A STRATEGY FOR BUILDING A SMART

SPORTS PLATFORM BASED ON MACHINE

LEARNING MODELS

Mingchan Gong *

School of design and art, Hunan Institute of Technology, Hengyang, Hunan,

421000, China

2000001200@hnit.edu.cn

Reception: 13/11/2022 Acceptance: 08/01/2023 Publication: 23/03/2023

Suggested citation:

G., Mingchan. (2023). A strategy for building a smart sports platform based

on machine learning models. 3C TIC. Cuadernos de desarrollo aplicados a

las TIC, 12(1), Page-Page. https://doi.org/10.17993/3ctic.2023.121.248-265

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

248

ABSTRACT

With the rapid development of big data technology, it has greatly changed the way

people get information, and also improved the speed and quality of information. In this

context, smart sports has become a new trend in sports development. This paper

creates an intelligent learning environment and builds a smart sports platform through

advanced concepts and technical means, which can effectively optimize the

integration and sharing of sports resources. Starting from the overall architecture

design of smart sports, the key technologies of machine learning model to realize

smart sports are sorted out. Through the five basic linking stages with machine

learning model as the core, the value innovation path of platform construction

structure is analyzed. The current status of sports resources application is studied,

and the data mining algorithm is used to calculate the user usage data of the smart

sports platform and improve the construction of the smart platform. Through the

construction of the smart sports platform, people shift from traditional reading books

and watching TV programs to getting information through intelligent mobile terminals,

and the proportion of attention to sports information is as high as 58.6%. This shows

that by building a smart sports platform, it can provide support and guarantee for the

sustainable development of sports.

KEYWORDS

Smart sports; machine learning modeling technology; resource sharing; platform

construction; sports information

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

249

PAPER INDEX

ABSTRACT

KEYWORDS

1. INTRODUCTION

2. SMART SPORTS PLATFORM CONSTRUCTION

3. SMART SPORTS SERVICES

4. SMART SPORTS PLATFORM DESIGN

5. MCHINE LEARNING MODEL ALGORITHMS

6. PLATFORM CONSTRUCTION ANALYSIS

6.1. Test Environment and Methodology

6.2. System Performance

6.3. Survey Results

7. CONCLUSION

DATA AVAILABILITY

CONFLICT OF INTEREST

REFERENCES

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

250

1. INTRODUCTION

Machine learning model analysis has increasingly become an important reference

for people's judgment and decision making, and is increasingly known and used.

Various analytical tools and analytical methods based on machine learning models

have emerged [1-4]. The idea of big data, which itself has undergone the process

from experiment to practice and from niche to mass, is also increasingly known [5-8].

The smart sports platform is based on information technology, including machine

learning model analysis, Internet of Things, and cloud computing technology, which

together build a public basic service platform integrating social, cultural, sports, and

environmental factors [9-12]. Its main functions include query service function, data

collection function, and data management function. The rapid development of

communication technologies such as machine learning models has made the

integration and application of resource information the key to the construction of

sports informatization [13-15]. The rapid development of science and technology

development under machine learning models will promote the application of mobile

terminals such as smartphones and tablets in the sports industry [16-18]. Machine

learning models are a collection of massive data, the volume of which is particularly

huge and cannot be processed by conventional software within a certain period of

time. Machine learning models are characterized by massive scale, high-speed flow,

and rich form [19-20]. The construction platform on the basis of machine learning

models should be unified to manage the collected data, and the multi-channel, multi-

level and multi-angle feedback query service window after data analysis. The platform

can achieve the sharing of sports information in the province and even in the country

or even globally, and build a service network covering the whole province, with the

characteristics of intelligence, innovation and extensiveness [21].

The literature [22] mentions that smart cities are the direction and advanced form of

future urban development, integrating the functions of digital, knowledge, ecological,

and creative cities. Big data is changing people's life, work and thinking, bringing

major changes in cities. Huge amount of data exists in all aspects and fields of the

city. The establishment of a technical system framework for smart cities based on big

data technology is discussed, and the feasibility of this technical system framework is

explored. The literature [23] outlines the significance of machine learning modeling

techniques in the construction of smart cities. Its important role in further promoting

the development of the Zhengzhou Airliner Economic Zone in China and the whole

process is analyzed. As well as how it can promote economic transformation and

upgrading, it is expected to provide ideas and support for the Zhengzhou Airliner

Economic Zone, China, to lead and promote the regional economy of Henan, China,

in the future and achieve continuous improvement in sustainable development.

Literature [24] et al. Also, a more reliable trust protocol can be established, which is

conducive to the implementation of machine learning model applications. The

literature [25] and others mention that with the advancement of urbanization and the

coordination of urban transportation, municipalities, economic industries.

This paper analyzes the application of machine learning models in the design of

smart sports and evaluates the impact of the planning and construction of smart

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

251

sports based on machine learning models on economic and social benefits. Through

the wide application of machine learning models in building smart sports platforms,

the government is able to run more smoothly, conveniently and efficiently in planning,

construction, industry, people's livelihood, society and cities. The current situation of

sports informatization construction, as well as the advantages and related strategies

of the construction plan of the smart sports platform based on big data, are mainly

analyzed and discussed.

2. SMART SPORTS PLATFORM CONSTRUCTION

The construction target of the smart sports platform is to gather various types of

venues and fitness groups of people's resources, and realize the digital fitness

consultation and guidance, sports and fitness activity resources, venue booking,

equipment purchase, personalized sports services and sports culture exchange for all

people and high social participation through the platform [26-27].

Figure 1. Smart Sports Platform Framework and Structure

The essential difference between smart sports services is the empowerment of

"wisdom". It refers to the use of advanced concepts, technology and other means to

create an intelligent learning environment, forming a precise, individual and flexible

education service system as shown in Figure 1. As can be seen, although "wisdom"

has different facets, "wisdom" represents, first of all, positive, innovative and

comprehensive capabilities on the outside, and on the inside, value understanding

and moral identity for good and common good. In order to better clarify the seven

modules of the application service layer, to open up and integrate the storage and

computing centered on big data, and to realize the effective operation of the modules

and the process of value output.

Control

Forecast

Decision

3D visualization

platform

Unified Data Processing

Engine

Form a real-time data source

Business process integration

Decision-making

level

Management

Business Layer

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

252

Figure 2. Creating a value innovation path for smart sports with machine learning models at

its core

As shown in Figure 2, the module value innovation is analyzed through five basic

linking stages with machine learning model as the core.

Demand identification: Based on the collection of basic data such as users'

physical health level, sports cognition and technical ability, sports learning demands

and sports preferences, the demand instruction for sports course learning is

intelligently generated.

Precise Matching: By analyzing and mining the system data, we provide precise

matching course learning solutions.

Intelligent push: When personalized learning solutions are intelligently pushed to

the platform's user interface, users can operate tasks and learn skills according to the

sports process and specific requirements of different scenarios in the module

architecture even in fragmented time periods with the help of ubiquitous networks and

mobile terminals.

Decision-making and implementation: Once a user enters a certain task point, the

coach will synchronously track and instantly share the learning process data through

intelligent perception, supervise and control the efficiency of sports implementation,

make timely professional assessment in three dimensions of "sports spirit, sports

practice and health promotion", and provide personalized counseling and suggestions

for non-standard problems, as well as instant evaluation and feedback.

The evaluation of sports completed by users in the client is instantly fed back to the

intelligent learning system to promote the quality of subsequent sports activities. In the

five basic linking stages, the sports courses through the platform fundamentally

stimulate the initiative and motivation of "value co-creation" between the platform and

users, and realize the "sports meaningful growth" with technical support in the new

era. Through the instant dialogue, two-way communication and joint exploration

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

253

between coaches and users, the skills learning willingness of users is fully achieved in

the new scenario, and targeted sports activities are effectively promoted.

3. SMART SPORTS SERVICES

From the perspective of service essence, the "wisdom core" of smart sports service

is a "people-oriented wisdom". It takes the sports needs of service recipients as the

core, and provides the most necessary, suitable, accurate and convenient

personalized sports platform services and health management services through the

systematic changes of service forms, service contents and service functions, thus

manifesting the human-centered service thrust.

Figure 3. Smart Sports Service Structure

As shown in Figure 3 its performance is specified in：

Openness of service form. With the support of modern information technology,

users only need to hold smart terminals to seamlessly connect and interoperate to

the open, collaborative and shared central system of smart sports service platform,

they can get "menu" sports resources and services anytime, anywhere, freely and

efficiently, so that sports are everywhere.

Wide extension of service content. In terms of content scope, the content of

intelligent sports services covers online and offline integrated sports model, sports

community interaction, intelligent venue management and other multi-dimensional

needs. In terms of content depth, sports skills output is more full and vivid under the

conditions of language, image, emotional awareness three-dimensional multi-

sensory technology, which can make the sports experience under the resonance of

sports thinking and sports psychological immersion more profound.

3. Integration of service functions. It is through the application of intelligent technology

and information technology, fully mining and analyzing the full record of data

information and automatically generate adapted personalized services to meet the

differentiated sports needs of the user body; through portable devices associated

with the user's heart rate, blood pressure, pulse oxygen saturation and other

physiological indicators, expression behavior recognition technology analysis of

sports emotions, perception and other psychological indicators of the dual

Students' skills learning demands and

preferences

The structure of students' sports cognition

and ability

Student Physical Health Test and

Diagnosis

Data collection

Need identification

Exercise Risk Intervention

>>>>>

Sports

Course Mode

Sports ground

Course teachers

Class time

>>>>>

Exact match

Data Integration

Data Classification

Data mining

Course guide

Online class

Online Q & A

Result inquiry

>>>>>

Smart push

Sports Core Literacy

Sportsmanship

Exercise practice

Health promotion

>>>>>

Decision

implementation

Supervision and

Early Warning

Evaluate and

adjust

Instant evaluation and feedback

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

254

assessment feedback, to achieve the scientific norm of the user's sports risk

Prevention and control, etc.

4. SMART SPORTS PLATFORM DESIGN

While leading sports reform and innovation, smart sports services are bound to

bring great impact and challenges to the traditional sports learning mindset, sports

management model and sports cultural environment.

Figure 4. Logical framework design for smart sports platform

As shown in Figure 4, the resource service layer refers to the application functions

based on the realization of data processing, mainly through modeling, to realize the

analysis, abstraction, verification, combination and other processing of data extracted

from the database, and the functions for event processing include brief event

processing, event query configuration, event flow processing, rule processing, etc.

The data layer, i.e. database, is a warehouse for organizing, storing and managing

data according to the data structure, and as the basic data storage area of the whole

platform, and guarantee the integrity and convenience of data extraction by users.

As the convergence layer of the platform data, the unified collection realizes the

convergence, filtering and data conversion of sports-related information such as

sports activities personnel, and then submits them to the front subsystem for

temporary storage, and submits them to the storage.

Users complete the selection of self-chosen content for sports courses within a

specified period of time through the Smart Sports Platform. The platform sets the

Sports

Personnel

Information

Database

Query service platform

Database

management

Storage

management

Link

management

Security control

Resource service

layer Event query

configuration

Data analysis Data abstraction Data verification Data combination

Event stream processingSimple event handling Rule processing

Data layer Sports Facility

Information

Database

Sports Industry

Information

Database

Industry

database

Third-party

resource

database

ETL

Unified collection

System debugging

Data Quality

Management Metadata management Front-end subsystem

System Management ETL monitoring

Coach

information

Basic information Athlete

Information

Simple event

handling

National

Fitness

Activities

Information

about various

sports clubs

Sports

Industry

Information

Traffic and

weather

information

Data service bus

Data Collection

Management Center

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

255

course content name for users to choose, and users choose the course with their own

situation and interest points. Users can also give feedback on courses that are not set

up through the platform's comment section and text the reasons for applying for the

course. The platform will give feedback to the sports coach, and the sports coach will

try his best to provide appropriate advice to support his own situation. This way not

only realizes the interaction between coaches and students, but also improves the

efficiency of communication.

5. MCHINE LEARNING MODEL ALGORITHMS

The data mining algorithm utilized in this paper contains a clustering analysis

algorithm and a BP neural network model algorithm, and the two are organically

combined to achieve data analysis of the construction of a smart platform. Thus, the

user usage data of the smart sports platform is calculated, which can better improve

the platform information and grasp the psychological needs of users.

Firstly, the data is classified and calculated using the K-means algorithm for the

measurement technique step. Its main steps for analyzing big data are：

1. K centroids are randomly extracted from different sample data to select the initial

cluster centroids.

2. Divide the sample cluster points by allocating each different data point to the center

nearest to that data point. In this step, the distance formula is introduced to obtain:

(1)

Based on the selected centroids of the clustered samples of each data, the

distance between each database sample data and these central sample parameters

is calculated using Equation (1), and the corresponding large data are re-classified

according to the minimum distance.

(3) Based on the mean value of each clustered sample data object, the distance

between each object and these central objects is calculated, and the corresponding

objects are re-divided according to the minimum distance recalculating the mean

value of the distance from the point in each class to the centroid of that class,

assigning each data to its nearest centroid. Forming the matrix with the minimum

data calculated each time, thenl:

(2)

where is the set of the found minimum values.

(4) In big data processing, by using K-means algorithm can obtain the clusters with

the smallest error criterion function of big data. By centering K sample data points in

(x,y)2=

∑

i=1

(xi−yi)2=x−y

,⋯,x

x21,x22,⋯,x2n

⋯

,xk

,⋯,xkn

xij

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

256

the space and clustering them, the information big data closest to the different

samples is finally categorized.

After using the K-means algorithm output results, then use the BP network

algorithm model algorithm to continue mechanical learning, training. It is able to map

and handle the more complex nonlinear relationships in the data samples in a timely

manner. In adjusting the BP neural network model, the following formula is followed:

where the adjustment formula for the output layer power system is:

(3)

The adjustment formula for the implied layer weight factor is:

(4)

The quadratic exact function model for pairs of input patterns in different large data

samples is:

(5)

Expression for the total accurate function for different large data samples:

(6)

Big data samples are standardized as follows: assuming that the type of input big

data information is and the sample is , for the input data is standardized

according to the steps of the following equation.

(7)

(8)

(9)

In the above equation, , and in the above equation

are the data after the normalization process. The standardization formula is shown in

Equation (10).

(10)

Where: is the output database sample data information; is the normalized

database sample big data; and is the extreme and extreme small values of

the output database sample big data; where .

Δωij =ηOp

k(1 −Op

k)(tp

k−Op

k)Op

ωij =ηOp

i(1 −Op

∑

i=1

ΔωkiO

∑

k=1

(tp

k−Op

∑

p=1

Jp=

∑

p=1

∑

k=1

(tp

k−Op

Xij

ij =

−

)



∑

i=1

−1

∑

i=1

(xij −

xj)

i= 1,2,⋯,N;j= 1,2,⋯,m

Zij

′

q(y

−y

min

+b)

(ymax −ymin +b)

y′

ymax

ymin

0<q<3;0<b< 2

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

257

In the validation of clustering analysis algorithms, F-measure can be chosen as the

evaluation criterion for determination, and here, two parameters, accuracy and recall,

are cited. Accuracy and recall are able to evaluate the accuracy rate of clustering

classification algorithms and are calculated as follows:

The accuracy rate is calculated as follows:

(11)

Recall rate calculation formula.:

(12)

Solve to obtain the value of F:

(13)

Through the above-mentioned calculation process, the data mining algorithm can

make the weights and thresholds in the neural network gradually adjusted to gradually

approximate the results precisely needed for the test system, and timely discover all

kinds of data information in the use of the intelligent sports platform. The data mining

computing service is a platform service that strips the business logic from the

computing application and provides the platform support service required by the

business logic to the developer. Corresponding to the node part of the application, it

provides data input adaptation and data pre-processing services, the processing node

part provides streaming data analysis, complex event processing and business rule

processing services, and the output node part provides data output adaptation

services. Combining the above algorithms, the SCS model is proposed. Developers

can define a streaming computing application that can run on the streaming

computing service system based on the elements of the SCS model.

6. PLATFORM CONSTRUCTION ANALYSIS

6.1. TEST ENVIRONMENT AND METHODOLOGY

The construction of smart sports platform meets the practical needs and greatly

improves the level of informationization of sports. In particular, universities should

reasonably build a platform under the framework of smart sports based on machine

learning model technology. In order to better build this platform, we should effectively

share the standardized data, create a "one-stop" service platform, do a good job of

data statistics and analysis, and provide a solid and reliable basis for decision-making.

After the above machine learning model algorithm, this part uses the Storm

application execution entity to test the dynamic adjustment of parallelism, and then

verifies the feasibility and effectiveness of the streaming SCS definition tool and SCS

execution engine by application examples.

precision(i,j) =

nij

ecall(i,j) =

nij

I=∑

nma x{F(i,j)

}

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

258

In the test of Storm application execution entity parallelism dynamic adjustment

algorithm, there are four virtual machines in the test cluster. One of them is used as a

Nimbus node and Zookeeper is installed on the Nimbus node, while the other three

are used as Supervisor nodes. All four virtual machines are installed with Ubuntu

12.04 OS, Open JDK 64-bit1.8.0_45, Storm version 0.94, and Zookeeper version

3.4.7.

The test environment for the dynamic tuning of the Storm application execution

entity parallelism is shown in Table 1:

Table 1. Application execution entity parallelism dynamic test environment table

The first step is to verify that the Parallelism Dynamic Adjustment module can

dynamically adjust the parallelism ratio between components with sequential order

according to the component execution events, and the second step is to verify that the

Parallelism Dynamic Adjustment module can optimize resource utilization by

dynamically adjusting the parallelism of each component. In our experiments, we set

the statistical interval of Storm to 5 seconds, set the topology builtin. metrics.bucket

size secs item in the configuration file of Storm.yaml as follows, set the high threshold

to 92%, the low threshold to 32%, and the count period to 10. For the test of dynamic

adjustment of entity parallelism of the Storm application execution, we validate it by

three different applications Generall, General2 and General3, which all correspond to

a Topology containing three nodes, i.e., one Spout and two Bolt nodes. The structure

diagram is shown in Fig.

Figure 5. Test application architecture diagram

The code logic for each of the three test applications is as follows:

In General:Hibernate for 1 second in the execute method of Boltl, hibernate for 1

second in the execute method of Bolt2, and set the initial parallelism of Spout, Boltl,

and Bolt2 to l; In General2:Hibernate 1 second in the execute method of Bolt1,

hibernate 3 seconds in the execute method of Bolt2, and set the initial parallelism of

Name/IP CPU/Memory/Disk Use

Dclab-1/192.168.

1.252

Intel(R) Xeon(R) CPU E5-2660 0

@ 2.20GHz / 3.6GB/42G Storm clouds zookeeper

Dclab-2/192. 168.

1.254

Intel(R) Xeon(R) CPU E5-2660 0

@ 220GHz/ 3.6GB/42G Tutor

Dclab-3/192.168.1.2

Intel(R) Xeon(R) CPU E5-2660 0

@ 220GHz/ 3.6GB/42G Tutor

Dclab-4/192.168.1.2

Intel(R) Xeon(R) CPU E5-2660 0 :

@ 220GHz/ 3.65GB/42.2G Tutor

Spout Bolt1 Bolt2

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

259

Spout, Bolt1, and Bolt2 to 1; General3:Hibernate for 3 seconds in the execute method

of Bolt1, hibernate for 1 second in the execute method of Bolt2, set the initial

parallelism of Spout to 1, and set the initial parallelism of Bolt1 and Bolt2 to 3. Submit

each of the above three Topologies to Storm for execution.

6.2. SYSTEM PERFORMANCE

In order to test the optimization of the parallelism dynamic adjustment module on

the use of system resources, we changed the Storm cluster to a single machine for

testing, and then we sampled the system information in 12 time periods. The first 6

sampling points Jmeter opened 2000 threads, and the last 6 sampling points Jmeter

opened 500 threads to send data. Then we record the CPU usage and memory usage

of the system at each of these 12 sampling points. First, we run Storm on the machine

without the parallelism dynamic adjustment module to sample the machine usage,

and then replace Storm with Storm with the parallelism dynamic adjustment module

and run it under the same conditions as before.

Figure 6. Usage Comparison Chart

As shown in Figure 6, in Generall, the processing time of two Bolts is the same, so

the parallelism ratio should be the same; in General2, the processing time of Bolt1 is

one-third of the processing time of Bolt2, and the parallelism ratio should be roughly

1:3; in General3, the processing time of Bolt1 is three times of the processing time of

Bolt2, and the parallelism ratio should be roughly 3:1. According to the change of the

number of Executors in the three applications in the figure: Generall remains the

same, and the ratio is still l:l: General2 increases by two, presumably because the

parallelism of Bolt2 increases from 1 to 3, and the ratio is 1:3; General3 decreases by

two, presumably because the parallelism of Bolt2 decreases from 3 to 1, and the ratio

is 3:1. That is, the parallelism between execution entities with sequential relationship

is dynamically adjusted according to the processing time proportionality. The

parallelism of execution entities is dynamically adjusted in the actual execution

process. By introducing the parallelism dynamic adjustment module (PDCM), the CPU

(a) CPU usage comparison (b) Memory Usage Comparison

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

260

usage and memory usage of the system increase (<2%) but not very significantly, but

when the real-time data volume is small, the parallelism dynamic adjustment module

can release the free memory resources (8%-11%) by stopping the idle Executor.

Through the smart sports platform route, on the one hand, users can independently

choose their favorite sports for physical exercise, which improves their motivation to

exercise independently; on the other hand, it facilitates the update of the sports

platform content.

6.3. SURVEY RESULTS

Through the construction of a smart sports platform, students are used as

experimental subjects in order to fully test the performance of the built platform. The

issue of individualization and differences of students is emphasized by giving students

the right to choose their own courses. At the same time, through the university's smart

sports platform, teachers can make timely adjustments to their teaching programs and

teaching plans to facilitate the completion of teaching tasks and teaching objectives.

（a）Statistical analysis of college students' ownership of mobile mobile terminal devices

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

261

（b）Survey on college students' access to sports information

（c）Survey on the purpose of college students' attention to sports information

Figure 7. College students' access to and purpose of sports information survey analysis chart

From Figure 7(a), it can be seen that among the mobile terminal devices owned by

college students, smart phones rank first, with almost one smart phone per person,

while the proportion of students who own laptops also reaches 62.30%. It can be seen

that the Internet access by smart mobile terminals has become one of the main

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

262

lifestyles of college students at present. From Fig. 7(b), it can be seen that through

building the intelligent sports platform in colleges and universities, the way of college

students getting information has changed dramatically, from traditional reading books

and watching TV programs to getting information through intelligent mobile terminals.

From Fig. 7(c), it can be seen that leisure and entertainment is the main purpose for

college students to pay attention to sports information, accounting for 58.6%, and

enjoying sports competition is another main purpose for college students to pay

attention to sports information, accounting for 51.00%; another more important reason

is to understand sports news, accounting for 49.2%. Thus, it can be seen that through

building a smart sports platform, college students start to become concerned about

sports information.

7. CONCLUSION

The emergence of smart sports to better provide convenient services for

participants. With a mobile client such as a smartphone or sports watch, people can

get reasonable sports advice anytime and anywhere. The smart sports platform

changes the way users exercise, and through interaction users can get more

reasonable and standardized exercise intensity and exercise combinations by

inputting their physical conditions and exercise patterns. In this paper, students are

taken as the experimental subjects, and according to the platform statistics it is known

that sports competition is the focus of college students' attention to sports information,

with a percentage of 51.00%. The percentage of understanding sports dynamics is

49.2%. Let more students understand the fun of physical education and physical

exercise through the smart sports platform under the machine learning model, and

also let teachers inspire more teaching inspiration through the platform and generate

good teaching interaction with students. Therefore, carrying out research on the

overall architecture and key technologies of smart sports, and seeking to translate and

apply its results in the fields of competitive sports and national fitness will certainly

play an important and positive role in the development of sports.

DATA AVAILABILITY

The data used to support the findings of this study are available from the

corresponding author upon request.

CONFLICT OF INTEREST

The authors declare that the research was conducted in the absence of any

commercial or financial relationships that could be construed as a potential conflict of

interest.

REFERENCES

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

263

(1) Wei, Y., Pan, D., Taleb, T., Han, Z., & Chen, X. (2016). An unlicensed taxi

identification model based on big data analysis. IEEE Transactions on

Intelligent Transportation Systems, 17(6), 1703-1713.

(2) R. A. de Assis, R. Pazim, M. C. Malavazi, P. P. da C. Petry, L. M. E. de Assis and

E. Venturino (2020). A Mathematical Model to describe the herd behaviour

considering group defense. Applied Mathematics and Nonlinear Sciences,

5(1), 11-24. https://doi.org/10.2478/amns.2020.1.00002

(3) Nallaperuma, D., Nawaratne, R., Bandaragoda, T., Singh, A. K., & Yang, B.

(2019). Online incremental machine learning platform for big data-driven

smart traffic management. IEEE Transactions on Intelligent Transportation

Systems, PP(99), 1-12.

(4) Jinxiang Xue. (2022). Design of language assisted learning model and

online learning system under the background of artificial intelligence.

Applied Mathematics and Nonlinear Sciences, (aop). https://doi.org/10.2478/

amns.2022.1.00027

(5) Mora-Sanchez, O. B., Lopez-Neri, E., Cedillo-Elias, E. J., & Rodriguez-Molano,

J. I. (2020). Validation of IoT infrastructure for the construction of smart

cities solutions on living lab platform. IEEE Transactions on Engineering

Management, PP(99), 1-10.

(6) Yukong Zhang, Hongwei Li and John D. Clark (2020). Experimental simulation

of mathematical learning process based on ‘chunk-objective’. Applied

Mathematics and Nonlinear Sciences, 5(2), 425-434. https://doi.org/10.2478/

amns.2020.2.00061

(7) Xiang, Y., Tian, X., Zhou, S., & Li, Y. (2020). Construction and application of

digital creative platform for digital creative industry based on smart city

concept. Computers & Electrical Engineering, 87, 106748.

(8) Liu, Y. (2022). Analysis of government public management information

service and computer model construction based on smart city

construction. Mathematical Problems in Engineering, 2022.

(9) Yilin Fan, Zheng Huo and Yaxin Huang (2022). Fed-UserPro: A user profile

construction method based on federated learning. Applied Mathematics and

Nonlinear Sciences, (aop). https://doi.org/10.2478/amns.2021.2.00188

(10) Su, L. X., Lyu, P. H., Yang, Z., Li, Y., & Cai, J. (2015). Scientometric cognitive

and evaluation on smart city related construction and building journals

data. Scientometrics.

(11) Kang, J., & Wang, X. (2020). The organizational structure and operational

logic of an urban smart governance information platform: Discussion on

the background of urban governance transformation in China. Complexity,

2020.

(12) Hui Jia, ISlam Nassar and Sarp Erkir (2022). English Learning Motivation of

College Students Based on probability Distribution. Applied Mathematics

and Nonlinear Sciences, (aop). https://doi.org/10.2478/amns.2022.2.0127

(13) Mora-Sanchez, O. B., Lopez-Neri, E., Cedillo-Elias, E. J., et al. (2020).

Validation of IoT Infrastructure for the Construction of Smart Cities

Solutions on Living Lab Platform. IEEE Transactions on Engineering

Management, PP(99), 1-10.

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

264

(14) Cho, O. (n.d.). Impact of Physical Education on Changes in Students'

Emotional Competence: A Meta-analysis. International Journal of Sports

Medicine.

(15) Chen, C., Peng, X., Li, Y., et al. (2021). Smart city community governance

based on blockchain big data platform. Journal of Intelligent and Fuzzy

Systems, 2021(2), 1-7.

(16) Xiang, Y., Tian, X., Zhou, S., et al. (2020). Construction and application of

digital creative platform for digital creative industry based on smart city

concept. Computers & Electrical Engineering, 87, 106748.

(17) Pal, A., & Hsieh, S. H. (2021). Deep-learning-based visual data analytics for

smart construction management. Automation in Construction,

131(November 2020), 103892.

(18) Machado, V. R., & Saldanha, M. (2015). The impact of sports mega-events on

health and environmental rights in the city of Rio de Janeiro, Brazil.

Cadernos De Saude Publica, 31 Suppl 1(suppl 1), 39-50.

(19) Escamilla-Fajardo, P., Alguacil, M., & García-Pascual, F. (2021). Business

model adaptation in Spanish sports clubs according to the perceived

context impact on the social cause performance. Sustainability, 13.

(20) Kaga, H. (2016). Positive Research on the Developmental process of

Modern Sports in Japan: on the Militarization of Meiji Jingu Athletic

Meeting. Computers & Geosciences, 91(194), 405-9.

(21) Zhuang, H., Zhang, J., Sivaparthipan, C. B., et al. (2020). Sustainable Smart

City Building Construction Methods. Sustainability, 12.

(22) Guan, S. (2020). Smart E-commerce logistics construction model based on

big data analytics. Journal of Intelligent and Fuzzy Systems, 40(2), 1-9.

(23) Li, W. (2019). Research on Intelligent Urban Management Information

System Framework Based on Big Data.

(24) Ma, Q., & Yang, Y. (2021). Research on the Construction of Smart Cities by

the Big Data Platform of the Blockchain. Journal of Physics: Conference

Series, 1883(1), 012144 (6pp).

(25) Zhao, Z., & Zhang, Y. (2020). Impact of Smart City Planning and

Construction on Economic and Social Benefits Based on Big Data

Analysis. Complexity, 2020(4), 1-11.

(26) Horani, M. O., Najeeb, M., & Saeed, A. (2021). Model electric car with

wireless charging using solar energy. 3C Tecnología. Glosas de innovación

aplicadas a la pyme, 10(4), 89-101. https://doi.org/10.17993/3ctecno/

2021.v10n4e40.89-101

(27) Chen, S., & Ren, Y. (2021). Small amplitude periodic solution of Hopf

Bifurcation Theorem for fractional differential equations of balance point in

group competitive martial arts. Applied Mathematics and Nonlinear Sciences,

7(1), 207-214. https://doi.org/10.2478/AMNS.2021.2.00152

https://doi.org/10.17993/3ctic.2023.121.248-265

3C TIC. Cuadernos de desarrollo aplicados a las TIC. ISSN: 2254-6529

Ed.42 | Iss.12 | N.1 January - March 2023

265