Spatial heterogeneity and its influencing factors of Douyin network: Attention to 5a-level scenic spots in China

It can be seen from the review of the relevant literature that, in terms of data sources, most studies have focused on using the Baidu index as a measure of network attention to conduct research on tourist attractions (Jiao et al., 2022; Su & Kang, 2022; Wang et al., 2022; Yan et al., 2021; Yao & Liu, 2020; Zhang et al., 2022). For example, Jiao et al. (2022) examined the distribution pattern and the driving mechanisms of network attention to 300 classic red tourism attractions in China, using the Baidu index as the data source. Some scholars have explored the effectiveness of online promotion of scenic spots using various social media tourism data (Duan et al., 2020; Knight, 2014; Lu et al., 2022a; Li, Tao, et al., 2021; Zang & Wang, 2020; Zhao, Zhao et al., 2022) as indicators of network attention. For example, Zhao, Zhao et al. (2022) analyzed spatial differences in network attention and its influencing factors for desert scenic spots at the A-level in China using data on the search engine index, tourism website index, social media index and short video index. In terms of research methodology, most studies focus on exploring the spatio-temporal differences in attention to scenic spots using various measurement indicators (Lu et al., 2022a; Zhang et al., 2023). For instance, Lu et al. (2022a) combined methods such as mathematical statistics, spatial analysis, and geographic detectors to analyze the spatio-temporal differences and influencing factors of tourism in Xi’an. Regarding the scope of the research, the study encompasses multiple scales, including national (Lai, 2022; Liang & Lijun, 2023), provincial (Alvarez-Diaz, M. et al., 2020), and urban (Ding et al., 2022; Xue & Bai, 2023) levels. Research has shown that network attention serves as an important indicator for measuring the level of information dissemination and public interest in tourist attractions. Not only does network attention benefit the positioning of tourist attractions, the precise planning of tourism, and the promotion of unique features of tourist attractions, but it also has significant implications for enhancing the quality of tourist attractions (Krisjanous, 2016; Luo et al., 2023).

Short videos not only serve as a means of entertainment and documenting daily life but also possess multiple functionalities such as fan interaction, professional learning, and live shopping. As a result, the download and viewing figures of short video applications have surpassed those of traditional news media (Ding et al., 2022). Research on “short videos + tourism” has also emerged in large numbers, such as Xu et al. (2023) exploring the factors influencing the willingness of short video platforms to search for travel information, providing theoretical and practical insights for improving users’ willingness to use short video platforms as a source of travel information, thereby making rational travel decisions. Shu et al. (2023) explained the transformation of virtual and reality in tourism short videos, highlighting insights for tourism marketing practices. Fang et al. (2023) pointed out that short-form travel videos are popular and can spark travel inspiration. In addition, scholars have studied the impact of destination short videos on customer attitudes (Cao et al., 2021), the impact of short videos on willingness to share online (Zhao, Shen et al., 2022), and the impact of short videos on tourists’ destination decisions (Jiang et al., 2022).

In this regard, as an emerging media form, Douyin (a Chinese social media platform) has garnered significant attention from a large user base due to its unique operational mechanisms and marketing strategies. By 2022, Douyin had reached a daily active user base of 500 million, establishing itself as a popular short video application (Liu et al., 2022). Academic research has shown that analyzing the user’s network attention can not only be used to study the power of dissemination and the effects of networked information on the Douyin platform, but also to some extent reflects the level of public’s interest and the variations in interest in different scenic spots. In a sense, the number of Douyin’s fans for a scenic spot is an important indicator for measuring the online development of the scenic spot and users’ attention, representing the quality of physical construction, influence, and attractiveness of the scenic spot. With the rise and development of the short video + tourism model, there has been an increasing amount of research in the field of tourism based on the perspective of Douyin short videos. For example, Mao et al. (2023) studied how the novelty of travel-themed short videos on Douyin and the shooting angles influence potential tourists’ behavioural intentions. Zeng and Li (2019) explored the constituent elements and the formation process of interactive communication in Douyin tourism short videos from a sociological perspective. However, most current research is mainly focused on examining the perspectives of problem solving and strategic thinking about customer attraction in the context of short video promotion; there is a lack of analysis of the relationship between short video promotion and physical development in the context of the digital economy (Zhang et al., 2019).

In summary, scholars both domestically and internationally have conducted increasingly comprehensive and in-depth research on network attention towards tourist attractions, which has laid the theoretical and methodological foundation for tourism-related studies. However, there is still relatively limited research on spatial differences in network attention towards tourist attractions and driving factors from the perspective of short videos + tourism. It is worth noting that 5A-level tourist attractions, as the highest standard for quality grading and development of tourist attractions in China, have a significant impact on the measurement of network attention levels towards tourist attractions. Currently, there is scarce research in academia on the attention and spatial differences of short videos towards 5A-level tourist attractions. Therefore, this study aims to use the number of fans on the Douyin platform of tourist attractions as a measurement indicator to explore the network effects and influencing mechanisms of network attention towards 5A-level tourist attractions in China. It hopes to objectively analyze the current development status of network attention towards 5A-level tourist attractions in China under the new scenario of short videos, and provide valuable strategic recommendations for the construction of related tourist attractions, the integration of virtual and real economies, and the enhancement of the soft power of tourist attractions.

This study takes into the account the 318 5A-level tourist attractions identified on the official website of the Ministry of Culture and Tourism of the People’s Republic of China as research subjects (Table 1). Meanwhile, utilizing the Baidu Maps coordinate system, the latitude and longitude coordinates of each attraction were obtained, resulting in spatial location data for the attractions (Figure 1).

Figure 1:

Spatial distribution of China’s 5A-level tourist attractions

The number of Douyin’s fans of a scenic spot represents the level of users’ attention and recognition towards the related tourist destination. Therefore, this study adopts the number of Douyin’s fans to characterize the development level of network attention towards tourism destinations. Through data comparison analysis, it was found that the official Douyin account of scenic spot and the Douyin travel blogger accounts with more than 10,000 fans have a higher quantity and quality of Douyin, and they can be regularly updated, indicating good validity of the data. Therefore, this study selected data of official accounts of 5A-level scenic spots and travel bloggers on Douyin with a fan count of over 10,000 (as of August 23, 2023). The data were filtered and organized, excluding invalid data. The specific search process is shown in Fig. 2. The influencing factor data were sourced from official statistical yearbooks, national economic and social development statistical bulletins, official websites of tourist attractions, etc. Some missing data were obtained from official platforms of various regions (such as the Communication Administration Bureau and the Department of Culture and Tourism).

Figure 2:

Data retrieval route

Examining the level of Douyin fans’ attention to China’s 5A-level tourist attractions, this study was conducted using 1394 Douyin accounts. A fan count rank-size distribution chart of the tourist attractions was created (Figure 3). From Figure 3, it can be observed that the goodness of fit for the number of Douyin’s fans is 0.74, indicating a relatively high level of fit. This suggests that the level of Douyin’s fan attention to scenic spots is in line with the rank-size rule. Additionally, q = 2.0409, indicating that highly ranked scenic spots are well developed, while lower-ranked scenic spots are less developed. Furthermore, the overall structure of Douyin’s fan attention to scenic spots exhibits significant imbalance characteristics. Specifically, the head of high-ranking scenic spots is characterized by a small number and scattered distribution, with no significant difference in Douyin’s fans between these spots. The majority of scenic spots have Douyin’s fans concentrated in the middle of the fitted curve, and the hierarchical levels are not clear. The tail of low-ranking scenic spots bends downwards in an almost vertical state, indicating a significant difference in the level of attention between these spots. In summary, the Douyin’s fan base of scenic spots exhibits a dispersed head, closely packed middle, and vertical tail characteristic on the rank-size curve, reflecting the widespread and significant development of high-ranking scenic spots, while the development of mid-to-low-ranking attractions is insufficient, and the overall structure of attention level is characterized by an imbalance.

Figure 3:

Rank-size distribution of network attention to Douyin’s fans

Research has shown that if a scenic spot is promoted or marketed with distinctive characteristics, it will generate viral videos and create a multiplied effect in online publicity, thereby attracting public attention to the scenic spot and further enhancing its popularity. Therefore, this study is based on the number of fans of Douyin accounts as an indicator, and through a comprehensive analysis of the short video effects on the Douyin platform of China’s 5A-level tourist attractions, a clearer understanding of the online promotional effects of these attractions can be obtained. From Figure 4, it can be observed that there are significant variations in the spatial distribution of different levels of attention in tourist attractions. Specifically, high-attention attractions are mainly distributed in Zhejiang Province, with the Hengdian World Studios and Putuo Mountain scenic spots performing exceptionally well. Tourist attractions with higher attention are mainly concentrated in the Yangtze River Delta, Shanxi, Shaanxi, and Henan provinces, as well as the Yunan and Hunan provinces. Tourist attractions with moderate attention are clustered in Beijing, Tianjin, Hebei, Shandong, Shanxi, Henan, Guangdong, Chongqing, and Sichuan. On the other hand, tourist attractions with low and lower-attention exhibit a wide range and a scattered distribution pattern.

Figure 4:

Spatial pattern of fan quantity and attention level on Douyin

Meanwhile, the Jenks natural break classification method was used to analyze and examine the spatial structure of the number of fans on Douyin; they were divided into five levels. This analysis aimed to understand the spatial differentiation of public attention. The results are presented in Table 2 and Figure 4. Additionally, to better illustrate the spatial agglomeration of attention to scenic spots, the kernel density analysis method was applied to analyze the spatial agglomeration characteristics of the number of Douyin’s fans in 318 5A-level tourist attractions in China. The results are shown in Figure 5.

Figure 5:

Density analysis of Douyin’s fans for China’s 5A-level tourist attractions

Note: The map is drawn based on the standard map of the Ministry of Natural Resources of China (Map Approval Number GS(2019)1822), and the base map remains unmodified. Data for Hong Kong, Macau, and Taiwan are not available.

According to Table 2 and Figure 4, overall, low-attention scenic spots account for 77.4% of the total number of scenic spots, while the proportion of high-attention scenic spots is 3.5%, and only 0.6% are high-attention scenic spots. This indicates that low-attention scenic spots still dominate in China and reflects the urgent need to further enhance the online popularity and dissemination of 5A-level tourist attractions. In terms of spatial pattern, there is a clear radiating pattern from provinces such as Zhejiang, Hunan, Anhui, and Shanxi towards the surrounding areas, forming three relatively dense spatial agglomeration areas: the Yangtze River Delta, Beijing-Tianjin-Hebei, and Sichuan-Shaanxi-Chongqing. It exhibits the characteristic of “low-level surrounding high-level.” There are 14 moderate-attention tourist attractions, accounting for 4.4% of the total. These include Jiuzhaigou Valley, the Palace Museum, Wugong Mountain, etc. This also objectively reflects the good endowment of the tourism resource to attractions and their potential impact on the attention effect of the Internet. The proportion of low and lower-attention scenic spots is as high as 91.6%, with a total of 291 sites distributed throughout the country. Representative scenic spots include Lingwu Shuidonggou Tourist Area, Tashilhunpo Monastery scenic spot, and Zhashilunbu Monastery scenic spot, indicating a clear underdeveloped state in western provinces such as Qinghai, Xinjiang, and Sichuan.

According to Figure 5, it can be observed that the overall attention on Douyin for 5A-level tourist attractions exhibits a spatial agglomeration pattern characterized by a structure of “one core and one belt.” Specifically, it not only forms a core area in the Yangtze River Delta, but also forms a core belt that encompasses Beijing, Tianjin, Hebei, Henan, Jiangxi, Hunan, and Hubei provinces. As for the core area of the Yangtze River Delta, centered around Zhejiang, it forms a core region of agglomeration and development, including Shanghai, Jiangsu, Anhui, and other areas. In addition, except for the Beijing, Tianjin, Hebei, Henan, Jiangxi, Hunan, and Hubei provinces region, the core density of scenic spots in other areas is relatively low. A comprehensive analysis reveals that the high popularity of China’s 5A-level tourist attractions on Douyin is predominantly concentrated in regions with developed tourism economies, abundant tourism resources, and distinctive branding campaigns. Furthermore, these highly popular attractions tend to generate a spillover effect on other regional tourist attractions.

Using the nearest neighbor index formula, an analysis is conducted on the spatial proximity of the level of Douyin’s fan engagement for 5A-level scenic spots. This analysis aims to determine the developmental status and types of spatial clustering of the scenic spots. It is worth noting that due to the limited number of highly popular tourist attractions, this study combines the high and higher-attention tourist attractions for calculation, as shown in Table 3. The research findings indicate that, as the concentration level decreases, the spatial distribution of Douyin’s fans of tourist attractions exhibits a pattern of initial agglomeration followed by dispersion, reflecting the underdeveloped quality of low-level tourist attractions. Specifically, low and lower-attention scenic spots exhibit a clustered state, while moderate, higher, and high-attention scenic spots tend to have a dispersed distribution state, indicating that such scenic spots have weaker interaction with the public on the Douyin platform. In addition, the number of scenic spots shows a significant decrease as the level of attention increases, and the concentration of scenic spots also decreases to some extent. This partially confirms the changing characteristics of the nearest neighbor index (ANN) of Douyin’s attention level for scenic spots, that is, the ANN values for scenic spots with low and lower-attention are all less than 1, while the ANN values for scenic spots with medium, higher, and high-attention are all greater than 1. The P-value for low-attention scenic spots is less than 0.01, indicating a significant clustering pattern in spatial distribution (Table 3).

Analyzing the level of Douyin attention towards 5A-level scenic spots and investigating influencing factors and their impact on the level of network attention towards these scenic spots holds significant practical implications for the rational development, integration, and construction of tourist resources. The level of network attention to scenic spots is a product formed by the foundation of tourism resources, the support of scenic spot construction, and the utilization of network-based publicity and marketing (Li & Zhang, 2018; Liu, 2016; Ren et al., 2021; Shu et al., 2020; Zhang & Wu, 2022). Therefore, when selecting influencing factors, not only should the connection with the development environment of scenic tourism, resource quality, and unique advantages be considered, but also the relationship with physical construction support should be taken into account. Furthermore, the effects of network publicity and professional media operation must also be considered. Through the review of existing literature on the factors influencing the network attention of scenic spots, and combining the objectivity and scientific validity of indicator quantification, as well as following the principle of data availability, this study selects eight indicator factors for the study, including marketing and market development of scenic spot, construction of scenic spot promotion platforms, regional economic development level, support of modern service industry, level of Internet development, convenience of transportation, regional population support, and vitality of tourism market development (Table 4).

Specifically, ticket prices are a representative indicator of the marketing market in scenic spots. As one of the important platforms for the marketing and promotion of scenic spots, the number of Douyin accounts can also reflect the level of scenic spot promotion platform construction to some extent. In addition, the network promotion of scenic spots needs to rely on the Internet and use qualitative and quantitative means of promotion to enhance the visibility of the scenic spot, so this study also considers the level of Internet development as an important influencing factor. The level of local economic development can provide support for the construction of scenic spot infrastructure and market-oriented operation, and it has an important economic support for improving the attention to the scenic spot network. Therefore, the regional GDP is selected to reflect the overall development level of the regional economy. The proportion of the tertiary industry added value can reflect the service support capability of the scenic spot. The passenger turnover represents the total amount of passenger transportation work in a certain period, and it is chosen to reflect the level of transportation development in the region where the scenic spot is located. The selection of regional population size serves as a reflection of the population magnitude supporting the geographic area where the scenic spot is located. In conclusion, this study comprehensively employs the aforementioned eight indicators as independent variables and investigates the mechanism of the influence of spatial heterogeneity in network attention to 5A-level tourist attractions in China, with the number of Douyin’s fans as the dependent variable. Regression analysis of the influencing factors is conducted using the geographic detector model, and the coefficients of the various influencing factors are ranked as follows: A2 > A1 > A3 > A8 > A6 > A7 > A4 > A5 (Table 5).

Based on Table 5, it can be observed that the construction of promotional platforms has a relatively significant impact on the network attention to scenic spots, while the level of Internet development has a weaker influence. Specific analysis: first, the tourist attraction promotion platform is an important channel for the attraction to communicate with the public. The public obtains tourism information by paying attention to hot topics in tourism. Therefore, the spatial differences in its network attention have a direct impact. Second, regions with higher levels of economic development often have higher levels of tourism development, scenic spot service quality, and modernization of online operations. In these regions, scenic spots with better service experiences also tend to attract greater attention from the public online. Furthermore, the development vitality of the tourism market has a significant impact on the network attention to scenic spots. The reason behind this is generally that the level of network attention is positively correlated with the public’s willingness to travel. To a large extent, the public tends to choose highly popular scenic spots to visit and engage in actual consumption activities, which in turn affects the local tourism revenue. Finally, the dependence of the attention to scenic spots on the level of Internet development is relatively weak, indicating that the Internet has a limited impact on tourists’ motivation. However, it is undeniable that the popularity and accessibility of the Internet still remain as the main factors influencing tourists’ willingness to travel. Moreover, the information network has become an important component of smart and digital scenic spot construction. Therefore, the level of Internet development is a progressive rather than a decisive factor.

The degree of explanatory power of the interaction effect of influencing factors on the spatial layout of network attention to 5A-level tourist attractions was analyzed using geographic detectors. From Table 6, it can be observed that the explanatory power of the interaction effect of the influencing factors on the spatial distribution of network attention to 5A-level scenic spots is higher than that of single factors, indicating the presence of two effects: dual-factor enhancement and nonlinear enhancement. This suggests that the spatial distribution characteristics of network attention to 5A-level tourist attractions are the result of the combined effects of multiple factors.

In the interaction results of the influencing factors, the q-values of q(A₁∩A₂), q(A₂∩A₄), q(A₂∩A₆), and q(A₂∩A₈) are all greater than 0.6, while the q-values of q(A₂∩A₃), q(A₂∩A₅), and q(A₂∩A₇) are greater than 0.85. This indicates that the construction of scenic spot promotional platforms is the dominant factor in enhancing interpretive power through interactions, reflecting the importance of developing scenic spot online promotional platforms.