significance.

2 ANALYSIS AND COMPARISON

OF PERFORMANCE MARKET

MODELS IN CHINA

According to the research of relevant institutions in

recent years, the main management modes of mass

cultural performances in China can be summarized as

the following three modes.

Wang, Y. and Nam-gyuCho, .

Research on the Operation System of Chinese Performance Market Based on Data Mining Technology.

DOI: 10.5220/0011735500003607

In Proceedings of the 1st International Conference on Public Management, Digital Economy and Internet Technology (ICPDI 2022), pages 311-316

ISBN: 978-989-758-620-0

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

311

(1) Management mode of free cultural groups

Today, with the increasingly developed market

economy, the vitality of various cultural groups

(including karaoke bars, performing arts halls, etc.)

has become increasingly prominent due to their

flexible mechanism and strong market adaptability.

Its advantage is that in order to maximize profits,

small-scale cultural enterprises will maximize the use

of performance venues, provide more rich

performances to the society, actively look for the

programs that are in attendance, increase their

income, and reduce the financial burden of the state

to a certain extent. But at the same time, it should be

noted that although the contracted management of

small cultural enterprises has great vitality, it is not

suitable for the management of large-scale

performances (Ozdagoglu, Oztas, Cagliyangil 2019).

However, small-scale cultural enterprises often

aim at making profits, and it is difficult to invest and

cultivate the performance market that can not get

returns in the short term. At the same time, the

standard for enterprises to choose programs is

whether they are profitable or not, so it is difficult to

guarantee the artistic taste of performances, which

will make some dirty and low-cost kitsch

performances flood among them, and fail to realize

the true original intention of meeting the audience's

cultural needs.

(2) Professional collectivization management

mode

Entrusting large-scale state-owned enterprises to

implement group management seems to have been

reformed in a short time, which has many elements of

enterprise management, but it is no different from the

direct management of the government in essence.

After long-term operation, all the shortcomings of

state-owned enterprises are fully reflected in the

performance management.

(3) Management mode of public institution

The advantage of this model is that it can

implement government instructions to the maximum

extent and the government decrees are smooth; The

disadvantage is that in the face of the lack of

motivation for market operation, it relies too much on

the government and finance, and its responsibilities

and rights are unclear. Employees lose the motivation

to pursue progress because they enjoy the "iron rice

bowl" for a long time.

3 DATA MINING

Nowadays, data mining research is considered to be

another new wave in the field of information

technology after the Internet. As a mathematical tool

for knowledge development and innovation, data

mining can be widely used in many social

information fields such as finance, market

development, medical diagnosis and decision-

making, traffic management, and enterprise

performance evaluation, so as to improve the

reliability and accuracy of data analysis in the above

industries (Rahimi, Sharifzadeh, Feng 2020). Today,

these mature technologies, coupled with high-

performance relational database engine and extensive

data integration, make data mining technology enter

a practical stage in the current data warehouse

environment.

It is convenient to divide the types of data mining

tasks according to the different goals of data analysts.

The classification given below is not unique, and it

can further divide more detailed tasks, but it

summarizes various types of data mining activities.

The implementation steps of data mining are shown

in Figure 1 below:

Figure 1: Implementation steps of data mining.

(1) Business Understanding

This initial stage focuses on understanding the

ICPDI 2022 - International Conference on Public Management, Digital Economy and Internet Technology

312

objectives and requirements of the project from a

commercial point of view, then transforming the

understanding into data mining problems, and

working out a preliminary plan to achieve the

objectives.

(2) Data understanding

The stage of data understanding begins with the

collection of original data, and the following

activities are familiar with data, identifying data

quality problems, exploring the first understanding of

data, or mining data subsets with deep meanings to

form assumptions about hidden information.

(3) Data preparation

The data preparation stage includes all activities

to construct the final data set from the original

unprocessed data. Data preparation tasks may be

carried out many times, and not in any specified order.

These tasks include the selection of tables, records

and attributes, and the conversion and cleaning of

data in modeling tools.

(4) Modeling

At this stage, we mainly choose various modeling

techniques and calibrate their parameters to reach the

optimal value. Usually, there are many methods for

the same data mining problem type. Some methods

have specific requirements on data form, so it is often

necessary to return to the data preparation stage.

(5) Assessment

Before the final release of the model, it is very

important to evaluate the model more thoroughly and

check each step of building the model, so as to make

sure that it has completely achieved the business goal.

A decision on the use of data mining results should be

made at the end of this stage.

(6) Release

According to needs, the publishing process can be

as simple as generating a report, or as complex as

executing a repeatable data mining process in the

whole enterprise. In most cases, the release is carried

out by customers, not the data analysts themselves.

However, even if the analyst does not execute the

release, it is very important for the customer.

4 ANALYSIS OF PERFORMANCE

MARKET OPERATION

SYSTEM

4.1 System Design Framework Based

on Data Mining

The data of performance industry has the

characteristics of sea quantification. For a long time,

due to the variety of business and the large number of

customers, not only has a large amount of important

business data been accumulated, but with the

expansion of business and the rapid increase of

customer volume, these data are growing at an

alarming rate every day. Based on this consideration,

the data mining technology of performance industry

should be built on the data warehouse system.

Any system using data mining technology has the

following functional modules: database or data

warehouse server, which manages the database or

data warehouse and preprocesses the data. User

graphical interface, that is, the interface where users

interact with data mining module. Users

communicate with the system through data mining

language to complete mining work (Chubukova,

Ponomarenko, Nedbailo 2020). The operating system

we designed is also composed of these functional

modules. According to a certain period, the data

mining server extracts data from the data warehouse

server, then carries out data mining according to

preset parameters and patterns, saves the generated

rules and patterns to the data mining server, and

provides the data mining results to the corresponding

managers through visual tools. Evaluation data

analysts can modify data mining parameters and data

update and mining cycles through the management

system. The president structure of the system is

shown in Figure 2.

Figure 2: System president structure chart.

The database in this system is the basic data

source. Its contents mainly include all kinds of

information related to customers, such as customer

background information, transaction history, etc. Is

the most primitive data. It is at the lowest level of

data, and keeps direct contact with the client due to

the need of dynamic update, which is generally the

business database of an enterprise, in addition to data

sources from other channels. In the data warehouse

Research on the Operation System of Chinese Performance Market Based on Data Mining Technology

313

system, traditionally, the workload is the largest, and

the problem in daily operation is the work of

extracting, transforming and integrating data from

business database to data warehouse. The reason is to

extract, transform and integrate data from different

kinds and forms of business, and finally store it in

data warehouse. And to maintain and manage the

quality of data.

4.2 Establishment of Prediction Model

The churn prediction model is a model for predicting

customer churn in the performance market, which

subdivides different prediction targets of different

users and makes the prediction results more forward-

looking. When the model is initially built, it is

necessary to study the data in the data warehouse and

obtain the data related to the loss prediction and

analysis (Szafrański, Zieja, Wójcik, et al. 2018). And

organize them according to the time granularity to be

studied, and explore the data. At the same time,

strengthen the discussion with the bureau. Through

exploring the data and understanding the demand, we

should be able to preliminarily determine the time

window structure of the forecasting model, the

definition of the group to be predicted, the forecasting

target, and select the index set sensitive to the

forecasting target.

The process of data understanding is an iterative

process. Choosing the appropriate time window

structure, groups to be predicted, prediction targets

and index sets is half the success of loss prediction.

Therefore, we must carefully scrutinize the data to

understand the work at this stage (Bimonte, Billaud,

Fontaine, et al. 2021). In the stage of data exploration,

the most valuable index set for loss prediction is

obtained. Then, by making derivative variables, the

data can more fully reflect the customer's behavior

changes. After defining the forecast target, the

forecast target is divided into several loss types, and

the priority of each type is defined to ensure that each

customer is in only one loss type state. Then, using

the data in the time window, mark each user with the

churn type.

In the data preparation stage, after making the

analysis table of time period A and time period B, the

operation of establishing the model can be started,

and the number of samples to be extracted in the

sample table is designed. Assume that 30,000

samples of 0 loss type need to be sampled, and 1,000

samples of other loss types need to be sampled.

Create an empty table to store the samples extracted

next. This table is called sample table for short. Pay

attention to tick the position of "Output should be

attached to the specified table" in the figure. The data

preparation process is shown in Figure 3.

Figure 3: Data preparation process.

The data of loss prediction model mainly comes

from two parts: detailed accounting list, user data, etc.

If the data warehouse of the system has been built,

these data can be provided by the data warehouse;

otherwise, a data mart can be set up separately to

provide data for data mining.

Take samples with loss type 0. When extracting,

first define the source table. On the input data page of

bivariate statistics, in the available input data box,

select a time period analysis table. In the filter record

condition box, the selection condition is "loss type

flag =0". On the sample page, select "create sample".

The sampling technology is "random selection of N

records", and the number of records should be filled

in "30000".

Input the sample table into the decision tree model

to start training. After the training, check the

confusion matrix output by the model to determine

the training effect of the model. If you are not

satisfied with the training effect, you can try to retrain

the model by adjusting the number of samples, the

proportion of samples, the input fields of the model,

the weights of the fields, and the parameters of the

decision tree algorithm to improve the effect of the

model. When modifying the parameters of decision

tree algorithm, the prediction effect of the model can

be improved by setting the cost matrix.

4.3 Solution Based on Data Warehouse

Platform

The performance market operation system generally

adopts a four-tier structure, as shown in Figure 4. The

operation system has established a unified enterprise

data information platform for the performance

industry. In this paper, the advanced data warehouse

technology and system analysis and mining tools

which are popular in the market in recent years are

used to extract useful information from the enterprise

historical data, provide services for the enterprise

ICPDI 2022 - International Conference on Public Management, Digital Economy and Internet Technology

314

customers, and comprehensively enhance the

operation level and competitive strength of the

enterprise, so as to embody the customer-centered

enterprise management philosophy.

Figure 4: Framework of operating system.

The architecture of the operating system adopts a

four-tier structure. Include:

(1) Data acquisition layer

The data layer defines the scope of system

management data, which mainly includes different

subsystems of operation system, including

application systems such as distributed data source,

ETL system, operation system, OLAP and front-end

applications.

(2) Data storage layer

The storage layer provides technical means and

data for data quality management, mainly including

metadata management, algorithm base, rule base and

intermediate information.

(3) Application layer

Provide various analytical applications of topic

analysis and thematic analysis, such as group

business analysis, data business analysis and so on.

(4) Data access layer

The access layer is the window and platform for

all kinds of business operators to access the operation

system, which is composed of two parts: the unified

access platform for users and various specific access

tools.

The basic idea of the system is to display rich

information through flexible configuration of the

underlying basic elements. In order to make the

system have good structure and maintainability, the

system is developed by java and jsp, and mature SSH

technology is adopted in the design, which is layered

according to MVC structure, which is convenient for

cooperative development, and also beneficial for

future maintenance and expansion.

In the process of model design, data quality is the

most important of all problems, so we must ensure the

authenticity of the data used, and make records to

compare the impact of each change on the results. In

this way, we can keep clear thinking and not get into

a dead corner. In the application process of the model,

it is necessary to have more contact and

communication with market personnel, so as to

ensure that the data handed over to market personnel

are truly usable data and avoid the disconnection

between development and application.

5 CONCLUSION

Data mining technology is the most powerful data

analysis method in the field of data warehouse at

present. Compared with the verification analysis of

OLAP, the analysis method of data mining uses

known data to find out the hidden business rules by

establishing mathematical models, which has been

successfully applied in many industries. The changes

of market and customer demand put forward new

requirements for the performance market operation

system. It is necessary to establish an efficient data

statistical analysis system to provide decision support

for the development of performance business. The

emergence of data warehouse and data mining makes

it possible for the performance industry to fully

exploit and utilize customer historical data

information. In this paper, the application of data

mining in performance industry is analyzed, and

some preliminary results are obtained, and an

analysis system based on data mining technology is

established The characteristics, architecture and

common models of data warehouse are analyzed. In

addition, OLAP and its multidimensional analysis

and data mining are described in detail.

REFERENCES

B Szafrański, Zieja M, J Wójcik, et al. Data analysis of

information system TURAWA in the aspect of data

mining methods[J]. Computer Science and

Mathematical Modelling, 2018(7/2018):45-52.

Bimonte S, Billaud O, Fontaine B, et al. Collect and

analysis of agro-biodiversity data in a participative

context: A business intelligence framework[J].

Ecological Informatics, 2021(4):101231.

Cui Z, An F, Zhang W. Internet Financial Risk Assessment

Based on Web Embedded System and Data Mining

Algorithm[J]. Microprocessors and Microsystems,

2021, 82(3):103898.

Cui Y. Intelligent Recommendation System Based on

Mathematical Modeling in Personalized Data Mining

Research on the Operation System of Chinese Performance Market Based on Data Mining Technology

315

[J]. Mathematical Problems in Engineering, 2021,

2021(3):1-11.

Chubukova, Ponomarenko, Nedbailo. Using data mining to

process business data[J]. Problems of Innovation and

Investment Development, 2020(23):71-77.

Ebrahimi F, Asemi A, Nezarat A, et al. Developing a

mathematical model of the co-author recommender

system using graph mining techniques and big data

applications[J]. Journal of Big Data, 2021, 8(1):1-15.

Klepac G, Kopal R, Mrsic L. Early Warning System

Framework Proposal, Based on Big Data

Environment[J]. International Journal of Artificial Life

Research, 2019, 9(1):35-66.

Liu Y, Huang J, Lu N. Research on Environmental

Monitoring System Based on Microservices and Data

Mining [J]. E3S Web of Conferences, 2020, 143(3):

02031.

Ozdagoglu G, Oztas G Z, Cagliyangil M. An application

framework for mining online learning processes

through event-logs[J]. Business process management

journal, 2019, 25(5):860-886.

Rahimi B, Sharifzadeh M, Feng X T. Ground behaviour

analysis, support system design and construction

strategies in deep hard rock mining – Justified in

Western Australian's mines [J]. Journal of Rock

Mechanics and Geotechnical Engineering, 2020, 12(1).

ICPDI 2022 - International Conference on Public Management, Digital Economy and Internet Technology

316