Data mining algorithm of experiential sports marketing based on cloud computing technology

Abstract

The internal connection and rule diversification of experience marketing data make it difficult to predict the future trend of data. Therefore, it is necessary to mine sports marketing data to guide future marketing strategies. In order to improve the effect of sports marketing data mining, this paper puts forward the algorithm research of experience sports marketing data mining in the cloud computing environment. In the cloud computing environment, based on the idea of data mining, a sports marketing monitoring system is designed and implemented to obtain a large number of evaluation data. The related data is extracted from the database of sports marketing evaluation system, and the data warehouse is constructed by data preprocessing. Using association rule algorithm to realize the data mining module of sports marketing evaluation system, mining the data in the data warehouse, dividing the data set into various data blocks, and then scanning each data block for association rule mining. The experimental results show that the mining algorithm can effectively mine different factors that affect the marketing status. The customer satisfaction obtained after the practical application of this method reaches more than 90%. Sports marketing enterprises can establish benign interaction between users and enterprises according to the mining results of this method, further meet the personalized and differentiated needs of consumers, thereby expanding the influence of enterprises and promoting the realization of marketing.

Keywords

Cloud computing experiential sports marketing data mining

1. Introduction

In recent years, with the continuous improvement of computer performance, the rapid popularization of computer information technology represented by the Internet, the rapid development of database technology and the wide application of database management system, people’s ability to produce, collect, store and process information and data has greatly improved, forming a large amount of data accumulation, and this trend will continue. The current database management system can efficiently input, query and count the massive data stored in it, but it cannot find the internal relations and rules hidden in the data, and cannot predict the future development trend according to the existing data [1].

It is difficult for people to extract valuable knowledge from massive data, which leads to the phenomenon of “rich data but poor knowledge”. Because of these requirements, in the late 1980s, data mining, DM for short and knowledge discovery in database came into being and developed rapidly. Data mining and knowledge discovery integrate database technology, statistics, artificial intelligence, pattern recognition, parallel computing, machine learning, visualization technology and other fields.

At present, China’s sports industry cluster is basically formed, and has 6 national industrial bases. However, in these clusters, the further cooperation and communication of industry, University and research still need some impetus. Under the background of big data, “data”, as a powerful driving force, will bring great changes to the sports industry in the collaborative development of production, learning and research. Driven by “data”, sports enterprises can not only share information from other enterprises, but also reduce the investment of scientific research funds; universities and scientific research institutions can input “data” talents for sports enterprises, while sports enterprises can provide a theoretical and practical transformation platform for universities and scientific research institutions [2]. Therefore, collaborative innovation is a win-win process for all parties. In order to better enable the buyer and the seller to exchange products, all marketing activities are driven by “data”, and the integration of this series of marketing activities driven by “data” is called “data” driven marketing mode. The “data” here is the data from the perspective of big data. Here, we need to distinguish two concepts, namely, the difference between business model and marketing model. “The general term of a certain kind of method adopted by an enterprise to realize the value orientation confirmed by the enterprise according to its business purpose.” It solves the development direction and strategy of the enterprise. Our common business models, such as e-commerce business model, include B2B, C2C, O2O, etc. In the process of marketing, marketing mode is the integration of the ways and methods that marketers use around product exchange. Common marketing models include experiential marketing, brand marketing, cultural marketing, and direct marketing and so on. From these two concepts, we can see that the perspective of business model is on the “enterprise planning”, while the perspective of marketing model is on the “product exchange”. Therefore, we can say that we study “the business model of sports industry” and “the marketing model of sports products”, not “the marketing model of sports industry” or “the business model of sports products”. There is an intersection between the two, but they cannot be confused.

Mobile communication technology is also becoming more and more perfect, which has a strong momentum of development and has a great influence on the public life. It has also become a business opportunity and an opportunity for enterprises to explore the market. However, Chinese enterprises lack systematic theoretical support to carry out marketing and are prone to fall into a dilemma. Therefore, only by mining the marketing customer data through artificial intelligence technology, analyzing and studying the marketing model of sports brands, and finding out the factors affecting the evaluation results of sports marketing, can the sports marketing ability and the number of sports marketing products be improved. Kauffmann et al. [3] proposed a data mining method for marketing product features based on natural language processing (NLP) technology. Using the emotion analysis, text data mining and clustering technology of NLP technology, the marketing data mining evaluation is carried out based on the emotion of different product characteristics of consumers. Combine product price and evaluation score to classify and rank marketing products, and extract big data input mining model from Amazon online review to realize data mining. Ordens and Silipo [4] designd a marketing data analysis platform based on machine learning. Visual programming software creates a real-time repository of projects and builds a marketing data analysis platform. Users can query data mining results through learning, sharing and visual code. Realize marketing data mining through customer churn mining analysis, emotion analysis, automatic image analysis, and search engine optimization. However, the data monitoring system is not established in the above research methods, the amount of data collected is insufficient, and the data mining results obtained are not ideal. In order to improve the effect of data mining, this paper proposes an experiential sports marketing data mining algorithm based on cloud computing technology. Innovatively in the cloud computing environment, based on the idea of data mining, a sports marketing monitoring system was designed and implemented innovatively to obtain a large number of evaluation data, and a data warehouse was built through data preprocessing. The data mining model of sports marketing evaluation system is established by using association rule algorithm, and the data in the data warehouse is mined. The data set is divided into different data blocks, and then each data block is scanned to complete association rule mining. The test results show that the mining algorithm can effectively mine the different factors that affect the marketing status, and the customer satisfaction obtained after practical application is high, which can provide reliable marketing strategy reference for sports marketing enterprises and improve the marketing rate.

2. Experiential sports marketing data mining in cloud computing environment

2.1 Cloud computing environment

Cloud computing technology is a computing platform that uses on-demand service mode, virtualization, load balancing, consistency management, network technology, distributed file storage and other technologies to connect multiple ordinary computers into a computing platform with high storage capacity, high computing power, high fault tolerance, and high scalability. Due to its characteristics of high scalability, low cost, high efficiency and high reliability [5, 6], this paper develops a distributed file system HDFS that stores large-scale sports marketing data sets based on the open source platform of cloud computing. First, develops a distributed file system HDFS that stores large-scale sports marketing data sets based on the open source platform of cloud computing. Then, based on the cloud computing development platform, the data mining algorithm is designed to determine that the mining object is the factor affecting marketing; Extract relevant data from the corresponding source database of the cloud environment, and integrate the data extracted from multiple data sources into a unified data warehouse. The fluctuation matrix of fuzzy abnormal data is calculated by ant colony algorithm. Finally, strong association rules are discovered through association rule mining to realize data mining.

HDFS is a distributed file system that can store large-scale data sets. It is developed based on the open source platform of cloud computing. Files are stored in the form of block data blocks, providing redundant copies for each data block. By default, block will be Data blocks are copied into three copies and stored in different storage locations in the cluster with specific policies [7]. When a block on the node is abnormal or lost, it is marked as damaged. Then, a new block is copied and stored on another data node using the intact copy block data block, and the invalid block is deleted and the name node is updated The mapping relationship between the upper data block and the data node storage node to ensure the data integrity in HDFS (see Fig. 1).

Figure 1.

HDFS file system architecture.

Name node is responsible for managing the namespace of files, the mapping relationship between files and data blocks, and the mapping relationship between data blocks and storage nodes; listening for the establishment and deletion of client’s namespace, the creation, read-write and deletion of files, the acquisition of file information and other events and executing corresponding commands; listening for data node, the end of the data block information, heartbeat response and error information and other requests are processed [8]. The data node is responsible for reading and writing block data blocks. DN receives the file read/write command from NN and executes it through the local file system client process; it reports the working status to NN regularly.

2.2 Data mining algorithm of experiential sports marketing

“Data” drives the “personalized” classification of the target market of sports products. In the process of marketing, marketers will treat different characteristics of sports consumer groups according to their needs, and finally determine different target markets through the division of the whole market. In the era of big data, “data” marketers should pay more attention to marketing products with the concept of “customers”, through the use of advanced data perception technology, the “data” of sports consumers is collected in an all-round way, so as to effectively classify the lifetime value of sports consumers’ customers, provide each sports consumer with “personalized” customized sports products or services, push different sports related information for them according to their preferences, and improve the loyalty of sports consumers, Increase the sports product exchange of every consumer to meet the sports needs of every consumer. In marketing, there is a famous saying about products: “you sell not a drill but a hole.” In the past marketing strategies, due to the difference of technical conditions and marketing ideas, most of them consider the marketing mix strategy from the perspective of the marketer, ignoring the needs of consumers [9]. In the background of “big data” era, due to the increasingly mature technical conditions, it is possible to “personalize customized” sports products for every sports consumer. Through the data generated by the sports consumers themselves, we can have a deeper insight into the needs of every sports consumer, some of which are hard to be found even by the consumers themselves. These are all characteristics of big data. Different from the acquisition and storage of traditional data, the focus of big data is on how to mine and process valuable data, so that it becomes an important resource for “data” marketers to make decisions. The “data” marketers can timely understand the needs and preferences of sports consumers and produce products more in line with the needs of sports consumers [10]. In the era of big data, enterprises can improve sports products by collecting “data” of sports consumers.

There are many ways to improve the application of sports products through consumer “data”, especially in the field of sports fitness and medical treatment. Enterprises can acquire “body data” from consumers’ bodies as much as possible with advanced sensors, use consumers’ commonly used social software to obtain online relevant information, integrate these messy data through big data processing technology, monitor users’ sports health status in real time, and make corresponding suggestions or prescriptions in real time. Therefore, in the future, the marketing of sports products will be revolutionized. It is no longer a fantasy to improve the production of sports products by collecting sports consumer data to meet the needs of consumers [11, 12]. Furthermore, using SPSS software is one of the most widely used and effective mathematical analysis methods. Cluster analysis refers to the analysis process of grouping the set of physical or abstract objects into multiple classes composed of similar objects, which is also an important human behavior. The clustering method can be divided into hierarchical clustering and nonhierarchical clustering. In this paper, hierarchical clustering is a technology for static data analysis, which is widely used in many fields, including machine learning, data mining, pattern recognition, image analysis and biological information. Clustering is to divide similar objects into different groups or more subsets by static classification, so that the member objects in the same subset have similar attributes, including shorter space distance in coordinate system. The core of building the evaluation system of “data” driven sports products is to build a comprehensive analysis model centered on the evaluation of sports consumer demand satisfaction. The establishment of evaluation mechanism is for better management. Through the evaluation of the health, sports consciousness, satisfaction and loyalty of sports consumers, as well as the overall effect of data marketing, the comprehensive collection of data, the establishment of a special evaluation team to evaluate the marketing benefits, to better control the marketing process of sports products (see Fig. 2).

Figure 2.

Marketing steps of sports products.

The 18 factors that affect marketing are divided into three categories. Generally, 15 is the default landing frequency, that is, the basis for selecting cluster analysis classification results. The first category is the activity participation category, which specifically includes age, gender, whether you have an account number, purchase method of sporting goods, whether you pay attention to the enterprise, what advertisements will attract the eyes, whether you pay attention to the sports enterprise, how you use them, and whether you share the purchase The second is the frequency of sports, only the frequency of sports; The third category is related information category, including the use time, whether to pay attention to advertising, advertising credibility, advertising attitude towards friends and information trust bias [13]. The final result of cluster analysis is very consistent with the reality. With the growing popularity, the factors that affect the current situation of marketing can be divided into the above three categories. From the actual situation, the first type of activities to participate in the impact of factors on the use of attitude and use; Such factors as whether to pay attention to the enterprise, the time of landing and whether to participate in the interaction determine whether the user has the possibility of becoming a potential customer of the enterprise. Only those who are willing to participate in the interaction, are willing to pay attention to the enterprise and have a long time of landing can easily develop into a potential customer group. The second category is the frequency of participating in sports, in which the people who like to participate in sports are generally loyal customers of sports goods, with high loyalty. Therefore, such people should be trained as key customers of marketing, while the customers who have the intention and idea of exercising become potential customers. This category has a large number of customers and huge potential. How to develop them into loyal customers is also a matter. The third category of relevant information factors shows the impact of the relevant information on marketing. The more specific the relevant information is, the easier it is to form information symmetry with consumers [14]. Then, it is helpful to further meet the personalized and differentiated needs of consumers, so as to promote the realization of marketing.

In the data preparation stage, the first step is to determine the goal of mining. Data mining personnel must repeatedly communicate with domain experts and end users, understand the tasks to be handled by data mining, be familiar with the background knowledge of relevant fields, and make mining plans [15]. Secondly, we need to determine the data source needed for mining, and then extract the relevant data from the corresponding source database. Finally, we need to integrate the data and store the data extracted from multiple data sources in a unified data warehouse. If we want to make an inertial system an open system with acceleration, we can change its internal structure. In order to transform the inertial system by transforming the internal part of the system, we should consider the following points:

First, reform the system of maintaining internal balance structure; Second, improve the quality of the internal elements of the system; Third, improve the potential energy difference in the system.

If $a_{1}$ and $a_{2}$ represent the acceleration of two objects, and $F_{1}$ and $F_{2}$ represent the external force acting on two objects, there are:

$\displaystyle{a}_{1}/{a}_{2}=\sum{F_{1}}/F_{2}$ (1)

This formula shows that when the mass is the same, the acceleration of the object increases with the increase of the external force acting on it, that is to say, the larger the external force is, the greater the acceleration is. The sports product marketing system is also a non-linear system. By introducing this conclusion into the system, it can also be explained that the exchange speed of sports products depends on the marketing system of sports products and the external forces acting on it. If $m_{1}$ and $m_{2}$ are the masses of two objects, they are:

$\displaystyle\sum{{a}_{1}/{a}_{2}}=\sum{\mathop{\lim}\limits_{0\to\infty}m_{1}% }/{m}_{2}$ (2)

This shows that the acceleration of two objects is inversely proportional to the mass when the external forces are the same. In the sports product marketing system, we can understand that the acceleration of the internal structure development of the marketing system is inversely proportional to the scale of the marketing system ${v}_{i}^{n-1}=\|{x_{ij}-m_{ij}^{n-1}}\|^{n}$ , Furthermore, the time series value $u$ and the flow fluctuation value $n$ of network traffic play are normalized. If the mining window length is $k$ , the data flow fluctuation sequence is recorded as $t=\{{{v}_{i}^{n-1}|{i=1,2,3,\ldots,n=i+1}}\}$ . If the detected network traffic characteristic value is generally large, it means that the network traffic value is fuzzy data and abnormal data. In order to better quantify the stability and abnormal value of network traffic data, the fluctuation matrix of fuzzy abnormal data is calculated by ant colony algorithm. The specific algorithm is as follows:

$\displaystyle l=\left\{{\begin{array}[]{l}f(x)=\sum{\sum{\frac{tw_{i}^{n}{v}_{% i}^{n-1}-1}{2\sum{{a}_{1}/{a}_{2}}\ast\sum{m_{1}}/{m}_{2}}}}\\ g(y)={a}_{1}/{a}_{2}\ast\sum{u\|{i-nk}\|F_{1}}/F_{2}\\ \end{array}}\right.$ (3)

If $\beta$ , $\eta$ are the reference value of the network traffic data respectively, the abnormal values mined will be standardized according to the maximum similarity principle, and the abnormal value fluctuation coefficient will be further standardized according to the distribution function operation principle. The specific algorithm is as follows:

$\displaystyle R=l^{-1}\left({a}_{1}/{a}_{2}+f(x)\left|{\sum{\mathop{\lim}% \limits_{0\to\infty}m_{1}}/{m}_{2}\ast g(y)-\eta\beta}\right.\right)$ (4)

Usually we need to find the strong association rules defined above through association rules mining. The mining algorithm of association rules is mainly divided into the following two steps:

(1)

Find frequent item set: find all frequent item set a with given minimum support;

(2)

Generate strong association rule: use the given minimum credibility to find strong association rule B.

Since the second step is based on the first step, the amount of data and calculation is not large compared with the first step, and there is little room for improvement. Therefore, the design of the first step algorithm is the focus of association rule mining research. The calculation formula of A and B correlation coefficients is calculated. The specific algorithm is as follows:

$\displaystyle A=\sum{\frac{\bigcap\limits_{\lim 0}{P(0)/P(1)}}{\oiiint{{w}_{i}% ^{n}{v}_{i}^{n-1}-\sum{u\left\|{i-nk}\right\|F_{1}}/F_{2}}}-1}$ (5) $\displaystyle B=\sum{\frac{\bigcap\limits_{\lim 1}{P(1+)/P(0+)}}{\mathop{\int% \!\!\!\int\!\!\!\int}\limits_{\kern-5.5pt{\lim 0}}{{v}_{i}^{n-1}\sqrt{{a}_{1}/% {a}_{2}+f(x)}}\left\|{x_{ij}-m_{ij}^{n-1}}\right\|^{n}}}$ (6)

If $P(i)$ represents the number of objects satisfying the $A=j\&\&B=N$ condition $({i=0,1;j=0,1})$ . In addition, $P(i+)$ represents the number of objects satisfying $A=j$ conditions, and $P(j+)$ represents the number of objects satisfying $B=j$ conditions. The relationship between $P(i+)$ , $P(j+)$ and $P(ij)$ can be expressed as follows:

$\displaystyle\Delta u=\sum\limits_{\lim=0}{(A+B)\frac{\prod B\oiint\limits_{% \lim=1}{p(ij)}-p(i)}{\bigcup{A\mathop{\int\!\!\!\int\!\!\!\int}\limits_{\kern-% 5.5pt{\lim=1}}{\log}P(i+)+P(j+)}}-1}$ (7)

An important idea of designing parallel clustering algorithm in cloud computing environment is data parallelism, that is, dividing the whole data set into several subsets, using the same clustering algorithm for each subset, clustering independently on a single node. But after the data set is divided, the data that belong to the same class will be divided into different subsets. The data set is divided into data blocks, and then each data block is scanned for association rule mining.

3. Experimental study

Taking the marketing mode of sports products based on big data technology as the research object, this paper analyzes the current research and development of big data by searching the latest big data at home and abroad and relevant literature of marketing mode, and tries to analyze and build the marketing mode of sports products based on big data technology. The test is to install VMware virtualization software on one server, and create Hadoop and spark platforms composed of six virtual machines (two master nodes and four slave nodes). The server parameters are as follows: Intel (R) Core i7 6700k @ 4.00 GHz processor, 32g memory, 2T hard disk space.

The author searches the contents of big data phase on CNKI, Baidu and other network resources, and borrows relevant books in qushida library to collect various forms of literature. And draw the average link tree diagram of sports marketing, as follows (see Fig. 3).

Figure 3.

Average link tree of sports marketing.

In the form of online questionnaire, I asked my friends to distribute the questionnaires to their friends and make sure that they were completed. Because some of the interviewees did not understand each other, they did not take back the questionnaires successfully. Finally, 110 questionnaires were distributed and 101 questionnaires were recovered, with a recovery rate of 91.8%, all of them were effective questionnaires, with an efficiency of 100%, In order to ensure the recovery rate and coverage of the questionnaire, this questionnaire imitates the form of six degree separated online questionnaire. The specific operation method is as follows: distribute the designed questionnaire to 20 different ages and occupations and in different places, and ask them to fill in the questionnaire carefully, and then spread the questionnaire to their five friends of different occupations and ages. This way can not only ensure the effective recovery of the questionnaire, but also a relatively wide range of respondents, including students and official business Staff, teachers, business people and other professions, age from I8 to 50 (because the relative age of the group is relatively young, the non-reference is not strong, so there is no over age group as the interviewee), which covers a wide range, can comprehensively represent the opinions of Internet users, and has a high reliability. Based on this, first of all, this paper investigates the change of sales volume of sports products in recent years in China. In the process of building the model, it analyzes and summarizes each link of the “data” driven marketing model through different ways and methods used in the marketing model of various cases. The details are as follows (see Fig. 4).

Figure 4.

Investigation on sales volume of sports products.

Based on the above findings, the “data” driven marketing model is further regarded as a general system, while the marketing of market analysis, target market positioning, marketing mix strategy and marketing management activities can be regarded as the “data” driving for the subsystem of this marketing system.

Based on this, the paper further integrates the factor information in the questionnaire, including 18 elements, which are gender, age, whether you have an account, use time, frequency of participating in sports, way of purchasing sports goods, whether you pay attention to the enterprise, whether you pay attention to the advertisement, what kind of advertisement can attract attention, advertising credibility, use mode, and wide release to friends The attitude of informing information, whether to pay attention to sports brand enterprises, whether to publish shopping experience and experience, whether to participate in product interaction, whether to trust the type of information bias, whether to support sports brand marketing; These 18 projects cover the basic information of the interviewed users, the user’s usage habits, and the factors that the user’s acceptance of the marketing methods will have an impact on the marketing. The purpose of these projects is to study how the enterprise can quickly and effectively transfer information, and establish a good interaction between the user and the enterprise, so as to expand the enterprise’s influence; This kind of marketing method should be more active in customer’s user experience, so as to better improve customer satisfaction and customer loyalty. The personal experience can provide more practical strategies for marketing, and the operational marketing strategies can also provide strong theoretical support for enterprises to carry out marketing. Through the analysis of the relevant situation of the respondents, only the age has significant difference. See the table for the specific situation. It can be seen from the above table that the age distribution is mainly between 18 and 30 years old, and the age distribution is characterized by youth. Of all the questionnaires, 260 were male and 260 were female.

Table 1

Information collection of respondents

Project	Content	Number	Proportion of people who exercise regularly	Proportion of occasional exercise	Proportion of people who do not exercise
Gender	Male	260	8.2%	5.0%	6.2%
	Female sex	260	9.1%	4.6%	5.4%
Age	Under 18 years old	63	6.5%	8.2%	3.1%
	18–25	87	3.8%	6.0%	4.6%
	25–30	127	3.6%	4.1%	5.0%
	30–35	73	6.5%	7.1%	6.6%
	35–40	82	4.6%	6.3%	5.1%
	Over 40 years old	88	4.5%	5.9%	4.7%

Table 2

Investigation on purchase of sporting goods

Feature description	Not buy	Online shopping	Physical store	Support degree %	Confidence level %
tid	20/2/1	0.890	0.662	60	71
item	60/5/9	0.612	0.510	63	68
Lid	12/3/8	0.541	0.648	58	52
Cnt	40/6/7	0.535	0.814	68	71
Nid	0/16/3	0.648	0.712	66	59
Link	6/5/2	0.647	0.481	59	63
f-t	17/3/8	0.910	0.648	94	64
Size	21/5/4	0.510	0.658	58	84
Nld	13/20/0	0.645	0.810	68	72
jslb	29/3/11	0.481	0.948	74	70
fsd	41/3/19	0.518	0.024	66	68

The way of cluster analysis is because cluster analysis is a process of dividing abstract object sets into multiple categories composed of similar objects. For the research of this paper, it is of great significance to analyze the factors that will affect the marketing effect and result of sporting goods, and to put forward feasible marketing strategies. Clustering the factors that affect the marketing effect can ensure that The pertinence and feasibility of the strategy can be used for reference to the influence of sports goods enterprises, and promote the development of sports goods marketing and sports goods enterprises in China. The quantitative analysis of the influencing factors of sports goods marketing can point out the direction for sports goods enterprises to gain advantages in marketing development and competition. This paper tries to draw scientific conclusions through data analysis, so that sports goods enterprises can adjust marketing methods in time to improve their market competitiveness. The concept of relational database has been familiar to most people. The main task of database system is to perform online transactions and query processing, and to query and update data content. In order to improve the efficiency of transaction processing, the design of table structure in database system should be as simple as possible, and the data should be as few as possible. The purpose of data warehouse is to facilitate query. From the perspective of data analysis, it can provide different forms of organized data to meet the query needs of different users. The data in the data warehouse should be as much as possible, and the data of the same subject should be put together. Relational database system is called on-line transaction processing (OLTP), and data warehouse system is called on-line analytical processing (OLAP).

Table 3

Comparison of data feature mining

	OLTP	OLAP
Features	Operation processing affair	Information processing
Face	Operator	Analysis
User	Database manager	Decision maker
Function	Daily operation processing	Senior management analyst
DB design	Based on E-R	analysis of resolution
Data	Application oriented current	Supports star/snowflake models
Access	Latest, detailed	Theme oriented, aggregated
Work unit	Two-dimensional, discrete	Multidimensional, integrated
User number	Read and write operations	Mostly read
DB scale	Simple transactions	Complex queries
First	Thousands of 100MB to GB	Hundreds of 100GB to TB
Quality	High performance, high availability	High flexibility
Summary	Transaction huff and puff	Query throughput, response time

Through the questionnaire survey and data statistical analysis, according to the different factors that affect the current situation of marketing can be divided into the above three categories. From the actual situation, the first type of activities to participate in the impact of factors on the use of attitude and use; Such factors as whether to pay attention to the enterprise, the time of landing and whether to participate in the interaction determine whether the user has the possibility of becoming a potential customer of the enterprise. Only those who are willing to participate in the interaction, are willing to pay attention to the enterprise and have a long time of landing can easily develop into a potential customer group. The second category is the frequency of users’ participation in sports. The users who like to participate in sports are usually loyal customers of sports goods with high loyalty. Therefore, such people should be trained as key customers of marketing, while the customers who have intention and idea of exercise become potential customers. This category has a large number of customers and huge potential, and will develop them into loyal customers. It should lie in the precise positioning of marketing and targeted interaction; The third kind of related information influence factors explain the influence of related information on marketing. The more specific the related information is, the more likely it is to form information symmetry with consumers. This requires that the construction of content should be scientific, reasonable, attractive and attractive, which will help to further meet the personalized and differentiated needs of consumers, thus promoting the realization of marketing.

Using the method as experimental group, literature [3] and literature [4] as control group 1 and control group 2, respectively, mined the effectiveness of the results using customer satisfaction detection methods. The comparison test results are shown in the following figure.

Figure 5.

Comparison of satisfaction with the test results.

It can be seen from Fig. 5 that the results of the marketing influencing factors mined by this method are effective, with customer satisfaction reaching more than 90% and high, while the other two methods have low customer satisfaction and poor mining results.

4. Discussion

According to the analysis results and case details, it can be seen that “data” provides sports consumers with a “multi-channel, multi-way” marketing model when building sports product channels. When consumers enjoy sports products, they can accept sports products including words, pictures, videos, etc. through all-round channels such as TV, radio, Internet, media, etc. With the continuous innovation of “data” technology, mobile devices also provide users with more different experiences. According to Alexa, a famous website ranking system, during the London Olympic Games, Sina, Tencent and Sohu’s daily average views of Olympic topics reached tens of millions of levels in a month. Among them, sina sports released 36769 text news, 20129 high-definition pictures and 5812 video news, and all news posts reached an amazing number. During the London Olympic Games, CNTV has an average of nearly 600 million page views per day, with an average of more than 35 million visitors per day. Users broadcast CNTV more than 600 million times through various terminal channels. The development of “data” technology has changed the way people interact with sports platforms. Consumers can obtain sports products through different channels and have more brand-new experiences, which will be more profoundly changed in the future development.

This paper analyzes the current sports marketing problems, including:

(1)
Lack of independent innovation strength, lack of high-tech content, the majority of sporting goods in China are still processed by agent and raw materials, simple imitation or even “imitation” of foreign famous brands, lack of brand awareness, low cultural value of products. Therefore, the competitiveness of Chinese sports brands is still weak compared with foreign famous brands.
(2)
Brand awareness is poor, which leads to the loss of domestic brands’ advantages in popularity and price. How to expand the popularity is the biggest problem faced by China’s sporting goods marketing; In addition to the famous brands such as Li Ning and Anta, there are only a few other well-known brands in China’s sports brand advertising, which reflects that in China, the publicity of sports products is obviously insufficient, and the way of publicity is single, only relying on TV advertising to obtain large-scale consumer groups is also the result of the lack of research on the consumer market. Sports goods manufacturing industry is generally regarded as one of the high-tech industries in the world. Many countries invest a lot of R & D funds in sports industry, constantly increasing the scientific and technological content of products and increasing the added value of products. Therefore, China’s sports goods should also pay attention to independent R & D capacity, invest scientific research funds, constantly develop new products, improve the scientific and technological content of products, and improve competitiveness from its own quality This is the most important way.
(3)
From the perspective of traditional marketing, the research on the marketing strategy of sports products is mostly based on the analysis and discussion of enterprises. To a certain extent, it lays a foundation for the theoretical discussion and further research on the improvement of sports products and services. However, the existing research on the traditional marketing model for sports products and services with the characteristics of public goods and services will have some conceptual definitions and theoretical conflicts. At the same time, the field of sports industry is lack of relevant comprehensive statistics.
(4)
While paying attention to the quality of products, we should change the marketing mode, so as to make the marketing of Chinese sports brand “better”. The customers of sports goods are people who pay attention to physical exercise and often participate in sports activities, including men, women, old and young people of all ages, but they all have something in common. Research shows that people who often participate in physical exercise have a mentality five to ten years younger than their peers. Similarly, people who often take part in sports are optimistic, sunny and willing to accept new things. People with this personality are usually happy to make friends, which is highly consistent with the people who use them. At the same time, through the classification and participation in topics, users can accurately distinguish whether they like sports or are potential customers of sporting goods. When marketing sporting goods, enterprises should pay more attention to the choice of target customer groups and try to achieve accurate marketing.
(5)
Storage of big data. Various types of unstructured and semi-structured data, such as video, audio, social and so on, are growing rapidly, and TB level data may be generated every day. The continuous growth of these “data” has brought great challenges to big data storage.
(6)
The security of sports information. Enterprises use perceptual technology to collect consumers’ personal “data” and make data processing, but if there is no relevant laws and regulations to limit it, enterprises for economic interests, making consumers’ personal privacy will be exposed.
(7)
Short board of big data talents. When analyzing and processing “data”, we need to have a very professional technical ability as support, and sports enterprise is also an enterprise with strong professional requirements, which requires not only sports knowledge, but also compound talents who master it professional technology.

The traditional marketing theory tells us that the concept of product as a whole has five levels, and sports products also have these five levels. The reason why consumers need sports is not because they need sports themselves, so sports marketers and producers of sports products must understand the real needs of sports consumers and potential consumers. This paper holds that the analysis of the needs of sports consumers should start from the spiritual and physical aspects. In terms of body, what sports consumers need is “health”. Whether tangible or intangible sports products, their essence is to meet the needs of consumers for their own health. For example, the national fitness center in Jinan provides venues and facilities, and sports consumers can play badminton in the venues, which is an intangible sports service product. Playing badminton is a whole-body sport. Scientific exercise can relax muscles and activate blood circulation, prevent and improve cervical spondylosis, and also play a role in reducing weight. These are actually the needs of consumers for “health”. For the badminton racket and protective equipment purchased for badminton, these tangible sports products are to meet the needs of consumers for badminton, as well as for “health”. On the other hand, sports consumers also need “spiritual needs”. The holding of the Beijing Olympic Games inspires the Chinese people. Sports consumers need this “sense of honor”; playing with friends is a need for “sense of belonging” and “sense of identity”; in order to win a game, the actual need is this “sense of achievement”. The human emotion is too complex, which needs to be discussed more systematically.

Through literature research and logical analysis, we can know that the number of users in China is large and the stability is strong. Sports goods enterprises have a large number of potential customers to carry out marketing. Marketing is one of the important marketing methods that sports goods enterprises can adopt. At present, Chinese sports products enterprises are faced with the problems of low production standardization, lack of brand awareness and lagging development of brand. Through this platform, we can expand brand influence, enhance brand awareness and promote the development of sports products enterprises’ brand. Through the successful marketing cases of wiggle, the largest sporting goods e-commerce in UK, in China, it can be seen that sporting goods enterprises should pay attention to the cultivation of professional teams, open different, build different contents and achieve the best marketing effect for different groups of people. It should be realized that marketing is not simply to send advertisements, but to combine the user’s usage habits, social events or heat Point topic development, using the characteristics of information viral communication.

The potential customers of sporting goods are those who like sports and insist on taking part in physical exercises. Their demand for sporting goods is active and purposeful. The purpose of sporting goods marketing is to attract customers’ attention to their own brands or products, which requires enterprises and users to establish an interactive mechanism to deeply understand customers’ needs. The purpose of marketing is that enterprises can search the related customers through the keywords sent by users, grasp the user’s dynamic situation immediately, and lock some consumers who are not sports goods consumers through keywords. However, the ultimate goal of marketing is to develop potential customers into loyal customers by positioning potential customers, which requires comprehensive use of keywords and targets Sign, square and other functions, precise positioning can improve the efficiency and effect of marketing. There are many differences on the definition of sports product concept. From the theoretical point of view, it can be basically divided into two views. The first view is that sports products are intangible, that is, sports products are sports services. The data warehouse reflects the content of historical data, not the data generated by daily transaction processing. Once the data is processed and integrated into the data warehouse, its data is little or no need to change. The most fundamental characteristic of data warehouse is the physical low storage of data, and these data are not up-to-date and proprietary, they come from other databases. The establishment of data warehouse is not to replace the traditional database supporting transactional processing, but to establish a more comprehensive and perfect information application environment supporting the application of high-level decision analysis. Although the data warehouse can be realized by both relational database and object database, because the relational database has been fully mature and very popular both in theory and in practice, it is mainly realized by the relational database management system at present.

Summary of the first view:

1.
Sports products are tangible sports material products and intangible sports service products provided by the sports market to meet the needs of customers;
2.
Sports material products exist because of the existence of sports intangible products, and sports intangible products can be realized because of sports material products;
3.
Sports products can be divided into commodity and non-commodity. Sports products of commodity nature can be exchanged to meet the needs of consumers, while sports products of non-commodity nature cannot be exchanged.

Summary of the second view:

1.
Sports products refer to sports services;
2.
Sports products refer to the sports services consumed in the sports activities that the rest of the demanders want to buy from the sports producers;
3.
Sports products are produced in the physical activities of the sportsmen.

5. Conclusion

In the cloud computing environment, this paper puts forward an experiential sports marketing data mining algorithm. First, it describes the current situation of sports goods company marketing development, and then through data analysis and case analysis, it further analyzes the marketing mode and experience of sports goods company. This study combines theoretical analysis, data analysis and case analysis to carry out business for sports goods company The correct mode of marketing is summarized, and the specific theoretical basis and countermeasures are provided for the marketing of sports goods. However, due to the limited conditions, this study only improves the applicability of sports marketing data mining, and does not significantly improve the mining speed. Future research can further improve mining efficiency while ensuring the reliability of mining results.

Footnotes

Funding

This work was supported by Zhejiang Provincial Federation of Social Sciences Research Project: “Research on the mechanism and countermeasures of coastal sports tourism to improve residents’ happiness in Zhejiang Province” (2022N63).

References

Salami

Bala

Sait

Ismail

. An energy-efficient cuckoo search algorithm for virtual machine placement in cloud computing data centers. J Supercomput. 2021; 77(11): 1333013357.

Muñoz

Vilalta

Yoshikane

Casellas

Martinez

Tsuritani

, et al. Integration of IoT, transport SDN, and edge/cloud computing for dynamic distribution of IoT analytics and efficient use of network resources. J Lightwave Technol. 2018; 36(7): 1420-1428.

Kauffmann

Peral

Gil

Ferrández

Sellers

Mora

. Managing marketing decision-making with sentiment analysis: An evaluation of the main product features using text data mining. Sustainability. 2019; 11(15): 4235.

Ordenes

Silipo

. Machine learning for marketing on the KNIME Hub: The development of a live repository for marketing applications. J Bus Res. 2021; 137(12): 393-410.

de Moraes

Schaefer

Schreiber

JNC

Thomas

Nara

EOB

. Algorithm applied: attracting MSEs to business associations. J Bus Ind Mark. 2019; 35(1): 1322.

Chatterjee

Goyal

Prakash

Sharma

. Exploring healthcare/health-product ecommerce satisfaction: A text mining and machine learning application. J Bus Res. 2021; 131(7): 815825.

Esser

Jernigan

. Policy approaches for regulating alcohol marketing in a global context: A public health perspective. Annu Rev Publ Health. 2018; 39(1): 385-401.

Overgoor

Chica

Rand

Weishampel

. Letting the computers take over: Using AI to solve marketing problems. Calif Manage Rev. 2019; 61(4): 156185.

Timoshenko

Hauser

. Identifying customer needs from user-generated content. Market Sci. 2019; 38(1): 120.

10.

Ahmad

Guzmán

. Negative online reviews, brand equity and emotional contagion. Eur J Marketing. 2021; 55(11): 28252870.

11.

Seo

Itoh

. Perceptions of customers as sustained competitive advantages of global marketing airline alliances: A hybrid text mining approach. Sustainability. 2020; 12(15): 1-15.

12.

Luo

Qiu

Pan

Yang

Tian

. Exploration of spa leisure consumption sentiment towards different holidays and different cities through online reviews: Implications for customer segmentation. Sustainability. 2022; 14(2): 664.

13.

Ban

Choi

Lee

Kim

. Investigating key attributes in experience and satisfaction of hotel customer using online review data. Sustainability. 2019; 11(23): 6570.

14.

Zaki

Mccoll-Kennedy

. Text mining analysis roadmap (TMAR) for service research. Journal of Services Marketing. 2019; 34(1): 3047.

15.

Huang

Yuan

Zhou

Zheng

. Analysis and evaluation of heat source data of large-scale heating system based on descriptive data mining techniques. Energy. 2022; 251(7): 123834.1123834.18.