online agriculture products store case study coepd

Academia.edu no longer supports Internet Explorer.

To browse Academia.edu and the wider internet faster and more securely, please take a few seconds to  upgrade your browser .

Enter the email address you signed up with and we'll email you a reset link.

• We're Hiring!
• Help Center

A FEASIBILITY STUDY FOR ONLINE MARKETING OF AGRICULTURAL GREENHOUSE PRODUCTS W.R.T. PUNE DISTRICT

One third population of our country is depended on the agriculture business directly or indirectly and hence it is proven that we are an agriculture country. Having more than 30.02%1 contribution in GDP, agriculture plays the significant role in Indian economy. In last two decades everything is being converted into digital platforms globally. All industries became the digital and adopted paperless transactions. Following the same way in marketing, organizations started selling their product on e-commerce platforms. Promotion activities have started in social media like e-mail, websites, messages etc. Being understood the convenience, effectiveness and efficiency of digitalization Indian Government also launched “Digital India” scheme under which Government is promoting the use of technology in organizational function. Even though it seems somewhat inconvenient to use these types of technologies in unorganized sectors like agriculture, Government of India has announced in its “Union Budget 2016-2017 central Government stated that, the Unified Agricultural marketing e-platform will be launched for wholesale markets in India and 100% FDI to be allowed through FIPB route in marketing of food Products produced and manufactured in India. Considering these changes are going to affect the marketing and trading of agricultural products tremendously, this study is initiating the discussion regarding the possibilities of the digitalization in agricultural business.

Related Papers

IAEME Publication

Traditionally, agricultural information exchange has been dominated by business media such as newspapers, television and magazines. But now the power of the 21st century is literally in our hands. Digital marketing for Agricultural produces is the need of the hour for developing country like India. The study aims to understand the farmer's knowledge in Social media and its uses for production and marketing of their produces. Data were collected from primary sources, from Pooranakuppam village, Manaveli village, Thavalakuppam village and Nallavadu village under Pondicherry. The study found most farmers in the mentioned villages uses social media for agricultural purposes, from production to marketing and other related areas. They faced challenges in accessing them at the same time

Agriculture occupies a central position in the Indian economy. It contributes 29.4 percent of GDP, employing 64 percent of the country's workforce. Food price volatility has had a dramatic impact on the food security of poor households in developing countries. Equally serious has been the impact of long-term trends such as a slowdown of agricultural productivity growth, urbanization and an increasingly sophisticated supply chain. These changes have led farmers to focus on exports and other means of commercialization for better returns. For this a positive attitude of farmers towards farm management and export benefits is essential. With this back ground, the present study is taken up with the rigor.

Opinions by other consumers influences considerably on consumers' buying decisions as well as on post-purchase product perceptions. This is called word of mouth (WOM) and it is assumed to be more efficient than conventional marketing tools like advertising and personal selling. With the dawn of the Internet, major changes to the way consumers express their opinions about product or services have seen a sea change. Customers can now interact with other consumers via social media, e-mail, instant messaging, homepages, blogs, forums, online communities, chat rooms and review sites. Such interactions now commonly known as electronic WOM (eWOM) also influence consumer behavior. The Internet's accessibility, reach, and transparency have empowered marketers in influencing consumers. There is also need for understanding to what extent e-WOM is similar or different from offline WOM. Influence of e-WOM and traditional WOM cannot be undermined in today's highly competitive business environment. Market is ruled by organizations that can influence consumer behavior in its favour. Consumers are inclined to consider their friends and relatives are more credible and trustworthy than any form of sales promotional activity like advertising, public relations, and sales people. Thereby consumer-dominated communications tend Expanding The Horizon of Marketing: Contemplating The Synergy of Both Traditional Word of Mouth and E-Word of Mouth http://www.iaeme.com/IJM/index.asp 205 [email protected] to be more powerful and influential than information from other sources. This indicates that individuals are more inclined to embrace the information disseminated through WOM than commercial promotion. The advances of information technology and the emergence of online social networking sites have profoundly changed the way information are transmitted and have transcended the traditional limitations of WOM. The effect of e-WOM on consumers is similar to that of traditional WOM. Previous consumer's reviews on products and services have been an influential tool as an information source for consumers. Effectiveness in influencing customer behavior of both traditional WOM and eWOM cannot be undermined. However, there are some differences between traditional WOM and e-WOM. In traditional WOM communication, the information is exchanged in face-to-face conversation while in e-WOM consumers need only to interact with their computers to post or search consumer reviews. That is senders and receivers of information are separated by both space and time. Thus, it is a many-to-many communication in which the information exchanged is more voluminous in quantity compared to information obtained from traditional contacts in the offline world. As traditional WOM comes from friends and relatives while the source is anonymous in e-WOM differences of influence between the traditional and the online process may also depend on other factors. The relative influence of WOM and e-WOM may be dependent on the volume of information obtained from both sources. When consumers obtain more information from traditional WOM than from e-WOM, WOM influence is likely to be higher than e-WOM influence. Equally, the reverse effect is expected for consumers who obtain more information from e-WOM than from WOM. The differences between online and traditional process could affect the degree of influence on consumer behavior. Understanding the strength of both traditional WOM and eWOM can broaden the horizon of marketing products and services of an organization.

Agriculture has been the mainstay of India's Economy for centuries. Even now, when the world is proliferated with Technology and Science in everyday lives of people, Agriculture has stood out and displayed stoic resilience. This is because Food touches the very basic need of humans. Most of the farmers around the world are using the farming or agricultural practices based on the knowledge and experience gained over generations and generations. With Food Security increasingly threatened with millions of humans adding up every day, Agriculture in modern times has to gradually transform towards leveraging increased automation and Technology for faster yields with less resources. The paradigm shift has the potential to transform the face of our villages and cities. Amethi Region in Uttar Pradesh, Northern India, a political Super brand of Indian politics, being known for affiliation to Nehru-Gandhi dynasty and previously ruled by Royal Maharajahs of Amethi since 10 th century is well poised and positioned to reap the benefits of modern Agricultural practices and be an Agricultural hub. India's stunning economic growth is not only gobbling up spaces like a giant sponge but gradually eating up the beautiful scenic lands of our villages also. Pristine vistas with rows and rows of green fields are becoming History by the day. This paper describes the various measures and methods by which Amethi can be an Agriculture leader by blending core Agriculture with modern Technology with priority to leave maximum open spaces " intact and green " for enjoyment by next generation people who are getting restricted in claustrophobic cities and towns.

Digital marketing includes those marketing efforts that send a message from a source (company) to a receiver (customer) through digital platforms and the internet. Technology is opening up a world of possibilities previously unavailable to both marketers and consumers. Also, options and reach are hugely enhanced by international e-commerce. Electronic commerce, commonly written as e-commerce, is the trading or facilitation of trading in products or services using computer networks, such as the Internet. Electronic commerce draws on technologies such as mobile commerce, electronic funds transfer, supply chain management, Internet marketing, online transaction processing, electronic data interchange (EDI), inventory management systems, and automated data collection systems. Modern electronic commerce typically uses the World Wide Web for at least one part of the transaction's life cycle, although it may also use other technologies such as e-mail. This study is to find the promotional activities and factors influence customer to E-Commerce sites. The sample size is of 120 respondents and simple random sampling method was used to data collection by using a self-administered questionnaire that was distributed through Google forms to respondents and sending mails to consumers through their past experience of purchasing in E-Commerce sites.

narendhar bollawatri

Consumer behaviour helps to identify the behaviour of the consumer upon that product and also for the market. The main objective of the study is to find the overall consumer behaviour. Data from the customers visiting big bazaar were collected as population study. The primary data was obtained through questionnaire and secondary data from the company records and through internet. The purpose of the survey process is to analyze the consumer behaviour of big bazaar. In order to achieve a competitive position in the market. The research work will propose a true idea that identifies key factors which are of great importance to customers. From the analysis it was found that the big bazaar satisfying consumer needs and wants.

Radosław Kotecki , Max Naderer , Sini Kangas

This two-volume collection brings together contributions from cultural and military history to offer an examination of religious rites employed in connection with warfare as well as their transformative and power- and identity-building potential across political communities of medieval Northern, Central, and Eastern Europe. Covering the period ca. 900 and 1500, the work takes theoretical, textual and practical approaches to the research on religious warfare, and investigates the connections between, and significance and function of crucial war rituals such as pre-, intra- and postbellum rites, as well as various activities surrounding the military life of individuals, polities, and corporates.

ACM SIGGRAPH Computer Graphics

Anthony Maciejewski

Modeling techniques for animating legged figures are described which are used in the PODA animation system. PODA utilizes pseudoinverse control in order to solve the problems associated with manipulating kinematically redundant limbs. PODA builds on this capability to synthesize a kinematic model of legged locomotion which allows animators to control the complex relationships between the motion of the body of a figure and the coordination of its legs. Finally, PODA provides for the integration of a simple model of legged locomotion dynamics which insures that the accelerations of a figure's body are synchronized with the timing of the forces applied by its legs.

RELATED PAPERS

Matheus Pacheco

Junior Vega

François Dubuisson

Jurnal Penelitian Kelapa Sawit

Zulfi Prima Sani Nasution

Revista Electrónica de Psicología Iztacala

Angel Pacheco

International Migration Review

Barbara Oomen

Plant Disease

Hafiz Haseeb

Gyula Simon

Geological Society of America Abstracts with Programs

Jeff Tepper

41st AIAA Fluid Dynamics Conference and Exhibit

Pino Martin

Mae Ann Ociman

Surface and Coatings Technology

Voitech Stankevic

Binod Ghimire

多伦多大学毕业证书 购买加拿大学历Uof文凭学位证书

Critical Care and Resuscitation

Dr Timothy Chimunda

DOAJ (DOAJ: Directory of Open Access Journals)

Wolfgang Rabbel

Turkish Thoracic Journal

Current Opinion in Environmental Sustainability

Kristal Jones

Physical Review D

Edward Shuryak

RELATED TOPICS

•   We're Hiring!
•   Help Center
• Find new research papers in:
• Health Sciences
• Earth Sciences
• Cognitive Science
• Mathematics
• Computer Science
• Academia ©2024

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Comput Intell Neurosci
• v.2022; 2022

E-Commerce Marketing Optimization of Agricultural Products Based on Deep Learning and Data Mining

1 College of Economics and Management, Northeast Agricultural University, Modern Agricultural Development Research Center, Harbin 150030, China

Zhuohang Zheng

2 School of Management, Guangdong University of Education, Guangzhou 510303, China

Associated Data

The data used to support the findings of this study are available from the corresponding author upon request.

China Internet plus agriculture was first put forward in 2015 by the Chinese government's work report, laying the foundation for the development of Internet plus agriculture and promoting the rapid growth of e-commerce marketing of agricultural products. The combination of agricultural product marketing and e-commerce effectively reduces the intermediate links of agricultural product sales. Many e-commerce professional villages have sprung up in some rural areas across the country, and the number of rural e-commerce stores has continued to grow. At this stage, rural e-commerce has become a new way of agricultural trade, and rural e-commerce has formed a unique rural e-store. At present, the e-commerce market share of agricultural products in rural stores is very large, and its advantages are favored by the government, scientific research institutions, and agricultural products processing enterprises. However, with the gradual development of rural e-commerce, it has also encountered many difficulties. Based on this point, this study applies deep learning and data mining to optimize e-commerce marketing. First, with the growth of the online scale of agricultural product transaction data, the creation of traditional shallow model cannot meet the needs of online data processing. Therefore, this study decides to use the deep learning theory for optimization. It has excellent performance in the technical fields of big data processing and image and voice processing and has strong construction ability, which can effectively represent the characteristics of the model. Combined with the characteristics of e-commerce agricultural products processing and consumer practice, this study designs and develops a new customer value evaluation model based on data mining and e-commerce agricultural products value characteristics in the field of e-commerce. By combining deep learning and data mining technology, this study applies it to the field of e-commerce, so as to promote the transformation of marketing optimization.

1. Introduction

The Internet economy has become a major theme of economic growth in the 21st century. The Internet and related technologies are leading the rapid development of economy and trade, surpassing traditional business activities and giving full play to their greatest advantages in daily life [ 1 ]. The emergence of e-commerce has opened a new era, which “destroys” the traditional mode of operation and has a great impact on the production, operation, and marketing of enterprises [ 2 ]. No enterprise or industry can ignore the changes brought by e-commerce [ 3 ]. Only by conforming to the development of e-commerce business and building a new marketing model we can occupy a strong position in the economic wave of the 21st century and obtain the main advantages of market competition [ 4 ]. The development of e-commerce has shortened the distance between enterprises and consumers, greatly reduced the marketing cost of enterprises, and promoted the income growth of enterprises [ 5 ]. At the same time, compared with traditional commerce, the convenience of e-commerce aims to stimulate consumer consumption, so as to further increase enterprise sales [ 6 ]. At the same time, e-commerce goes beyond the space limit and enables enterprises to reach more consumers and attract more consumer groups [ 7 ]. Under such extensive influence, agricultural products must also meet the needs of the development of the Internet and integrate more Internet content into the original marketing model [ 8 ]. Based on this point, this study analyzes and optimizes the e-commerce of agricultural products by combining deep learning and data mining technology [ 9 ]. Due to the fast growth and variability of transaction data and the weak corresponding relationship between sales factors and sales level, the data preprocessing technology based on fuzzy membership is used to optimize it, so as to further optimize the deep learning algorithm [ 10 ]. By adding the decentralized restricted self-coding method, this study proposes a low membership deep learning sales prediction model—super crown model (SICM), which effectively solves the factors affecting product sales and balance level [ 11 ]. Then, aiming at the problems of unclear target users and low efficiency, this study carries out data mining on the agricultural product customer data provided by enterprises. To identify potential new business customer groups, the personalized needs of customers are taken into account and the consumption probability of customers is predicted [ 12 ]. The process of agricultural products should be sorted out to obtain the best recommendation effect, realize reasonable marketing, effectively improve the performance of agricultural products, or carry out online sales. Through such treatment, we hope to help promote the development of e-commerce for agricultural products [ 13 ].

2. Related Work

This study proposes an online marketing model of agricultural products based on deep learning—crown model (ICM), which uses two-layer automatic coding network to capture samples [ 14 ]. The labeled sample set is used to extract the classification features of unlabeled training samples; BP is used to adjust the parameters of the whole network, get the best parameters of the loss function, and realize the dynamic classification and prediction of online sales of agricultural products [ 15 ]. A depth weighted K-means clustering algorithm based on an enhanced depth clustering neural network is proposed to realize telecom customer segmentation [ 16 ]. This algorithm solves the problem that the traditional K-means algorithm is sensitive to starting the clustering center, further improves the grouping performance, and can be easily extended to large-scale data [ 17 ]. Comparative experiments show that the clustering algorithm is superior to other mainstream clustering algorithms in customer segmentation and can provide strong information support for resource planning operators, which fully shows the practical significance of the system [ 18 ]. An improved depth neural network (DNN) model SDNN is proposed in the literature. The parallel calculation based on GPU shows that by establishing the model and implementing the algorithm without affecting the prediction effect, the training time of SDNN prediction model is reduced by about 73.28%, and the efficiency of DNN calculation is further improved [ 19 ]. Literature has confirmed that a relu activation function is more suitable for deeper network models than sigmoid activation function, while DNN and SDNN based on the relu activation function are more suitable for modeling complex problems. The literature makes an in-depth study on the development of e-commerce operation mode at home and abroad and expounds on the role and importance of big data in the future development of e-commerce [ 20 ]. The future development framework of e-commerce is established based on big data, so as to predict the mode and marketing trend of agricultural products in the future.

3. Theoretical Basis

3.1. deep learning.

The research on online sales prediction of agricultural products involves the application of information science pattern recognition in the field of e-commerce. As for the research on commodity sales prediction methods, the research at home and abroad is mainly based on the analysis and comprehensive prediction of commodity sales prediction related methods, such as gray theory, artificial neural network, and time series. Such methods have low prediction accuracy when dealing with a large amount of data, and such models usually lack long-term effectiveness and scalability.

An automatic encoder (AE) was proposed by Rumelhart in 1986. It is a deep learning method with a fast learning ability. Its main purpose is to learn the compressed distribution feature expression of a given data set. Its essence is to use the characteristics of the hierarchical structure of artificial neural network (ANN), use the hierarchical unsupervised self-monitoring pretraining and parameters for system optimization, and extract the hierarchical features from large-scale complex input data, so as to obtain a distributed original data distribution. Feature representation deep learning neural network structure shows the high efficiency of processing high-dimensional big data. A simple AE consists of three components: encoder, decoder, and hidden layer. The biggest difference between it and the traditional neural network, which directly uses the predictive learning results, is that the automatic coding network only pays attention to the hidden parameters of layer weight and does not carry out classification operation.

Assuming that an unlabeled training data sample V is input, the encoder maps V to the hidden layer H . V is converted to H activated by formula ( 1 ), and V and H satisfy the following formula:

Among them, s f is a nonlinear activation function, which usually uses the logic function sigmoid function. It is a common S-shaped function in biology, also known as an S-shaped growth curve. Its definition formula is as follows:

θ ={ W , b } is the parameter set. H reconstructs the input value V into y through reverse coding, and H and y satisfy the following formula:

where θ 1 = { W 1 , b 1 } is the parameter set and s g is the decoder activation function. The typical formula ( 2 ) is selected. The relationship between parameters W 1 and W T in equations ( 1 ) and ( 3 ) is W 1 = W T The reconstruction error J AE is reduced by continuously optimizing and updating θ and θ 1 . J AE is defined as follows:

where V is the input data set and L is the reconstruction error function, and the expression of L is as follows:

Automatic encoder learning uses a simple identity function to realize the complete reconstruction of the original data, while for complex data structure, it needs to learn in complex nonlinear function, because the self-encoder cannot meet the needs of complex data structure and complex learning nonlinear function. Therefore, most advanced learning rarely uses coding, which effectively improves the stability of the model. Sparse self-coding is to add sparse constraints according to the structural characteristics of artificial neural network on the basis of automatic encoder, so as to use the least number of hidden layer units to express the input characteristics, that is, hidden layer nodes. It is suppressed most of the time and active only a few times, so when the number of hidden layer units is large, the structural part of the input vector is still visible. The main purpose of unsupervised vector learning is to use the sparse vector set of ∅ i encoder for automatic learning:

Among them, i  = 1, 2,…, k ( k is the number of input nodes) and α i is a linear correlation coefficient.

The average activation value of hidden layer unit j is as follows:

Among them, i  = 1, 2,…, m ( m is the number of input nodes, m  =  k ) and a j (2) ( x i ) represents the activation amount of hidden layer unit when the input is x .

The penalty factor is expressed in the form of relative entropy KL:

The general cost function of sparse automatic coding network is as follows:

Among them, J ( w , b ) is the global loss function, λ is used to control the importance of the second part of the formula and prevent overfitting, and μ is the penalty coefficient. The higher the penalty is, the sparser the result is:

Among them, J ( W , b ; x i , y i ) is the unit loss function:

The output error is obtained according to formula ( 11 ):

δ i (3) represents the error of the output layer and α i (3) is the activation function:

Thus, the error of hidden layer element is as follows:

3.2. Multilabel Classification Method of Data Mining

As one of the most commonly used data mining scenarios, the recommendation system can recommend products of interest according to customers' personal needs and consumption habits, improve product order rate, and realize reasonable marketing. Data mining often uses classification algorithms to build recommendation systems. The classification task obtains the classification model based on the known samples in order to classify the unknown samples. Common classification problems usually include two classification and multiclassification. Both of them are single label classification problems; that is, each sample in the training set has a label. If the label set L  = 2, it is a two classification problem; if L  > 2, it is a multiclassification problem.

For a given sample X and a corresponding set of tags Y , N ( X ) is used to represent the sample K closest to the sample x. The main steps of implementing the ml-knn algorithm are as follows:

• (1) Calculate the label set as shown in the following equation: Y = y 1 , y 2 , … , y j . (15)
• A priori probability of each label is calculated as follows: P H L = s + ∑ i = 1 m y j ∈ Y i s × 2 + m , (16)
• where s is the smoothing term, usually 1.
• (2) Calculate the number of nearest neighbors N ( t ) marked L to K corresponding to each sample in sample set X : C L = ∑ a ∈ N t y a L . (17)
• (3) Calculate the conditional posterior probability P ( E C ( L ) L |H b L ) of each sample in N ( t ).
• (4) Calculate the k-nearest neighbor N ( x ) of the unknown sample.
• (5) Calculate the maximum a posteriori probability of unknown samples: y = arg max b ∈ 0,1 p H b L | E C L L . (18)

The label marked with 1 is the prediction result of the classification.

Like KNN, the main problem of ml-knn algorithm is the selection of K value. This algorithm is not suitable for big data classification.

Xgboost model is a combination of several decision trees. Adding a tree each time you split a feature is equivalent to adding a new function. Each time a tree is added, it is expected to improve the classification effect, that is, to ensure that the value of the loss function is reduced after adding a new function. However, in the process of adding trees, the number of leaf nodes may be too large and the model may be overfitted. Therefore, a regularization penalty should be added to the loss function to limit the number of leaf nodes. Finally, the objective function of the model consists of loss function and regularization:

F is a set of K decision trees, and each f is one of them. It is defined as ( x ) =  wp ( x ). The parameters are leaf node w and tree structure ρ , where ρ ( x ) is the index of leaf node where x is located.

3.3. Design of Data Mining Steps

Data mining is not a simple process of mining results by establishing a data-driven model but a process of repeated iteration and continuous improvement. Generally speaking, the data mining process is as follows:

•   Understanding the Project . First, understand the project requirements from the business perspective, then analyze the project from the technical perspective of data mining, and put forward a preliminary scheme to meet the business requirements by combining theory with practice.
•   Understanding Data . After setting the project objectives, it is necessary to sort out the relevant data, then describe and analyze the data, and make a qualitative evaluation of the data quality.
•   Data Preprocessing . The original data are processed in the appropriate final data set for modeling processing.
•   Modeling . Select appropriate modeling methods and adjust model parameters according to the quality and scope of the data itself. Once the algorithm is selected, the training data set will create a model. In the training process, the relevant parameters of the algorithm are continuously adjusted to obtain the best model. After modeling, use the test data set to test or predict the model.
•   Evaluation . Conduct a comprehensive and hierarchical evaluation of the model. The evaluation method should consider the differences between algorithms and application fields and select appropriate evaluation parameters. If the verification results meet the requirements, that is, the model really meets the business requirements, then go to the next mining step. Change the direction of data analysis and make plans.
•   Model Implementation . Modeling is not the ultimate goal of mining. After getting an ideal model that meets the business requirements, we also need to understand the business of the model and provide interpretation and guidance to support the actual business application.

4. Current Situation of E-Commerce Marketing of Agricultural Products

4.1. investigation methods, 4.1.1. determination of sample size.

Determine the Sample Size . In the investigation of an e-commerce rural store, the operators of e-commerce rural stores were the main respondents. As there are many types of e-commerce products mainly engaged in agricultural products in rural e-commerce shops in e-commerce villages, in order to ensure the correctness and objectivity of the survey, the author uses the method of multistep sampling to determine the residents of e-commerce village who finally conducted the survey. In this process, the deff design effect is used to realize simple random sampling. The confidence level is set to 0.95, and the sample size is set to 0.5 and 0.3, which are expressed by P and Q, respectively. The error sampling limit is controlled at about 5%, the deff value is 1.3, and the final sample size is 293. In addition, 10 e-commerce villages in a county were selected as the survey items. Due to the influence of various factors, the recovery rate of this survey is not 100%, but 90%, so the actual number of questionnaires is 297/90% = 330. Therefore, in this survey of 10 e-commerce villages in a county, the total scale of 330 e-commerce rural stores is taken as the sample.

4.1.2. Sample Allocation

As the survey is a nonrepeated sampling survey, the sampling units in the first stage are all e-commerce villages in a county (6 towns, 3 ordinary townships, and 1 ethnic township), with a total of 4 streets, 183 administrative villages, 4 neighborhood committees, and 15 communities. Ten e-commerce villages are selected as the survey elements, that is, the sample units in the second stage. In the follow-up distribution of this survey, an appropriate proportion of households, namely, rural online stores, will be selected according to the sample units in the second stage. In the sample, considering the subtle differences in the distribution of rural e-commerce stores, the probability sampling method is selected, and the actual random sampling is carried out according to the proportion of household administrative summary in the total number of rural households. Finally, the total number of households actually received the query, that is, the number of e-commerce rural stores, is taken as the test unit of the third stage.

In addition, this study also conducted separate interviews with e-commerce village partners, Ali operation center, and county government staff, added countermeasures to the questionnaire, and conducted investigation and analysis, making the results more reasonable and scientific.

4.1.3. Sample Selection Method

On the one hand, the second stage sampling of rural local electricity store operators is divided from regions to rural electricity stores. The survey is divided into two parts: t county and administrative village. Each e-commerce rural store is taken as the unit for the general survey. On the other hand, a multistage sampling survey is conducted on local e-commerce rural partners, from regional divisions to administrative villages, then to e-commerce rural stores, and finally, to e-commerce rural partners. In the process of regional division, according to the division of administrative regions, this sampling method makes the distribution of samples more uniform. In the sampling process of administrative villages, a typical sampling survey of 10 administrative villages in t county is adopted to determine a reasonable sampling interval.

4.1.4. Questionnaire Design Method and Process

A total of 20 questions were raised in this questionnaire, including five questions: the operation and management system of rural electric stores, the quality and brand of rural e-commerce products, the logistics and transportation of rural agricultural products, the technical talents of rural electric stores, and the effectiveness of e-marketing of rural electric stores. Among them, there is no special limited investigation on the types of agricultural products in e-commerce rural stores. A total of 330 questionnaires were distributed and 330 were recovered. Under the supervision and guidance of the investigators, there were 330 valid questionnaires, and the recovery rate and effectiveness rate of the questionnaire were 100%.

4.2. Data Processing

According to the purpose and content of the study, SPSS version 22.0 analyzes the relevant data and summarizes the research results. The data of the same group were compared with the paired t -test samples, and the data of different groups were compared with the sample t -test. The significance level was set as P  = 0.05.

4.3. Research Results

As shown in Figure 1 , more than 45% of online store farmers in a county choose Taobao, 9% choose pinduoduo, 11% choose Xianyu, 15% choose Jingdong, and other farmers choose wechat.

Proportion of sales platforms.

More than half of the farmers have opened online stores for 1–3 years and have just entered the online sales team recently. Only a relatively small proportion of online shoppers have been online for more than three years. Most of the employees of the online store are relatives, so the customer service, packaging, and warehousing work are completed internally. Because most of them are self-employed, they rarely recruit foreign personnel. Influenced by the education level and education level, the reputation level of online stores is different. Only 16% of the respondents were crown-level online stores, and only 50% were diamond-level online stores.

As shown in Figure 2 , in the survey of online sales of agricultural products in rural e-commerce, 55.7% of rural e-commerce adopts the wholesale and retail trade of agricultural products, 23.5% of rural e-commerce only adopts the wholesale trade of agricultural products, and 16.7% of rural e-commerce only adopts the retail trade, as listed in Table 1 . The results show that in order to protect their own economic interests, rural e-commerce mostly adopts the way of wholesale and retail, while the operation and management of agricultural products in e-commerce have not been fully developed.

Basic information of online store.

Forms of online sales of agricultural products by rural e-commerce.

In the rural e-commerce agricultural product marketing management survey, up to 59.7% of rural e-commerce store operators said that they could only use a small amount of information technology to achieve governance, and 23.5% of rural e-commerce owners said they could make full use of big data and other advanced information technologies for management. About 16.8% of rural e-commerce operators said that they could not use advanced information technology to achieve information governance, as listed in Table 2 , which fully shows that there is a certain lag in the management system of the current e-commerce business model.

Technical management of e-commerce marketing of rural agricultural products.

Market attention is one of the main factors affecting business growth. The number of search clicks is proportional to the number of followers, and the higher the number of searches, the more followers. Taking company a as an example, comparing company a with the other three companies in the same industry, we can see the current concern of the nut market in a certain city. Market attention is analyzed from four aspects: the number of search clicks, the number of followers, the number of transactions, and the number of favorites, as shown in Figure 3 :

Market scale.

Company a has no obvious advantages in the number of search clicks, followers, and collectors, but it is higher than the other three companies in the number of transactions. Most of the reasons are related to customer service behavior and website interface. Good customer service attitude, short web response time, and website interface have attracted consumers' attention to ensure that customers will not have “running orders.” If consumers are familiar with a brand, they will continue to pay attention to the brand. If there is a demand, the brand will be the first choice, and consumers will conditionally look for the brand and buy it.

As shown in Figure 4 , the number of search clicks is directly proportional to the number of followers, and the total click-through rate of each promotion and the conversion rate of concerned brands will also increase. Therefore, companies with a large number of collections may not conduct multiple transactions for customers. The survey shows that some consumers will continue to browse stores after collecting goods. If they find products with lower prices, more discounts, and better evaluation, they will give up collecting and focus on another product.

Brand conversion rate.

Second, after investigation, the problems of agricultural products existing in rural e-commerce mainly include the high price of agricultural products, poor quality of agricultural products, low brand awareness of agricultural products, few types of agricultural products, and insufficient after-sales service level. Among them, the survey proportion of low quality of agricultural products and low brand awareness of agricultural products ranks first and second, as listed in Table 3 . It can be seen that in the process of e-commerce of agricultural products, the quality and brand of e-agricultural products are now the most prominent problems.

Problems of rural e-commerce agricultural products.

Through the survey, it is found that the 330 rural e-commerce stores surveyed rely on the e-commerce marketing income of agricultural products every year, as listed in Table 4 .

Annual income of rural e-commerce relying on agricultural product marketing.

4.4. Problems in E-Commerce Marketing of Agricultural Products

The survey found that rural online stores often have some problems that are unfavorable to the e-commerce marketing of agricultural products, such as poor quality of agricultural products and poor brand construction effect. For example, a survey found that among the operators of Taobao rural stores, 48.6% of the agricultural products of Taobao rural stores were purchased from surrounding farmers, 31% of the agricultural products were directly supplied by suppliers, and only 20.4% of the agricultural products of rural online stores were produced by themselves. Purchasing from farmers often lacks scientific identification and quality control, which is easy to lead to quality problems of agricultural products, while moderate brand promotion activities are only carried out occasionally. Overall, the problems of agricultural products in rural online stores mainly include the high price of agricultural products, poor quality of agricultural products, low brand awareness of agricultural products, few varieties of agricultural products, and low after-sales service level of agricultural products. Therefore, the e-commerce transaction quality and brand construction of agricultural products in rural online stores need to be improved.

5. Related Systems and E-Commerce Marketing Optimization Methods of Agricultural Products

5.1. system design.

During the operation of the basic layer, it includes information mode, service mode, and products, as shown in Figure 5 . Products are taken as the core and information mode and service mode are taken as auxiliary support means. The foundation layer is the most important and basic part of the whole platform, which will play a decisive role in the future development of e-commerce of fresh agricultural products to a certain extent.

Basic layer operation strategy.

The whole field of modern e-commerce is a big data era. The information process of various industries involved in e-commerce usually adopts the “cloud information” mode, including the efficient collection of “cloud information,” the organization of information within the platform, the release of efficient information, information feedback market, and information exchange based on this information.

As the core part of the e-commerce operation mode of fresh agricultural products, the core layer mainly includes logistics distribution mode, marketing mode, and profit mode, with sales mode as the core, logistics distribution mode as the guarantee, and profit mode as the goal, as shown in Figure 6 .

Core layer operation strategy.

5.2. Algorithm Design of Marketing Model

The specific implementation steps of SICM algorithm are as follows:

• Preprocess the collected online transaction data, which is divided into training data set and test data set.
• Write the test data set, that is, mark the specific level of sales indicators, which is divided into two parts: one part is used to drive the SICM model and the other part is used to test the performance of the SICM model.
• Because the online transaction data has the characteristics of large capacity, variability, and high speed, the model determines the number of hidden layers and neurons in each layer in the small-scale automatic coding network through repeated experiments, so as to achieve higher sales and accurate classification.
• The training data set is used as the input vector of sparse self-encoder network for unsupervised pretraining.
• In the softmax classifier, labeled and unlabeled data are used as input vectors to drive the classifier.
• BP is used to improve the parameters of deep learning network to make the parameters reach the global highest level.

When λ  ≠ 0, the weight value attenuation function is activated, and the weight update rule is as follows:

where η is the learning rate. Generally speaking, taking η as a small value will lead to gradual error convergence, and η is determined by the specific scale of the network. In this study, η  = 0.05 is selected.

The weight update formula is as follows:

where ∆( l ) and ∆ b ( l ) are the initial values, and the value is 0. The weight is obtained according to the following formulas:

5.3. Analysis of Experimental Results

Logistic regression is one of the most common models for CTR prediction problems, which is widely used by major companies. Decision tree is the basic model of GBDT. In this experiment, the decision tree model is added. In terms of CTR prediction, the GBDT model is the top noun in major competition platforms. As a representative of integrated learning, GBDT also has a research value. The structure of the neural network model is 2022-1000-2, and the initial weight of the model is random [0, 1]. Table 5 lists the parameter settings for each shallow learning model.

Parameters of shallow learning model.

The comprehensive comparison results of various methods are listed in Table 6 :

Comparison results of various methods.

Compared with the benchmark classifier of shallow learning, the improvement effect percentage of deep learning is shown in Figure 7 .

Percentage of relative improvement effect.

5.4. E-Commerce Marketing Optimization Method of Agricultural Products Based on Relevant Systems

Big data are used to improve the user experience. Big data can play a more humanized role in precision marketing, improve the user experience from obtaining product information to completing the purchase, and make users feel safer and more confident in the whole consumption process. Using big data to let consumers identify brands, the thinking stage is the cornerstone of cultivating brand loyalty. In real life, when buying products or services, most people tend to choose brands they are familiar with, rather than brands they have never heard of. Therefore, brand building is very important. Integrating advertising, business news, etc., and using big data can effectively improve brand awareness, such as sending messages to web pages, consumer shopping carts, and other places. At this stage, the relationship between enterprises and customers is relatively weak. If the products, services, or marketing activities of competitive products are more attractive, such consumers are more likely to use competitive products. Using big data to analyze the consumer identification stage, on the premise of identifying brand customers, customers decide whether to buy or not. At the first purchase, it indicates that the customer is familiar with the products of the brand. However, in the process of purchase, customers constantly waver and question their choice. If they can buy products at the most satisfactory price, after the first purchase, customers will see the value of products or services and master the data to be evaluated in their mind. Using big data to evaluate potential customers can take reasonable measures to win the favor of consumers when they are dissatisfied and let them buy and consume again and again. If the customer is satisfied with the product or service they bought for the first time, they will have a preference for the product or service and buy it again.

6. Conclusion

The rapid development of e-commerce has affected various social industries, from traditional electronic products and clothing to daily necessities and food. At present, more and more industries are affected by the Internet, and the depth of the impact is becoming greater and greater. At the same time, the state is also gradually increasing its support for e-commerce of agricultural products. Fresh green products and delicious food from all over the country can be delivered to the door through the e-commerce platform. The convenience of e-commerce of agricultural products can stimulate consumers to consume and further increase the sales of rural online stores. This study takes the data of rural online stores in a county as the research object, adopts the methods of literature retrieval, case analysis, problem investigation, and data analysis, and puts forward an optimization scheme of e-commerce agricultural products marketing based on relevant theories, in combination with in-depth learning and data mining technology, thus it will be helpful to e-commerce agricultural product marketers.

Acknowledgments

This paper was funded by the key project of Heilongjiang Province Philosophy and Social Science Research Planning “Research on the Effective Path of Heilongjiang Province's Brand and Strong Agricultural Construction under the Rural Revitalization Strategy” (no. 21JYA005).

Data Availability

Conflicts of interest.

The authors declare that they have no conflicts of interest.

• Preferences

PESTLE Analysis - COEPD - PowerPoint PPT Presentation

PESTLE Analysis - COEPD

Coepd - center of excellence for professional development is a primarily abusiness analyst and pmp training institute in the it industry of india head quartered at hyderabad. coepd is expert in pmp training and business analyst training in hyderabad, chennai,pune, mumbai&vizag. we offer pmp and business analyst training with affordable prices that fit your needs. coepd conducts 4-day workshops throughout the year for all participants in various locations i.e. hyderabad, pune. the workshops are also conducted on saturdays and sundays for the convenience of working professionals. for more details please contact us: visit at center of excellence for professional development 3rd floor, sahithi arcade, s r nagar, hyderabad 500 038, india. ph# +91 9000155700, [email protected] – powerpoint ppt presentation.

• What is PESTLE Analysis?
• There are many companies in the world who conducts PESTLE analysis on their brands in order to ascertain strategies for the future or to understand the market before launching their products.
• It is a fundamental tool used for market planning and strategizing that must be carried out to comprehend market trends and the systematic risks involved in the market.
• It is a macro environment which is used to understand the impact of the external factors on the organization.
• PESTLE stands for
• Technological
• Environmental
•  Political stability
•  Tax policy
•  Employment labor law
•  Environmental regulations
•  Trade restrictions
•   Tariffs etc.
• It affects the decision making process of the organization.
• A global player must examine the economic issues in the world and find opportunities with such conditions for business growth.
•  Economic growth
•   Interest rates
•  Inflation rate
• It represents the culture of the society that an organization operates within.
• It includes demographics, age distribution, population growth rates etc.
•    Age distribution
•   Population growth rate
•    Emphasis on safety
•    Career attitudes
• It is related to new inventions and development. 
• Affect the cost and quality of the outputs.
•    Automation
•   Technology incentives
•    Rate of technological change
•    RD activity
• It consists of both external and internal sides. There are certain laws that affect both sides of the business environment.
•   Consumer law
•   Antitrust law
•   Employment law
•   Discrimination law
•   Health and safety law
• It refers to ecological and environmental aspects such as climate, weather and climate change.
• Mostly it is useful for the certain industries like tourism, farming, agriculture etc.

PowerShow.com is a leading presentation sharing website. It has millions of presentations already uploaded and available with 1,000s more being uploaded by its users every day. Whatever your area of interest, here you’ll be able to find and view presentations you’ll love and possibly download. And, best of all, it is completely free and easy to use.

You might even have a presentation you’d like to share with others. If so, just upload it to PowerShow.com. We’ll convert it to an HTML5 slideshow that includes all the media types you’ve already added: audio, video, music, pictures, animations and transition effects. Then you can share it with your target audience as well as PowerShow.com’s millions of monthly visitors. And, again, it’s all free.

About the Developers

PowerShow.com is brought to you by  CrystalGraphics , the award-winning developer and market-leading publisher of rich-media enhancement products for presentations. Our product offerings include millions of PowerPoint templates, diagrams, animated 3D characters and more.

• Participant Payment

• ClassRoom Training
• Online Training
• Corporate Training
• Knowledge Partner
• Testimonials
• On Job Support
• Visakhapatnam
• Web Designing
• Data Science
• Digital Marketing
• BA Placements
• COEPD Clients
• Placement Reviews
• Success Stories
• Certifications
• BA Certifications
• CPOA Certification

IVY Global Academy

• Certified IT Business Analyst (CITBA)
• Certified IT Business Analyst Practitioner (CITBAP)
• Certified IT Product Owner (CITPO)
• Certified IT Scrum Developer (CITSD)
• Certified IT Scrum Master (CITSM)
• Job Openings
• Online Exams
• Our Endeavours 1954 Batches | 12160 Participants | 91.5% Placements | 6 Corporates | 4 B-Schools |
• Post Your Requirement |
• Verify Certification |
• Verify Awards |

 Weekend premium : Batch starts on 08 June 2024, For a free demo please contact +91 9742306227 / +91 89779 43515::   Hyderabad: New Batch Starts on 08 June 2024. Call Help Desk +91 89779 43515 / +91 8374676888.::   Hyderabad: Free Demo by Industry Expert. To know more details, Please Call Help Desk +91 89779 43515 ::   Chennai: Weekend premium batch starts on 08 June 2024 , Call Help Desk +91 88850 24387::   Bangalore: Weekend premium batch starts on 08 June 2024, Call Help Desk +91 8374465444 ::   Pune: Weekend premium batch starts on 08 June 2024, Call Help Desk +91 77220 37493.::   Bangalore: Free Online Demo by Industry Expert .Call Help Desk +91 9154829630::   Mumbai: Weekend premium batch starts on 08 June 2024, Call Help Desk +91 7414917172.::   Delhi: Weekend premium batch starts on 08 June 2024, Contact 9963555711::   Please drop an email to [email protected] for any Issues or Queries regarding COEPD Service Extensions and Certificates.::  

BA Training

BA Online Training

Welcome to COEPD! Business analyst training online/classroom course

is a primarily a community of Business Analysts. Objective of COEPD is to minimize project failures by contributing in the areas of Business Analysis. All BAs who are committed towards this cause, gathered and formed this COEPD Community. Through COEPD, we are striving to bring awareness of Business Analyst role and also the benefits of having a BA in project. As a part of this, we are imparting Business Analysis knowledge to all enthusiastic professionals, who are keen in getting into this BA role.

COEPD conducts 4-day workshops throughout the year for all participants in various locations i.e. Hyderabad , Visakhapatnam , Chennai , Pune , Mumbai , Bangalore & Delhi-NCR The workshops are also conducted on Saturdays and Sundays for the convenience of working professionals.

COEPD has innate leaders running the organization with an irrevocable human element to serve the ever changing dynamic environment of the IT industry. COEPD takes personal care in grooming our BA aspirants from the initial phases of conducting counseling sessions to the final phases of nurturing them into the IT industry Business Analyst role, our forte.

Business Analyst a key role in creating project success stories to every IT company in the industry. We at COEPD teach the subject with professional interest keeping in view of the challenging areas of the IT Industry.

COEPD workshops will be conducted and delivered by industry experts having more than a decade of work experience in handling IT projects in the ranks of a Project Manager, Principal Business Analyst , Senior Business Analyst and CIO.

COEPD delivers training through various channels: Workshops, Online Trainings, Corporate Trainings, Classroom Trainings and Certificate Trainings.

Training Locations

COEPD Talent in Corporates

Our Happy Participants Say it All

"Providing much needed clarity on topic with practical Case Studies." -T Shekar

"Coepd seems to be the topmost & qualitative institute in Hyderabad." - Kiran

"Faculty is highly knowledgable about the current IT industry." - Rakesh A

"The best BA Training Institute that I have ever come across and strongly recommend my friends to COEPD." - Vikrant

• BA Trainings
• BA Internships

More About Coepd

• Our Workbook
• Our Talent in Corporates

IIBA® , INTERNATIONAL INSTITUTE OF BUSINESS ANALYSIS®, CCBA® , CBAP®, and CERTIFIED BUSINESS ANALYSIS PROFESSIONAL® are registered trademarks or certification marks owned by International Institute of Business Analysis in one or more countries. ECBA™, CBATL™, Entry Certificate in Business Analysis™, Certification of Capability in Business Analysis™, and Certified Business Analysis Thought Leader™ are trademarks owned by International Institute of Business Analysis. EEP and the EEP logo are trademarks owned by International Institute of Business Analysis.