Role of AI for application of marketing teaching –A research perspective

2.1 Development of artificial intelligence technology in marketing

According to the data of the National Bureau of statistics, in 2018, the national marketing turnover reached 31.63 trillion yuan, including 30.61 trillion yuan for goods and services [8]. In 2018, the contribution rate of China’s service industry to GDP growth reached 59.7%. As an important part of modern service industry, the marketing service industry continues to maintain a steady growth trend, the market scale continues to expand, the market structure is further optimized, and new technologies, businesses and models continue to emerge, which has become an important driving force to promote the integration of marketing and industry [9]. In 2018, China’s marketing service industry continued to maintain steady growth, and the market scale further expanded. The annual operating revenue of the marketing service industry was 3.52 trillion yuan, an increase of 20.3% year on year. China’s marketing service industry has supported 31.63 trillion yuan of marketing transaction volume. The marketing service industry, such as transaction service, information technology service, advertising creativity, marketing logistics, intelligent marketing, consultation and training, has become increasingly diversified and specialized, and constantly promotes the innovative development of marketing [10].

With the in-depth application of big data and artificial intelligence, marketing is gradually realizing precision marketing, intelligent marketing and creating greater business value. According to AI media consulting data, in 2017, China’s Internet advertising market scale reached 30.1 billion yuan, and in 2018, it reached 350.9 billion yuan, a year-on-year increase of 16.6%; in 2020, it is expected that China’s Internet advertising market scale will reach 441.48 billion yuan, and the proportion of mobile advertising will continue to rise, known as the mainstream [11]. 76.5% of Chinese mobile Internet users said that the Internet was the main channel for them to contact with marketing advertisements, while the influence of traditional media gradually declined. With the rapid development of the marketing industry, the marketing war in the industry continues to be hot, and the information exposure rate of the marketing industry remains high among the netizens. The marketing based on big data and artificial intelligence effectively realizes personalized, precise, intelligent and interactive marketing. The three effects of big data marketing and intelligent marketing are also relatively recognized by Chinese netizens.

With the popularity of smart phones and the deep coverage of the logistics industry, the serious shortage of marketing talents began to highlight [12], among which the most popular is marketing technical service talents. As a strategic emerging industry of the country, the marketing specialty has played an important role in leading the change of business model, optimizing industrial transformation and upgrading, improving the demand for information consumption, promoting the development of modern service industry and information economy, becoming a new driving force for economic development, and providing a new space for mass entrepreneurship and innovation. As an industry application-oriented, subject integration and model innovation major, the experiment, practical training and practice teaching are the necessary teaching contents throughout the whole process of marketing undergraduate teaching. There are many employment opportunities for college graduates majoring in marketing. They can be engaged in website design, website maintenance, network editing, enterprise commodity marketing planning, customer management, marketing management, marketing activity planning and operation, market marketing system development and maintenance, etc.

2.2 Marketing technology under the background of artificial intelligence

Teachers lack practical experience and the ranks of practical teachers are not perfect. Under the background of artificial intelligence, China is in a critical period of rapid economic development and continuous innovation of the real industrial structure and production mode [13]. In the current innovation period, new requirements are put forward for posts in the industry. The demand for innovative, compound and productive talents is increasing rapidly. Teachers of marketing major are generally rich in theoretical knowledge and lack of practical experience. This leads to the fact that the teachers of marketing can’t teach the practical cases for the students, and spend most of their time on the simulation teaching software, which deviates from the progress and development of the technology of the times.

2.3 Training mode of marketing technical service talents

We will improve teachers’ practical skills and build a “double teacher” teaching staff. The application-oriented colleges and universities should actively innovate the training mode of marketing technical service talents and improve the practical skills of teachers. Regularly organize marketing professional teachers to participate in the practical activities of enterprises, improve teachers’ understanding of practical skills, and increase teachers’ working experience in enterprises. Application-oriented undergraduate colleges can also hire senior managers and professional technicians to teach in schools, promote the communication and exchange between school teachers and enterprise personnel, and improve the practical skills of teachers.

Insist on promoting the connection between specialty and industry, systematically design the practical training teaching system, integrate the practical training resources inside and outside the school, make full use of the advantages of network technology and digital resources, introduce the real projects of the industry, insist on the combination of teaching and practical combat, and cultivate the core mainstream media resources with solid professional foundation, proficient basic skills, strong practical ability, innovative spirit and understanding of global overseas Promotion mode, with the export-oriented enterprise digital marketing multi-channel overseas social media account management, real-time data perspective and intelligent report management capabilities of marketing senior application talents.

3.1 Neural network model

From the perspective of modern marketing teaching system, colleges and universities have made in-depth research on the learning system based on neural network, with the help of quantitative data, so as to realize the understanding of marketing system, social relations, modern rational quantification, rationality, knowledge system, values, through neural network learning and training [15]; the next step is to determine the results according to information processing, Predict the development trend, and put forward the corresponding methods: including the appropriate reform of the dynamic adjustment of the teaching methods of Ideological and political courses in Colleges and universities, the role of demonstration, the reasonable application of network media, the exemplary role of students, the harmonious construction of campus culture and other measures [16]. BP neural network is a kind of bionic algorithm, which has two main characteristics: the first is information transmission, the second is error back-propagation, there is no interaction between neurons and change the value with the genetic effect, repeated until the required error, the training meets the expected proportion range matrix.

As shown in Fig. 1 is the structure diagram of the neural network, where X₁, X₁, . . . , X _n represents the input value of forward propagation, Y₁, Y₂, . . . , Y _m represents the value of back propagation, w _ij and w _ik represent the parameters of back propagation. The back-propagation neural network represented in this figure is a nonlinear function, the independent variable is the input value of the network, and the dependent variable is the output value of the network. A function from n to m dimension is constructed. Using neural network to realize system data standardization and network intelligence, the implementation steps are as follows [17]:

OPEN IN VIEWER

Fig. 1

Structure of neural network.

Step 1: Network initialization. According to the type, the (X, Y) algorithm determines the number of nodes n, the number of hidden layer nodes L, and the number of printing layer nodes M. The connection and scale a between the type layer w _ij and the print layer w _ik neurons, the invisible layer range B, the print range and the given access rate are initialized.

Step 2: Hide layer output. According to matrix X, Y, the type of node number n is determined, and the contact connection between layer w _ij and w _ik range is used to calculate the proportion of output h of hidden layer. H j = f ( ∑ i = 1 n w ij x i - a j ) , j = 1 , 2 , . . . , l(1)

Step 3: Printout. The output h, B predicted by evolutionary algorithm is 0. O k = ∑ k = 1 l H j w jk - b k , k = 1 , 2 , . . . , m(2)

Step 4: error calculation. According to the prediction of output 0 and expected output y, the prediction error e is calculated. e k = Y k - O k , k = 1 , 2 , . . . , m(3)

Step 5: updated scale. Based on the e-prediction error update algorithm w _ij and w _ik . Where is the learning rate.

Step 6: update scope. Update the range a, B according to the prediction error E.

Step 7: determine whether to end. If the standard is not met, return to step 2.

This neural network is a parallel algorithm based on probability theory, which has many advantages, one is classification ability and multi-dimensional processing ability [18]. PNN is a feed-forward bionic algorithm. Its theory is based on Bayes minimum risk criterion. The algorithm is developed from a radial basis and is very suitable for pattern recognition.

There are some problems in neural network modeling. The selection of feature vector must reflect the characteristics of the problem correctly. It is very important to contain enough information in the function. It is necessary to establish a learning system in the neural network and add the quantification of data processing, so as to realize the reasonable and quantitative understanding of social relations, modern rationality, knowledge system and values through the neural network learning and training. The next step is to predict the development trend according to the information results, and then put forward the corresponding countermeasures. The input eigenvector of bionic algorithm is social relationship, marketing strategy, knowledge system and value concept [19].

The collected 33 * 4 dimensional matrix data, before selecting multiple (such as 23) samples, is used as network training samples, and some (such as 8) samples are selected as verification samples. The input layer of neural network adopts three ratio method, and the output is the result of system simulation and recognition. The code of simulation type corresponds to Table 1.

Based on the bionic algorithm of the classification and prediction model of Ideological and political culture system in Colleges and universities from the perspective of modern media, a two-layer network algorithm - Classification layer (cluster) and competition layer is established by MATLAB. PNN network creates a direct function provided by net = newpnn MATLAB (P, t, spread). As mentioned above, input vector and threshold processing are completed after input to the function [20]. Figure 2 shows the fault tolerance diagram of algorithm sample training. After loading the data, select the type of print matrix, and then convert the required classes into vectors. In this paper, the algorithm aims to train the network. If the results are consistent with the use of built-in functions, the implementation results are as follows. In the depth research of intelligent bionic algorithm, there are two factors that limit the accuracy of the algorithm. One is the number of training data, the other is the selection of s value. According to the research results [21], the more samples there are, the more types there are, the more accurate the prediction results. Through simulation and literature research, it is shown that the convergence speed of BP neural network is stable and the additional fault tolerance of samples needs to be improved. The algorithm is robust to the above problems. The prediction results are shown in Table 2.

OPEN IN VIEWER

Fig. 2

Fault tolerance diagram of algorithm sample training.

As a teacher, he must understand the learning situation before class, and understand the basic situation of the students in the given class in detail. He can design some forms or questionnaires to investigate the intelligence of the students according to the eight intelligence situations of the students. For example, the following table can be used to let students evaluate their intelligence. The specific survey table is shown in Table 3, which shows the pre class preparation score of a marketing trainee.

Before the test, the teacher should explain the eight kinds of intelligence and fill in the form, so that the students can fill in more targeted. In addition, teachers can also use their spare time to communicate with students, learn together, play together, etc., so as to better understand students’ artificial intelligences, and prepare a notebook to make a special record of each student’s strengths and weaknesses in the eight intelligences, so as to teach prisoners in teaching. It may take a lot of work for teachers to do so, but once teachers understand the situation of students’ intelligence, it will be very helpful to use the theory of Artificial intelligences for teaching in the future. Before class, there are also other arrangements for students. Teachers can group students according to different subjects and students’ intelligence, and then assign corresponding tasks to each group, such as: students in the group with strong language and observation intelligence can arrange them to go to the local shopping mall to investigate the pricing situation of crystal merchants and make records; students in the group with good mathematics and spatial intelligence can assume a new shopping mall They set prices for Shangjing [22]; the group of students with music intelligence can let them consider the price of goods from the rhythm and hearing; the students with better self-knowledge and intelligence can let them feel the price of different local shopping malls as consumers; the students with better interpersonal intelligence can let them do group organization and cooperation; the students with better physical intelligence can do the same Learning can make you run more to do a kind of Market Research on the price of new commodities in the near future. This arrangement is based on their own strengths and intelligence. Students are willing to do it, and it is relatively easy to do it. No matter how well they do it, as long as they do it seriously, teachers will reward their labor achievements with corresponding spirit or small gifts to improve students’ enthusiasm for learning.

In the classroom, according to the preparations made by different intelligent groups before class, the teacher first asks the representatives of each intelligent group to make a report, reports the summary of the preparations made by each group before class, and makes a simple record on the blackboard, then organizes the students to analyze and evaluate the reports of each intelligent group. If there are different opinions, depending on the situation and time, several of them can be selected In the end, the teacher summarizes the basic knowledge of pricing strategy, the application of pricing strategy and the matters that should be paid attention to when pricing according to the students’ reports and the results of the debate. In the whole classroom teaching, the teacher only plays the role of leading and guiding, and should make full use of the characteristics of different intelligent students to guide, so that different intelligent students can participate in and give full play to, and timely encourage and appreciate their active participation, so that the whole classroom can participate in a wide range, live and orderly.

4.3 After-class consolidation of marketing teaching

Marketing is a very practical subject, and the knowledge learned in class is limited, so it is necessary to consolidate and further deepen the knowledge and practice after class. The teacher’s consolidation and assignment after class are also based on the intelligence of different students. According to the grouping situation of different intelligent students before class, we will continue to make in-depth discussion on the problems not covered in class after class and observe the pricing strategies of commodities in local shopping malls to understand the various pricing methods and applications of commodities. Finally, according to the research results of each intelligent group, a Market Research Report or summary report on commodity pricing is written by a group with better language and logical intelligence. The teacher makes a corresponding evaluation of this report, and makes a record of the usual performance of each student participating in this process, as a part of the examination results of this course. At the same time, this report will be posted in the academic report column of the class for students to study together, so that students have a sense of achievement from hard work.

4.4 Marketing test of artificial intelligence technology

Marketing test is no longer a single mode, but according to the intelligence of students. The determination of learning and achievement is not only a test paper, but also in various forms. I divide the students’ marketing scores into four aspects [23–26]: ① preparation materials before class; ② classroom performance; ③ homework after class; ④ midterm and final examination. In the examination, the principle of “my exam, I am the master” is adopted to give the students the right of independence, less rote questions, more objective choice questions, judgment questions and case questions that fasten the current affairs. In the examination questions, we can also choose more questions reflecting different intelligence according to the different intelligence of the students, so that they can choose the questions that are suitable for themselves to do. In this way, we can learn Students do not have the same questions, not only to avoid plagiarism, but also to really test their academic performance, so as to achieve the test also Shaanxi music. Using the theory of artificial intelligences in the teaching of marketing can reflect the students as the main body, fully mobilize the initiative of students’ learning, let every student participate in the classroom, let every student get the joy of success, and improve the efficiency of the classroom. But all of these are based on the teachers’ understanding of the artificial intelligences of learning cows and the handling of teaching materials. The more clearly the teachers understand the artificial intelligences of the students they teach, the more successful their application in teaching will be. This puts forward higher requirements for our teachers, which requires teachers to be able to deeply understand the multi intelligence situation of each student, change our traditional outlook on students and talents, teach students according to their aptitude, and evaluate our students more scientifically and better. I believe that as long as we pay R, there will be a good harvest tomorrow.