Research on the application of NLP-driven public opinion sentiment tendency analysis in precision marketing of new energy vehicles

Abstract (background sentence, research objectives, specific implementation description + research results).

Introduction

Since the beginning of the new century, the rapid development of China's new energy industry, represented by new energy vehicles, has shown the good momentum of China's industrial upgrading and the strong resilience of the economy. Since 2015, China's new energy vehicle production and sales have ranked first in the world for nine consecutive years, and it is one of the new industrial tracks with the fastest growth and the best development prospects in the new energy industry. However, in recent years, although the manufacturing of new energy vehicles has been significantly improved in terms of environmental protection and energy saving, it has still attracted much attention in terms of battery attenuation, slow charging speed, and high maintenance costs. Extensive public opinion discussions have been carried out on charging facilities and after-sales service, which has reduced users' brand perception of new energy vehicles and affected the marketing strategies of automobile companies.

Figure 1: Monthly sales and growth rate of new energy vehicles

It can be seen that in today's era of online new media, online public opinion often affects consumers' purchase decisions. The paper will conduct NLP natural language analysis by collecting user reviews on a number of mainstream video platforms and professional automotive websites, including but not limited to Douyin, Bilibili and other major video websites, as well as professional automotive information and forum websites such as AutohomeCombined with brand positioning, publicity, policy, and public awareness, that is, the specific implementation of the concept of green environmental protection, the impact on the public opinion orientation of new energy vehicles is specifically analyzed. Based on the above research and analysis, the relevant strategies to promote the development of domestic new energy passenger vehicles are discussed, so that enterprises can more accurately understand the public's attitudes and perceptions towards new energy vehicles, formulate more effective marketing strategies, and enhance brand image, so as to achieve the purpose of improving customer satisfaction.

Theoretically, this paper can not only verify the predecessors, but also analyze the sentiment tendency of the comment data of network users with the help of network data collection technology and NLP technology, quantify the trend of public opinion, verify the operability of the technology, and enrich the research literature on public opinion of new energy vehicles. Moreover, in practice, it helps automobile companies to formulate more effective marketing strategies, improves the brand influence and user perception of automobile companies, and then drives commercial profits.

Literature Review

The status quo of big data precision marketing

In the era of online new media, online public opinion will affect consumers' consumption behavior, and Chen Jing believes that digital media, such as the Internet and social media, provide consumers with a wealth of information acquisition channels and consumption methods [1]. Mainstream online media platforms will use big data technology to collect and analyze user information, understand consumers' daily preferences, and accurately push corresponding content for them to guide consumers. Fang Fusheng believes that more valuable information can be obtained from massive information, and through the collection and collation of this valuable information, it can better serve the marketing work of enterprises [2]. Mainstream short video platforms such as Douyin and Kuaishou use big data technology to collect information and accurately promote content that users are interested in as their main business strategy.

Overview of NLP sentiment analysis techniques

NLP is an important direction in the field of artificial intelligence and computing, which is used to study the methods and theories of effective communication between humans and computers in natural language. Jia Guosheng believes that NLP sentiment analysis technology can help businesses understand consumers' evaluation of products in the economic field, and promote product improvement to make them more in line with consumer needs [3]. As the core field of NLP, sentiment analysis can help us explore the subjective emotions of texts, and Chen Siyuan believes that NLP technology can not only analyze the literal meaning of the text, but also analyze the emotional trend behind the literal meaning [4]. Analyze the emotional orientation of consumers to the product and improve the market value of the product for the enterprise.

Figure NLP Technology Roadmap

This paper is based on theNLP task - sentiment analysis carried out by Tran S Forme R, using the Transformer architecture, because the task is text classification, only the Encoder part of the architecture is referenced, which is shown by the left part of the figureThe model analyzes consumer evaluation based on consumer evaluation of new energy vehicles as input emdedding, and NLP sentiment analysis technology as Positional Encoding. Each new media network platform acts as Multi-Head Attention, providing more data support for the final output. It shows the important role of NLP sentiment analysis in driving consumer consumption.

Current status of public opinion analysis and research

In the development process of the new energy vehicle industry, not only the research on the technology and policy of new energy vehicles, but also the impact of NLP sentiment analysis on the development of new energy vehicles. Song Chunyan believes that in-depth analysis of consumer behavior characteristics can provide enterprises with accurate market insights and market dynamics, and further optimize marketing strategies and market positioning [5].。 Hou Heshui believes that with the optimization of the dimension and system of consumer reviews, user reviews also have more in-depth reference value [6]. Based on the research of the former, it is found that Song Chunyan only analyzes consumers' emotional attitudes, and Hou Heshui only focuses on the analysis of consumers' purchase intentions, both of which only stay under the influence of public opinion on consumers, and the analysis of car companies' perception and response to user reviews is relatively insufficient.

Theoretical basis

In this study, the AISAS model was used as the analysis model.

Attention: Han Mengdi believes that the marketing methods proposed by enterprises in the corresponding market can help enterprises increase their brand influence and market share[7]. Through various marketing means, such as showing the advantages of the appearance characteristics of new energy vehicles through creative advertising or using the influence of celebrities, increase the exposure of products, let consumers pay attention to the information of the brand or product, make it stand out from the mass of products, and attract the attention of consumers or potential consumers.

Interest: Jia Jia explores the phenomenon of different consumer attention to different attribute areas of different visually complex products [8]. It is extremely difficult to find that it is extremely difficult to attract the attention of consumers, so marketing means are extremely important to attract the attention of consumers, so as to further show consumers the advantages, characteristics and unique value of the product, such as environmental protection and energy saving of new energy vehicles, intelligent driving systems and other content to stimulate consumers' interest.

Search: Long Liujiang pointed out that rational consumers will conduct pre-purchase information searches to ensure the utility of purchasing the product for themselves[9]. Therefore, when consumers are interested in a product and want to know more about it, they will actively search for more details about the product and brand through search engines, social media and other channels, and consumers will focus on comparing car performance, user reviews, after-sales service and other aspects in order to make more informed purchase decisions.

Action: Wang Ying believes that the message conveyed by the media penetrates into the hearts of consumers and can effectively move and persuade consumers [10]. After understanding and considering, consumers will choose to buy products that meet their expectations according to their own needs, and new media marketing provides consumers with more convenient consumption channels through online and offline interaction and integration, and promotes purchases by holding preferential activities, financial loans and other programs.

Share: Wang Xilian pointed out that under the current communication pattern with new media as the main communication channel, consumer behavior not only includes information acquisition and purchase behavior, but also includes the use of multiple channels to share brand reputation, and it is necessary to include it in the evaluation system [7]. After purchasing and using, consumers often share their feelings on new media channels such as social media, which will become a new source of information to attract the attention of others, thereby attracting the attention of more potential consumers. Positive user reviews can further enhance the brand image and market recognition of new energy vehicles, thereby increasing consumer trust and attracting more consumers to buy new energy vehicles. At the same time, these comments also play an important role in the optimization and upgrading of enterprises and the adjustment of business strategies.

Figure AISAS Flow Chart

In the automotive industry, more and more consumers choose to rely on online information and social media to understand products, Fu Shuai pointed out that car buyers will be invaded by various information when obtaining automobile information, so that their attention is constantly weakened when obtaining information[8], so public opinion has become an important factor for consumers to understand products, whether it is positive public opinion or negative public opinion, will affect consumers' purchase decisions, affect the sales of new energy vehicles.

From the perspective of consumer purchase decisions, positive evaluations can increase consumers' willingness to purchase new energy vehicles. Consumers often refer to various reviews to understand a car. If the reviews are positive, they are more likely to choose to buy; Conversely, negative reviews can cause consumers to hesitate or abandon a purchase. In response to these evaluations, Ren Hengxin pointed out that, on the one hand, through the study of the influencing factors of consumer satisfaction of new energy vehicles, automobile companies can better understand the needs of target groups, formulate targeted product improvement strategies, improve product satisfaction, and enable enterprises to seize opportunities in the fierce market competition [9].。 On the other hand, for consumers, their satisfaction can accurately express their demands for new energy vehicle products and services, prompting automobile companies to continuously improve and provide products and services that can better meet the needs of consumers. Xue Hanxin pointed out that the higher the perceived value of a product, the more conducive it is to the formation of positive cognitive and emotional attitudes, and the more likely it is to generate the purchase intention of the product [10].

From the perspective of market competition, Bai Mei pointed out that the competitiveness of the new energy vehicle industry can be studied from the perspectives of market competitiveness, integrity of the industrial chain, public infrastructure guarantee ability, innovation ability, and market openness [11]. However, ignoring the impact of public opinion on the competitiveness of the new energy vehicle market, new energy vehicles with good evaluations will increase the share of their brands in the market.

To sum up, the public opinion of the new energy vehicle industry is an important factor affecting the development of new energy vehicles, which can effectively promote the development of the field of new energy vehicles.

Reaserch Methods

3.1 Research Subjects

This study focuses on the evaluations and public opinions expressed by users on major platforms on the models of six brands that occupy a leading position in China's new energy vehicle market, such as Tesla, Xiaomi SU7, and BYD, as the main analysis object, and carefully selects a series of research indicators to construct a database as the follow-up data processing and public opinionThe foundation of data analysis. The reason why the top six models in China's new energy vehicle market are selected as the research objects is that these brands and models show a high degree of representation in China's new energy vehicle market, and the comments made by users on various platforms can also reflect the current situation of the new energy vehicle market relatively truthfully.

3.2 Data acquisition and preprocessing

In this study, data crawling technology was used to crawl the public information of various platforms. The data mainly comes from crawling a number of mainstream video platforms and professional automobile websites, such as Douyin, Bilibili and other major video platforms, as well as professional automotive information and forum websites such as Autohome and Knowing Chedi. After collecting the original data, strict data cleaning was carried out to screen the relevant content on the platform, such as the expectations of new energy vehicles, the experience of use, suggestions for correction, and the impact of automobile sales and policies, including removing duplicate comments, screening invalid information, and correcting erroneous data, to ensure the representativeness, availability and authenticity of the data. Finally, the collected data is classified and sorted, which provides a solid data foundation for us to conduct in-depth research on the analysis of users' public opinion on new energy vehicles.

3.3 Data Analysis Methods

The analysis mainly includes two parts: theoretical basis and NLP sentiment analysis.

In the theoretical basis part, by studying the research methods and conclusions of relevant scholars in the literature review, the empirical methods suitable for this study are refined and summarized, and a new framework of sentiment analysis based on AISAS is constructed on the basis of understanding the theoretical principles and empirical principles.

In the NLP sentiment analysis part, this study mainly uses NLP natural language analysis to analyze the impact on the public opinion orientation of new energy vehicles。 Firstly, according to the collected data, the NL P task was performed based on the TRAN Sformer stack, and the positive and negative corpora were customized for training, and the model file was generated. The second is to construct an emotion distribution map to intuitively count the overall comment sentiment; The third is to calculate the average emotional prediction value of each specific new energy vehicle and obtain the corresponding emotional attitude of each user; The fourth is to further create a sentiment analysis diagram based on the data, and pave the way for the final word cloud diagram elaboration while concretely demonstrating the first two data. Here's a prediction: the proportion of negative reviews on new energy vehicle models is higher than that of neutral or positive reviews.

4. Research process

4.1 Sentiment distribution map and predictive value analysis

First, the study integrated nearly 60,000 reviews and comments of six models to obtain an emotional distribution map, of which 33,277 were negative comments, more than half; There were 19,619 positive reviews, accounting for about 30%; There were 8,033 neutral evaluations, accounting for about 15%.

Figure Diagram of the distribution of vehicle sentiment after summing

Here, the predictive value of each review was evaluated by NLP text analysis, and the predicted value of each review indicated the emotional strength of the review, and the study set the range on a scale of 0-1, where 0 represents negative emotion and 1 indicates positive emotion. The following are the average forecasts for the six major models:

Figure Predicted average in sentiment analysis table

The results show that except for the Xiaomi SU7, the average sentiment prediction is 0.5895, and the other models are dominated by negative evaluations, with the prediction value below 0.5 points, and there are no models dominated by neutral evaluations in the data results.

4.2 Sentiment analysis graph research

In order to better support the above mean, this study uses NLP text sentiment analysis based on the above review data and uses the sentiment analysis graph to more intuitively display the sentiment score and distribution of reviews in each model.

Figure Overall sentiment analysis chart of the reviews of the six major new energy vehicle brands

Overall, the positive and negative score range of the six major models is polarized. First of all, the area in the picture is almost dominated by red, and the depth of the color hints at the intensity of the emotional tendency in the review, whether it is a positive or negative comment, red indicates that the emotional description of the review of various models is more prominent and obvious. Secondly, by observing the sentiment score on the vertical axis, the distribution of the higher score, that is, the comments with strong emotional tendency, is mainly concentrated in 1.0, and the distribution is gradually sparse and the number decreases greatly with the decrease of the score, while the comments with weak emotion are mainly concentrated in 0.0-0.2, and the distribution is relatively dense, and the number accounts for a large proportion. Thirdly, according to the comment number on the horizontal axis, it can be seen that there is a certain degree of difference in the sentiment score of the comment in different serial number segments.

From the perspective of specific models, the evaluation of Ideal L7, Wenjie M7, Wuling Hongguang, Tesla and BYD Qin PLUS is more in a negative and negative form.

Figure Ideal L7 sentiment analysis diagram

As can be seen from the figure, the review score of ideal L7 is mainly concentrated in 0.0 points, and the distribution is relatively balanced between 0.2-0.6 points, and the intensity of emotional tendency is generally low, so the negative emotions of the whole evaluation are more obvious.

Figure M7 sentiment analysis diagram

As can be seen from the figure, the comment scores of M7 are mainly concentrated in 0.0, 0.5 and 1.0 points, but the color is darker around 0.0 points, and it is worth noting that the blue part of the area in 0.5-0.9 points indicates that the negative attitude intensity of this part is large, and the negative sentiment of the whole evaluation is also more obvious.

Figure Wuling Hongguang sentiment analysis diagram

As can be seen from the figure, Wuling Hongguang's review scores are mainly concentrated in 0.0 and 1.0 points, which are more dense at 0.0, and in the blue part of 0.5-0.9 points, the positive and negative review scores are more evenly gathered, but the review scores are generally low, and the number of negative evaluations is significantly ahead of the positive evaluations.

Figure Tesla sentiment distribution chart

As can be seen from the figure, Tesla's review scores are mainly concentrated in 0.0, 0.5 and 1.0 points, although there is a certain degree of difference in emotion, but the overall review score is low, that is, negative reviews are also dominant.

Figure BYD sentiment distribution map

As can be seen from the figure, BYD's emotional score is mainly concentrated around 0.0 points, and this part of the area shows a trend of lower scores, the denser the dots, and among the 0.5-1.0 points, there are fewer comments with strong emotional tendencies, and most of the comments are still low.

Figure: Xiaomi SU7 sentiment analysis diagram

Among the six major models,The more prominent is XiaomiSU7,As can be seen from the figure,The positive comment area with high emotional tendency intensity is mainly concentrated around 1.0,And the number is relatively large,At the same time,Although some areas have blue,That is, the distribution of negative emotions,But their emotional score is low,The corresponding emotional performance is weaker, It shows that the influence of negative reviews is small, and the performance of positive reviews is relatively good.

4.3 Positive and negative word cloud display under descriptive statistics

Based on the data presented in the first two steps of the research process, the research enters the keyword extraction stage of user feedback analysis, where the positive and negative sentiment comments are first classified through the code, and then the high-frequency words of the two types of comments are counted, which paves the way for the research results of the next part while improving the differentiation of user evaluation keywords.

Figure Word cloud diagram on positive reviews

The image above represents keyword extraction under positive reviews. From high frequency to low frequency, the key words are: Xiaomi, price, battery life, feeling, intelligent driving, charging, safety, car owner, high speed, new model, automatic, leading.

Figure Word cloud diagram about negative comments

The image above shows keyword extraction under negative comments. From high frequency to low frequency, the key words are: kilometer, charging, no, car pickup, battery, discount, abnormality, maintenance, insurance, comparison, subsidy, air conditioning. These are high-frequency feedback words from users, so they will be highly valued in the context of AISAS elaboration in the following article.

5. Results and Discussion

5.1 Results

The results of the study reveal some key trends and characteristics in the public opinion of new energy vehicles. First of all, we found that among the six major new energy vehicle brands, negative reviews accounted for a high proportion of 55.46%, while positive reviews accounted for 30% and neutral reviews accounted for only 14.55%. The reason for this result may be that some new energy companies ignore the focus on customer evaluation and pay more attention to automobile research and development.

Using NLP text analysis, we evaluated the sentiment intensity of each review on a scale of 0-1, and the results showed that except for the Xiaomi SU7, which had a relatively high sentiment prediction value (0.5895), all other models had a low sentiment prediction value, indicating that negative reviews were dominant. This finding echoes the view mentioned in the literature review that consumers' support for new energy vehicle products is relatively high, but the product support is gradually declining due to the influence of negative events and online public opinion.

The sentiment analysis chart visually shows the sentiment score and distribution of each model review, showing the polarization of emotional tendencies. The sentiment scores of most models are dense at both ends (0.0 and 1.0), which may be related to consumers' high expectations for new energy vehicles and the gap in actual use. As mentioned in the literature review, consumers' willingness to buy new energy vehicles is influenced by the evaluation of the surrounding population, which may explain why there is such a clear polarization in the reviews.

We found that the reasons for the negative evaluation of new energy vehicles include product cost performance, comfort, negative review guidance, and false or exaggerated publicity by car companies. This finding is consistent with the literature review mentioning that consumers' evaluation of NEVs focuses on technological development, policy and market environment analysis, but ignores consumers' evaluation of NEVs. Despite the large number of negative reviews, the sales of new energy vehicles have continued to grow, which may be related to technological innovation, policy promotion, and increased consumer acceptance of the concept of green mobility. This result echoes the rapid development of the new energy vehicle industry mentioned in the literature review, showing the good momentum of China's industrial upgrading and the strong resilience of the economy.

In terms of the analysis of the perception of reviews by car companies, the survey results show that car companies do not attach much importance to user reviews, and some car companies pay more attention to product R&D innovation and market strategies, while ignoring the importance of user reviews. This finding is consistent with the view mentioned in the literature review that public opinion, as an external factor in the development of new energy vehicles, can help new energy vehicle companies determine their production behavior according to their own costs and benefits, as well as in combination with the market economy and regulatory environment. These research results not only theoretically enrich the research literature on public opinion of new energy vehicles, but also provide a new perspective and theoretical structure for public opinion research in the field of new energy vehicles in practice, which is helpful to achieve the purpose of improving customer satisfaction.

5.2 Problem Discovery

In terms of negative evaluation of new energy vehicle models, the reason why the proportion is relatively high is first of all in the cost performance of the product, consumers are more willing to screen out the better one in the same category, even if some models have no defects in this regard, but because there is no outstanding point, it greatly reduces the expectations of users, and it is easy for users to have negative evaluations. Secondly, in terms of comfort, some brands have improved the user's experience to a certain extent by improving technology, but due to unreasonable cost or configuration, the original comfort has been weakened. Then, in the original negative review guidance, because consumers pay more attention to negative evaluations, it is easy to have doubts in personal trade-offs, which eventually leads to distrust of their own models, and some car companies do have false and exaggerated publicity, and consumers cannot identify the real situation, resulting in a more ambiguous final evaluation and a more negative attitude.

In terms of sales, the reason why they still show good sales on the basis of a large number of negative evaluations is because of the continuous optimization of scientific and technological innovation, especially in the core technology, which has created an advanced and intelligent atmosphere among users and enhanced their attractiveness to consumers. In terms of policy promotion, at present, China is vigorously promoting green development, especially in the field of new energy vehicles, such as financial support for car companies, user consumption subsidies, and publicity of promotional documents, which have promoted the vigorous development of the new energy vehicle market. But in general, this also reflects the wide variation in the perception of car companies based on user evaluations.

In terms of the importance of car companies based on user reviews, some car companies pay more attention to product R&D and innovation, and the long-term formulation of market strategies, while ignoring the importance of user reviews. At the same time, the instability of the positive and negative orientation of reviews and the inconsistency of review feedback channels also greatly reduce the efficiency of comprehensive analysis of reviews by car companies.

5.3 Research Recommendations

Based on the above analysis, after review and comparison, for example, the reason why the predicted value in the sentiment analysis table of Xiaomi SU7 is higher is because the appearance of Xiaomi SU7 is more in line with the aesthetics of contemporary young people, and it is also because Xiaomi SU7 gives customers more choices in vehicle configuration, and can customize the car configuration and diversify aesthetics. In terms of price, the pricing is reasonable, and before the release, the popularity of Xiaomi cars is also relatively high. Therefore, the new energy vehicle industry can meet consumers more in terms of personalized configuration, improve comfort, appearance conditions, and cost performance. Communication with customers is also very important, not only focus on innovation and research and development, but also need to establish contact with consumers, strengthen after-sales service, according to the suggestions provided by customers, change or give a reply, which is conducive to the contact between customers and customers, customers and manufacturers, which can greatly increase the customer's desire to buy, and then enhance brand sales.

Based on the above problems, car companies can take the following measures:

To improve the outstanding features of products, car companies can deeply understand consumer needs, improve the cost performance of products through technological innovation and cost optimization, and find and highlight their own unique selling points, so as to improve users' expectations and satisfaction; Optimize comfort and configuration, increase R&D investment in comfort to ensure the rationality of configuration, and avoid excessive stacking to improve user experience; Actively respond to negative reviews, and the official responds to consumers' doubts and dissatisfaction in a timely manner. False or exaggerated propaganda should be corrected to ensure the authenticity of the propaganda content; By attaching importance to user reviews and feedback, car companies can increase the importance of user reviews on the platform, include them in the reference scope of product improvement, actively respond to them and take measures to solve them, and enhance users' trust in the brand.

According to the AISAS model, to improve the marketing ability of big data, it is suggested that the new energy vehicle industry can increase publicity efforts first; Demonstrate product advantages and arouse consumer interest; When customers are attracted, it will be easier to turn potential users into consumers and search for products; The product performance is very high, and it is easier for customers to choose to take action to buy after the search is completed; Enterprises do a good job of communication with customers, and have a good attitude towards consumer service, which is easier for consumers to recommend products to others.

6.Conclusion

Through empirical analysis, this paper reveals the main problems and trends in the public opinion of new energy vehicles. Despite the negative reviews of new energy vehicles, their sales continued to grow, indicating that the market demand for new energy vehicles is strong and positively affected by policies and technological innovations. However, the lack of perception and response to user reviews by car companies may affect brand image and market competitiveness. Therefore, it is suggested that car companies should pay more attention to the voice of consumers, improve the cost performance and comfort of products through technological improvement and innovation, and strengthen communication with consumers to improve customer satisfaction and market competitiveness. In addition, policymakers should continue to support the development of the new energy vehicle industry and promote the healthy development of the industry through policy guidance and market incentives. On this basis, the research in this paper not only theoretically enriches the research literature on public opinion of new energy vehicles, but also analyzes the sentiment tendency of network users' comment data with the help of network data collection technology and NLP natural language processing technology, which provides a way to quantify the trend of public opinion.

7.Reference

CHEN Jing. The influence and analysis of new media marketing on consumers' purchase decisions[J].Time-honored brand marketing,2024

FANG Fusheng. Analysis of enterprise marketing strategy in the era of big data[J].Market Outlook,2024,(10):23-25.)

JIA Guosheng. Research on aspect sentiment analysis based on deep learning method[D].Jilin University,2023.

CHEN Siyuan. Application of NLP technology in enterprise procurement management under the background of big data[J].China Logistics & Procurement,2024

[5] Song Chunyan. Research on consumer behavior characteristics and sentiment analysis of new energy vehicles based on text mining[D].Guizhou University,2024.

[6] Hou Heshui. Analysis of user reviews of new energy vehicles based on sentiment analysis and theme extraction[D].Shanghai Normal University,2024.

[7] Han Mengdi. Research on the marketing strategy of M snack food company from the perspective of consumer behavior[D].Henan University,2022.

[8] Jia Jia. Research on the influence of product description page layout on consumers' attention, memory and preference[D].Southwest Jiaotong University,2021

[9] Long Liujiang. The influence of consumers' pre-purchase information search behavior on the choice of sales mode of manufacturer's platform[D].Southwest Jiaotong University,2023

[10] Wang Ying. The influence of mass media communication contact on consumption[J].China Media Science and Technology,2014

[11] Wang Xilian, Wang Hong. Research on brand reputation evaluation system of automobile evaluation from the perspective of AISAS model[J].Modern Business,2024,(15):7-10.)

[12] Fu Shuai. Analysis on the business monetization path of automotive new media based on AISAS model[J].World Automobile,2024,(05):46-49.)

[13] Ren Hengxin. Research on product satisfaction improvement strategy of new energy vehicle based on online reviews[D].Chongqing Technology and Business University,2023

[14] Xue Hanxin. Research on the influence mechanism of perceived product innovation on the purchase intention of new energy vehicles[J].Journal of China University of Petroleum(Social Science),2023,39(06):84-91.)

[15] Bai Mei. Price Theory & Practice,2020,(01):25-31