Tourist attractions usually refer to areas with high-quality tourism resources that can provide tourists with tourism activities (Chen, 2012). At present, there is no unified definition of tourist attractions. In this study, we will follow the definition of tourist attractions in the national standard Quality Classification and Evaluation of Tourist Areas (points) (GB/T17775-1999), issued by the State Administration of Quality and Technical Supervision. In this standard, the “tourist attraction” is defined as an independent unit established with the approval of the administrative departments at or above the county level (including the county level); it should have a unified management organization and a clear scope, with functions of visitation, sightseeing, vacation, recreation, and knowledge seeking, as well as providing the corresponding tourism service facilities.

Most previous studies share a common view that tourist attractions have multiple attributes. Some scholars suggested that tourist attractions have both natural and social attributes, while others believe that tourist attractions have cultural and natural attributes. Besides, some other scholars divide tourist attractions into those with managed and unmanaged attributes (Zhang, 2004), charging and free attributes (Chen, 2012), or natural and man-made attributes (Li et al., 2023). Based on the existing research, this study considers tourist attractions to have two main attributes: natural attributes and artificial attributes. The natural attributes are represented by various tourism resources and environmental factors that exist and develop in the tourist attractions; while the artificial attributes are represented by various human activities such as operation activities and facility construction in the tourist attractions.

After determining the definition and attributes of tourist attractions, this study will first analyze the impacts of major public health events on the elements of the different attributes of tourist attractions. To make the analysis more systematic, we divided the duration of major public health events into three stages: pre-disaster, in-disaster and post- disaster. The impacts of major public health events in each stage will be analyzed to provide a theoretical analytic basis for the construction of the resilience evaluation system of tourist attractions.

The term “resilience” comes from the Latin resilire, which refers to the ability of a system or individual to recover from shocks or disturbances. Resilience was first applied to studies of physics. In the 1970s, Holling introduced it into studies of ecology and defined it as engineering resilience, which has been used to describe the ability of ecosystems to reorganize themselves and the rate of return to a stable state after the disturbance (Lamhour et al., 2023). With the deepening understanding of resilience, ecological resilience and evolutionary resilience appeared. In recent years, the connotation of resilience has been continuously expanded and enriched. Relevant studies of resilience have become popular in the fields of urban planning (Meerow et al., 2016), regional economy and tourism (Li et al., 2019). Tourism resilience is the extension of the resilience connotation in the field of tourism research, so it mainly refers to the ability of the tourism industry to cope with crisis risk, restore system functions and restructure the industrial structure when it encounters destructive shocks. The construction of a tourism resilience system is the basis for the tourism industry to achieve resilient growth and high-quality development, and is also one of the most important measures for promoting the high-quality development of tourism. As a micro-unit of the tourism industry, the resilience characteristics of tourist attractions are consistent with the resilience of the tourism industry, emphasizing the ability to prevent risks before disasters occur, the ability to maintain stability when disasters occur, the ability to return to normal after a disaster, and the ability to adapt to the new environment. Therefore, this study defines the development resilience of tourist attractions as the ability of tourist attractions to prevent risks under normal conditions, to withstand shocks and disturbances when disasters occur, and to restore system functions and adapt to new changes under the abnormal post-disaster conditions to achieve sustainable development.

In terms of tourism resilience research, the existing research objects of tourism resilience are mostly the tourism economic system, tourism industry, urban tourism system, community tourism resilience, and others. The research contents mainly include the definition of the concept connotation, the influencing factors and influencing mechanisms of tourism resilience, the evaluation of the tourism resilience level, and the promotion strategy of tourism resilience (Wang et al., 2020; Liu et al., 2021; Noorashid and Chin, 2021). In terms of the connotation of tourism resilience, scholars mainly discuss the concept and characteristics of tourism resilience from the perspectives of adaptive cycle theory, sustainable development theory, vulnerability theory and stakeholder theory. In terms of the influencing factors and the mechanisms of tourism resilience, scholars have explored many factors that could influence the level of tourism resilience. For instance, Fang et al.have found that several factors such as specialization and diversification can affect the level of tourism resilience. In other words, the level of tourism resilience can be promoted by measures such as readjusting the industrial structure and increasing the diversity of tourism products (Fang et al., 2023).

In terms of evaluating the tourism resilience level, scholars have mainly established index systems to assess the tourism resilience of a study region after the selection of the influencing factors. Different methods have been used to determine the weight of each index, such as the expert scoring method (Shwedeh et al., 2022) and the entropy evaluation method (Yang et al., 2022; Wang et al., 2023a). In terms of the promotion strategy of tourism resilience, scholars have mainly carried out practical and comparative studies on tourism destinations, tourism economy and tourism enterprises, and proposed the upgrading strategies from the aspects of development ideas and promotion countermeasures. For example, Tang and Guo (2018) proposed that improving tourism resilience requires a reasonable governance mechanism, which means that a region needs to improve its resilience levels in other fields first, such as the economy, society, ecology and organization, before it can improve its tourism resilience. In another study, Wei et al. (2022) discussed the measures to improve tourism resilience from three dimensions: technology guidance, demand orientation and policy guarantee. In terms of research methods, both quantitative and qualitative methods have been used to analyze tourism resilience. From the perspective of qualitative research methods, scholars mostly use questionnaire surveys and case analysis to analyze the tourists’ perceptions of the epidemic and willingness to travel (Madeira et al., 2021). From the perspective of quantitative research methods, scholars have mainly used the expert scoring method, entropy method, kernel density method and DEA-Tobit model to measure and analyze the factors influencing tourism resilience (Huang and Wang, 2022).

In summary, the existing studies on tourism resilience mainly focus on the macro level, such as regional tourism resilience and tourism industry resilience, but there is a lack of studies on the development resilience of tourist attractions at the micro level. The regional tourism resilience constructed by the existing research mainly includes indicators from the aspects of regional economic development level, population size, tourism resource richness, tourism facilities, transportation infrastructure, tourism employment population, tourism economic reconstruction, talent reserve, and science and technology level. In the specific evaluations, the statistical data such as regional per capita GDP, total population, number of A-level tourist attractions, number of tourism facilities, railway and highway mileage, number of employees in tourism enterprises, proportion of fixed asset investment, number of colleges and universities, and number of patent authorizations have been selected as the operational evaluation indicators. As a micro-component of regional tourism, the development resilience of tourist attractions is restricted by the overall resilience level of regional tourism, but it is also affected by the types of tourist attractions, business models and other factors. Therefore, the applicability of resilience research methods and evaluation models that have been applied to large-scale research areas for studying the development resilience of tourist attractions is questionable. However, there are currently few relevant studies on tourist attractions, and a resilience evaluation model that is suited for tourist attractions has not been constructed.

Considering the analysis above and the importance of tourist attractions in the development of China’s tourism industry, this study takes tourist attractions as the research object, systematically analyzes the influencing factors and paths of major public health events on the development resilience of tourist attractions, and constructs an evaluation index system in order to provide a scientific reference for improving the development resilience of tourist attractions and to help the tourist attractions achieve sustainable development.

On January 8, 2006, the State Council of China issued the “National Public Emergency Overall Contingency Plan”, which divided public emergencies into four categories: natural disasters, accident disasters, public health events and social security events. Since 2000, China has experienced several major public health emergencies, such as SARS and COVID-19, which have had a great negative impact on the development of tourist attractions. The impacts of major public health events on the development of tourist attractions are multifaceted and their characters vary at different stages. Tourist attractions will show resilience by coping with these different impacts at the different stages. In order to assess the level of the resilience described above, this study first systematically combed the impacts of major public health events on tourist attractions and analyzed the relationships between the development resilience of tourist attractions and these impacts. This analysis provided a theoretical basis for the scientific construction of an evaluation system for the development resilience of tourist attractions under the background of major public health events.

From the perspective of impact dimensions, the impacts of major public infectious disease events on tourist attractions are mainly reflected in two aspects: business activities of tourist attractions and travel preferences of tourists (Fig. 1). In terms of affecting tourism business activities, the impacts of major public health events such as COVID-19 on the business activities of tourist attractions are mainly reflected in cost pressure. For instance, during the duration of the COVID-19 event, due to the sharp decline in income, tourist attractions experienced increasing pressures on many aspects such as salary payments, loan repayments, advance investments, order refunds and project construction; meanwhile, the costs of prevention and control of the pandemic, the business development during the pandemic, the uncertainty of employees’ arrival and the lag of accounts also contributed to the increase in costs.

In terms of influencing tourists’ travel preferences, the epidemic had impacts on the development of tourist attractions by changing the tourists’ psychology, travel preferences, consumption habits, and other aspects. Firstly, because of the ineffectiveness and unknown nature of the epidemic, tourists’ perception of the tourism destination image may change from safe to unsafe, and the tourists’ self-protection awareness and risk perception in the face of potential physical health hazards will be enhanced. These changes would lead to a change in the tourists’ travel preferences (Lee and Kim, 2023). Secondly, tourists’ panic, fear and other emotions caused by the pandemic may manifest as various concrete behaviors, such as staying indoors, going out less, shortening travel distances and reducing face-to- face contact with others. Along with these behaviors, tourists’ consumption confidence and willingness will also decline. Additional factors such as the differences in pandemic prevention policies between tourist sources and destinations, the probability of infection during tourism, the sanitary environments of tourist destinations, and the intensity of personnel can all affect tourists’ travel and consumption choices (Li and Huang, 2023). After the outbreak of the pandemic, the changes in tourists’ travel preferences, such as shorter travel distances and higher demand for hygienic safety, make health tourism, inner-provincial tourism, rural tourism, self-driving tourism and short-distance tourism become more popular, while indoor tourism and long-distance tourism become less popular. As a result, under the influences of changes in the tourists’ travel preferences, the speed and degree of recovery from the pandemic varies among tourist attractions due to their differences in resource types, traffic locations, ways of visitation, and hygienic security conditions.

From the perspective of the impact stage, the impact of major public health events involving infectious diseases on the development of tourist attractions can be divided into three stages: Pre-disaster (before the disaster), in-disaster (during the disaster) and post-disaster (after the disaster). The direct impact of infectious diseases is mainly manifested in the pre-and in-disaster stages, but these effects will continue until the end of the disaster, and the disaster may re-emerge and have a further impact on the development of tourist attractions. Therefore, only by effectively preventing the possible risks of the disaster in the pre-disaster stage, by resisting and adapting to the impacts of the disaster in the in-disaster stage, and by updating the tourism products and service facilities of the tourist attractions according to the changes in tourists’ travel preferences in the post-disaster stage, can tourist attractions enhance their own development resilience and ability to cope with the risks.

Based on the definition of the development resilience of tourist attractions and the analysis of the impacts of major public health events on the development of tourist attractions, this study proposes four abilities (prevention ability, resistance ability, recovery ability and renewal ability) that comprehensively characterize the development resilience of tourist attractions, and constructs the relationship between it the impact of major public health events (Fig. 2). Specifically, in the pre-disaster stage, the impact of the pandemic is manifested as risk, which tests the risk prevention ability of tourist attractions. In the in-disaster stage, the pandemic risk is converted into the actual impacts mainly on tourist attraction management and tourists’ travel preferences. The abilities of resistance, recovery and renewal are used to reflect the strength of tourist attractions in dealing with these effects. Specifically, the strength of tourist attractions’ resistance ability determines the degree of loss caused by the negative impacts of the pandemic; and the strength of the recovery and renewal abilities of tourist attractions determines whether the tourist attractions can quickly recover from the negative impacts of the pandemic and adapt to the changes in the tourists’ travel preferences and needs. In the post-disaster stage, the impacts of the pandemic are mainly shown as the sequelae, that is, tourist attractions recover slowly and tourists have psychological shadows that persist after a severe pandemic shock. Therefore, tourist attractions need to quickly resume operations and update tourism products and services according to the changes in market demands to eliminate the remaining negative effects of the pandemic, thereby promoting their recovery and sustainable development.

Prevention ability refers to the ability of tourist attractions to prevent sudden major public health events, among which risk test ability, risk control ability and emergency reserve level are the key factors affecting the level of prevention ability (Wang et al., 2023b). Firstly, conducting risk prevention tests regularly is an important measure to explore the systemic risk status of tourist attractions in the face of major public health events and improve their risk identification. This task can build an efficient safety risk prevention system for tourist attractions and improve the coping ability and flexibility of the tourist attractions. Secondly, the risk management and control plan is a concentrated reflection of the emergency performance during disaster events by the government, enterprises and other entities, including the release timing of early warning information, the completeness of the plan, and the implementation results of the plan. This plan is an important component for effectively integrating resources and improving risk prevention capabilities. Thirdly, due to the unpredictability of major public health events, the tourist attraction should undertake emergency reserve planning to cope with the losses and maintain the necessary expenditures when the risk comes.

Resilience refers to the ability of tourist attractions to cope with external shocks and interference, which is mainly affected by emergency response ability, the quality of the tourist attraction, the scale of the local tourist market, the consumption level of the local market, the control level of operating costs, ownership of management subjects and diversification of revenue (Huang, 2007; Liu et al., 2017; Dong et al., 2018; Zhang et al., 2022). Firstly, the emergency management of public health events is a long-term and complex process, in which rapid emergency response capability is the key to disaster mitigation and an important manifestation of the resilience level of tourist attractions. Secondly, the level of tourism resources, the level of service facilities construction, the quality of tourism products and the level of health and safety protection are important factors affecting the choices of tourists during the duration of public health emergencies. The levels of tourist attractions can reflect the levels of these factors to a large extent, so they can be used as indicators of the resilience of the tourist attractions. Thirdly, during the epidemic period, the radius of tourists’ travel activities is greatly reduced, and tourists are more inclined to peri-outings, and surrounding area and intra-provincial tours. Therefore, under the influence of an epidemic situation, the status of the local tourist market in the overall market of the tourist attraction has increased, becoming one of the main driving forces for the development of the tourist market of the tourist attraction. Because residents are more inclined to travel a short distance under the influence of the pandemic, a strong local spending power can promote the faster recovery of tourism income in tourist attractions and provide a guarantee allowing the tourist attractions to resist the interference of the epidemic. Fourth, because the overall operating income of the tourist attraction shows a rapid downward trend during the pandemic, effective cost control becomes an important factor to enhance the tourist attraction’s resistance to the impact of the pandemic and enhance its development resilience (Marko et al., 2020). Fifth, due to the differences in emphasis on social and economic values between state-owned and private tourist attractions, the tourist attractions of different management entities are also affected differently. Under the interference from major public health events, the loss of small-scale and non-public tourist attractions is generally higher than that of public-owned tourist attractions due to the shortage of large loans, fewer subsidy channels and greater influence of market loss in the former. In addition, in terms of the profit model, the current tourism industry generally has problems of insufficient innovation and a single business product. During the pandemic, the main business income of most tourist attractions declined significantly, while tourist attractions with an online service ability, online sales systems, and relatively diversified services and industries can make full use of their own channel advantages to fully develop e-commerce, cultural and creative businesses, which can make up for the loss of income from the main business. Therefore, diversified profit models are an important way for tourist attractions to reduce their losses under the influence of the epidemic.

Recovery ability refers to the ability of tourist attractions to return to the original balanced system state after being disturbed by the outside world. The durability of tourism facilities, the degree of openness of tours, the accessibility from external traffic, the level of management decision-making, and the flexibility of marketing are important factors that ensure the recovery ability of tourist attractions (Ruan and Li, 2018; Yang et al., 2019). First, due to the lack of maintenance during the pandemic, the quality of various tourism facilities will be affected to varying degrees. After the tourist attraction is reopened, the tourism facilities with good durability can be quickly put into use to help the tourist attraction recover quickly. Secondly, due to the great changes in the consumption characteristics and needs of tourists after the pandemic, outdoor tourism products are more popular than indoor tourism products, and gradually become the mainstream direction of tourism development after the pandemic. Therefore, tourist attractions with more outdoor tourism products are more in line with the tourism consumption needs of the public under the influence of the pandemic, so they show stronger development resilience. At the same time, as tourists are more inclined to drive to the suburbs, tourist attractions with good traffic accessibility and a short driving time from the urban area are more favored by tourists than tourist attractions with poor traffic accessibility. So better accessibility is more conducive to the recovery and development of tourist attractions during the pandemic. In addition, in terms of tourist attraction management, a scientific and efficient management decision-making level is the basic guarantee for enabling the tourist attraction to reorganize the internal structure of the team and adjust the functional attributes under the influence of the pandemic. This management flexibility can effectively alleviate the pressure brought by the pandemic, optimize the state and structure of the tourist attraction, and promote the sustainable and healthy development of the tourist attraction after the pandemic (Li, 2021). Furthermore, to stimulate market demand and enhance the attractiveness of tourist attractions, steps such as adjusting the supply of tourism products, upgrading marketing strategies and other methods can also become important ways for tourist attractions to actively respond to the impact of the epidemic and restore development.

Renewal ability refers to the ability of tourist attractions to reorganize their internal and external structures and functions, and to create new development paths and new models in order to adapt to the new environment in response to external shocks. Tourism revenue reconstruction ability, financing ability, talent reserve level, intelligent construction level and mechanism innovation are the main factors influencing the renewal ability of tourist attractions (Tian, 2016; Yu, 2021; Yang et al., 2022). Firstly, tourism revenue is an important manifestation of the tourism economy, and the reconstruction force reflects the degree of repositioning and adaptation of the tourism economy in response to shocks. A higher level of tourism revenue for reconstruction can not only enhance the ability of tourist attractions to resist risks, but it can also allow them to use multiple channels and multiple methods to provide support for the tourist attractions, weaken the damage caused by shocks, and improve market service capabilities. Secondly, capital is the core element of tourism production, operation and consumption, so its scale, structure and layout have a profound impact on the quality, efficiency and direction of tourism development. Improving the financing capacity of tourist attractions will help to promote the sustainable development of tourist attractions after the pandemic, in addition to optimizing the financing structure and enhancing the competitiveness of the tourist attractions. Thirdly, in response to the market changes after the pandemic, the high-quality talent reserve structure can promote the optimization of resource allocation and the improvement of the management and operational level of tourist attractions. This situation can provide favorable support and guarantee for the post-pandemic renewal and development of tourist attractions (Tian and Zhao, 2022). Fourth, under the influence of major public health events, strengthening the application of new technologies such as cloud computing, big data and VR in the daily operation of tourist attractions can effectively promote their intelligent transformation and development and optimize their internal and external development environment. In addition, the exploration of new resilient development strategies for tourist attractions in the post-pandemic period depends on the continuous innovation of development ideas and development models, especially mechanism innovation (Ahmad et al., 2023). Mechanism innovation can stimulate the enthusiasm of relevant bodies, such as tourism resources, enterprises and financial capital, by establishing a sound scientific and effective operation mode, optimizing the combination of various components of the tourist attraction and various other production and operation factors, so that all stakeholders have the same goal, thereby forming a high degree of synergy, and maximizing the economic, social and ecological benefits.

The DEMATEL is a decision-making and evaluation experimental method, which can be used to study the logical relationships between the elements in a complex system to simplify the system analysis (Sun et al., 2022), and lay the foundation for proposing more scientific and reasonable strategies and suggestions. Compared with the traditional AHP method, which ignores the interactions between indicators when determining weights, the DEMATEL method considers the relationships between indicators and calculates the importance of each indicator on this basis. There is a certain correlation between the dimensions and the influencing factors in the evaluation index system for the development resilience of tourist attractions constructed in this study. Therefore, we chose to use the DEMATEL method to determine the weights of the evaluation index for the development resilience of tourist attractions. The process involved six specific steps.

(1) Determination of the set of influencing factors (index). The set is:

where S is the set of factors which influence the level of development resilience of tourist attractions; s_i is influencing factor i; and i is the sequential number of the influencing factor.

(2) Establishment of direct influence matrix X. The matrix X is made up of the values of the influences between s_i and s_j. Those values were determined by relevant experts who were invited to compare and score the interaction relationships between each pair of two influencing factors according to their knowledge and experience. This study used the “0-3 scale” scoring method, in which 0 points means s_i has no impact on s_j; 1 point means the impact of s_i on s_j is at a low level; 2 points means the impact is at a medium level; and 3 points means the impact is at a high level. The matrix is:

where X is the direct influence matrix; and x_ij is the average value of the scores that the experts used to evaluate the influence of factor s_i on factor s_j . Since the index has no effect on itself, the value of x_ij is zero when i=j.

(3) Normalization of the direct influence matrix. In this step, matrix X is normalized into a normalization matrix G by the following formula:

where G is the normalization version of matrix X; g_ij is the normalized value of x_ij; and i and j are the row number and column number of the matrix, respectively.

(4) Calculation of the comprehensive influence matrix T. On the basis of the normalization matrix G, the matrix T consists of all the relations, including the direct and indirect influences between each factor and the others. The matrix T is calculated as:

where T is the comprehensive influence matrix; t_ij is the sum of the initial direct influences and all the indirect influences which are made by factor s_i on factor s_j; and I is the unit matrix.

(5) Calculation of the influencing degree f, influenced degree e, centrality degree m and severity degree n of each factor. The influencing degree f is the sum of the rows in matrix T. The formula is:

where f_i represents the degree of the comprehensive influence of factor s_i on the other factors.

The influenced degree e is the sum of the columns in matrix T. The formula is:

where e_j represents the degree of comprehensive influence from all the other factors on factor s_i.

The centrality degree m indicates the degree of relationships between each factor and the others, so m is the sum of f and e. A factor with a greater value of m means it has a stronger relation with the other factors, while small values of m mean weaker relationships. The formula is:

where m_i represents the centrality degree of factor s_i.

The severity degree r indicates the severity of the influence of each factor, so r is the difference between f and e. A factor with a greater value of r means it has greater influence on the other factors and is assumed to have higher priority, while a factor with a smaller value of r means it receives more influence from the other factors and is assumed to have lower priority. The formula is:

where r_i represents the severity degree of factor s_i. When r_i is greater than 0, the index has an influence on the other indexes, and is a cause index. On the contrast, when r_i is less than 0, the index is affected by the other indexes and is a result index.

(6) Determination of the weight of each index. The normalized method was used to calculate the weight of each index. The formula is:

where w_i is the weight of factor s_i.

According to the evaluation index system of tourist attraction resilience development that is constructed in Table 1, nine experts and scholars engaged in tourism management, tourism economy, tourism geography, urban and rural planning and other related fields were invited to score the interaction strengths between the factors influencing tourist attraction development resilience, and the direct influence matrix was constructed. Then, according to the DEMATEL method, the influencing degree, the influenced degree, the centrality degree and the cause degree of each influencing factor were calculated, and the weight of each evaluation index was calculated according to the centrality (Table 2).

In general, among the four dimensions for the evaluation of tourist attraction development resilience, the dimension with the greatest weight is resilience ability, followed by renewal ability, then recovery ability, and the lowest weight is for prevention ability. This result shows that resilience ability is the most important embodiment of the resilience level of the tourist attractions. Among the 20 evaluation indicators, the index weight of tourism revenue reconstruction power is the largest, at 0.062; the weights of 10 indicators such as mechanism innovation strength are greater than 0.05; the weights of eight indicators such as local tourist source market size are between 0.04 and 0.05; and the index weight of tourism facility durability is the smallest, at 0.037. The analysis of the results for each dimension are as follows.

(1) Analysis of prevention ability. In terms of centrality, the index that has the greatest impact on the prevention ability is the risk control ability, while the influences of other indicators are relatively small. Furthermore, risk control ability is also in the top position in the whole index system, which indicates that risk control ability is in the core position in the index system and is the core indicator affecting the pandemic prevention ability of tourist attractions. In terms of the degree of causation, the three indicators that affect the prevention ability are all result indicators, indicating that these three indicators are more susceptible to the other indicators. Among them, risk control ability is the most strongly affected by other indicators, so it will be strongly restricted and driven. Therefore, the tourist attractions can improve their risk control ability by improving the other aspects of their ability, which will enhance their pandemic prevention ability.

(2) Analysis of resistance ability. In terms of centrality, the revenue diversification degree and operating cost control level have the highest centrality values among all the indicators of resilience. This indicates that revenue and cost are the two key indicators for tourist attractions to enhance their resilience, as well as the main driving forces and guarantees for improving the resilience of tourist attractions. In terms of the degree of causation, the quality of tourist attractions, the scale of local tourist source market, the consumption level of local market, and the ownership of management subjects are positive, so they are the cause indicators. Although these four reason indicators affect the other indicators to some extent, the quality of tourist attractions and ownership of management subjects have only small probabilities of changing in a certain period of time; while local tourist source market scale and local market consumption level are determined by the regional population distribution pattern and socio-economic development level, and non-scenic areas can be changed. The emergency response capability, revenue diversification degree and operating cost control level have negative cause degrees and are easily affected by other indicators, so they are the result indicators. The two indicators of revenue diversification degree and operating cost control level have high centrality but are affected by the other indicators. Therefore, it is necessary to accurately analyze the other indicators that affect these two and make corresponding improvements. This will effectively improve the revenue diversification degree and operating cost control level of the tourist attractions, helping them to achieving the purpose of enhancing their resilience.

(3) Analysis of recovery ability. In terms of centrality, marketing flexibility and management decision-making level have the highest centrality, so they have the greatest impact on the resilience of tourist attractions, indicating that these two indicators are the core indicators. The second is the degree of tour openness, which indicates that the resilience of tourist attractions is greatly affected by the degree of tour openness. In terms of cause degree, external traffic accessibility has the highest positive cause degree, so it is the cause index with the greatest impact on the other indicators, indicating that traffic conditions are the basic guarantee for the rapid recovery of tourist attractions after the outbreak. The causes of marketing flexibility, tourism facility durability and management decision level have negative values and are outcome indicators, which indicates that although market flexibility and management decision-making level are the most important indicators of tourist attraction resilience, they are also easily affected by the other indicators.

(4) Analysis of renewal ability. In terms of centrality, the five indicators of renewal ability are at a high level among all 20 indicators. Among those five, the centrality of tourism revenue reconstruction power and mechanism innovation strength are the highest, indicating that these two indicators have the greatest impact on the renewal ability of tourist attractions, and are the key indicators in this dimension. The second highest are financing ability and smart construction level, indicating that the renewal ability of tourist attractions has high requirements for capital status and a smart construction level. From the perspective of cause degree, the mechanism innovation strength has the greatest influence on the other indicators, so it directly affects the changes in the other indicators, and is the most important cause indicator in the renewal ability. In addition, financing capacity and smart construction level are the two indicators that are relatively susceptible, so they are the links that need special attention to improve the renewal ability of tourist attractions.