The traditional statistical data of the catering industry have some shortcomings, such as incomplete classification information and time lag. With the rapid development of internet technology, big data represented by mobile phone signaling data, GPS trajectory data, and POI (point of interest) data have been constantly enriched and improved. As a new spatial data source, POI data have the advantages of a large amount of data, wide coverage, and high recognition accuracy and is easy to obtain (Zhang et al., 2021b).

Figure 1 shows our data acquisition and processing method. We divide the target city into different regions to obtain the street locations of Japanese restaurants across different regions. Next, we aggregate these street data into the city data set and establish a database of Japanese restaurants in the city. We use dianping.com and map.baidu.com as our sources for Japanese restaurant data.

We use the key search term “Japanese cuisine” in dianping.com and map.baidu.com; meanwhile, our collection rule is “whether customers choose to eat Japanese cuisine” and the collection objects are from 332 cities in China. The data set was collected from 28 May to 30 May 2022. Then, the data set is checked and rechecked for data loss. Map matching and database establishment are performed via ArcGIS and Google Earth, respectively, (Qin et al., 2019; Tian et al., 2021).

In the process of data acquisition and cleaning, we obtain the text data of Japanese restaurants from dianping.com (including restaurants that provide sushi and other dishes). After eliminating duplicate data and data that did not meet the search requirements, a total of 82,687 restaurant text data were included as part of this study. We found that the obtained restaurant data could be queried through batch comparison with the restaurant location data provided by map.baidu.com. The data used met our subsequent research needs.

In this study, the influencing factors are divided into economic factors, spatial factors, information factors, and humanistic factors (Yang et al., 2019; Guo et al., 2020b; Tian et al., 2021).

Economic factors: For the service industry, the pursuit of profit maximization is the factor preferred by producers for determining production location; that is, maximizing the difference between income and cost. Sufficient market demand is the first consideration in the spatial choice of the catering industry. Cities with large population size and high population density are the first choice for restaurant locations. Under sufficient market demand, the purchasing power level of the region is also an important factor in its location choice. The higher the disposable income of people in the region, the stronger the purchasing power, and the greater the possibility of the location will be. As an important part of the expansion cost of service enterprises, land rent is also one of the necessary conditions for location.

Spatial factors: Traffic accessibility is an important factor affecting the layout of the catering industry (Yang et al., 2018; Zhang et al., 2021a; Zhou et al., 2021). Good traffic accessibility is an important channel to communicate with consumers and retail facilities, and only areas with good accessibility can ensure that consumers can easily reach the retail business center. Simultaneously, smooth logistics distribution can save transportation costs for retail businesses, and good traffic accessibility can ensure the distribution of raw materials (Zhang et al., 2021b).

Information factors: Consumers often consider the place of origin of images and popularity of exotic foods, personal consumption concepts, eating habits, food quality, and impressions of a restaurant’s decor and then reach a purchase intention and choice preferences. Of these criteria, consumer knowledge of exotic food mainly comes through public media such as Japanese variety shows, animation, film, and television (Thorogood, 2020).

Humanistic factors: War can expose people to exotic food and deepen the impression and association of exotic cultures (Wallin and Sandlin, 2020). To understand the influence of the War of Aggression against China on the spread and development of Japanese cuisine in China, the cities affected by the war were assigned a value of 1 and the cities not affected by the war were assigned a value of 0. Moreover, consumers can directly contact the origin of exotic food through travel, which also increases familiarity with specific exotic foods.

The socioeconomic factors screened in this study are shown in Table 1. Considering the integrity of urban, economic, and social data, we used data from the Statistical Yearbook of 2021 for analysis.

According to economics, consumer demand is the quantity of goods or services that consumers are willing and able to buy at a certain price level. There must be an effective demand with both purchasing desire and purchasing power. The main factors include the price of goods, consumer income levels, consumer preferences, and consumer expectations of the price of the goods. Studies conducted in different environments have indicated that price is one of the main barriers to consumer access to dietary diversity benefits (Herforth and Ahmed, 2015; Vittersø and Tangeland, 2015).

By analyzing the average price of Japanese restaurants, we found that 17% of restaurants have an average price lower than 50¥, 63% have an average price of 50–211¥, and 20% have an average price higher than 211¥. Figure 2 shows that the pricing characteristic of Japanese cuisine is spindle-type, which signals that the main customers of Japanese cuisine are middle-income people, including the middle class and the “new middle class.” Therefore, wealthier areas may have more opportunities to enjoy more diverse and higher quality foods, while small and medium-sized cities and their residents are at a disadvantage in this sense. The eastern cities of China, which are economically dominant, have more potential consumers and greater consumer demand. Therefore, these cities are more likely to have a spatial agglomeration of restaurants and dishes. This finding is in line with general knowledge that big cities often have greater quality of life and wellbeing.

The distribution of restaurant locations and the field intensity of Japanese restaurants is basically bounded by the Hu line (The Hu line is a line that divides China’s population by density; it was proposed by Chinese geographer Hu Huanyong in 1935. According to the analysis of statistical data in 2000, the National Conditions group at the Chinese Academy of Sciences indicated that the southeast side of the Hu line accounts for 43.18% of China’s total land area, with 93.77% of the population and 95.7% of the GDP, an area with an overwhelmingly high-density economic and social function), but its distribution characteristics are more significant than population density.

In addition, in cities east of the Hu line, the total number of restaurants is 80,212, accounting for 97% of the national total, whereas the total number of restaurants in western cities is 2,475, accounting for only 3% of the national total (Table 2). If NR represents the number of Japanese restaurants, 168 cities have 0 ≤NR <100 restaurants, accounting for 51% of all cities (Table 2), and these cities are mainly distributed in central and western regions; 164 cities with 100 ≤NR <4,963 restaurants, accounting for 49% of all cities, are mainly distributed in eastern coastal areas (Figure 3A).

Figure 3B shows that there are 309 cities with dish styles (DS) between 4 ≤DS ≤7, accounting for 93% of the total number of cities. There are only 23 cities with 0 ≤DS <4 styles, but 82 cities have seven kinds of dishes. In terms of dish styles, there is a significant two-level differentiation in the service characteristics of Japanese cuisine.

Figures 3C,D show the spatial imbalance of the service capacity of Japanese cuisine between cities. In cities east of the Hu line, the field intensity value is basically higher than 0.08, and the service ability is strong; in western cities, the field intensity (FI) value is basically lower than 0.08, and the service ability is weak. Overall, there are 61cities with 0 ≤FI ≤0.08, accounting for 18.4% of all cities, and these are concentrated in central and western regions; there are 271 cities with 0.09 <FI <41, accounting for 81.6% of all cities, and these are mainly distributed in the East.

In the socioeconomic environment, the characteristics or identity of the media often determine the characteristics of catering services. The impact of non-equalization of the external environment, such as social economy, on producers and consumers became the main criteria for screening socio-economic environment factors in this study.

GeoDa is a free and open-source software tool that serves as an introduction to spatial data science; thus, we used GeoDa to solve the model. Before the establishment of the spatial econometric model, the spatial correlation of the sample data was tested. Table 3 shows that Moran’s I are 10.818, 5.768, and 5.989, and their significance levels are all lower than 0.001, indicating that the spatial distribution of restaurant locations, the formation of dish styles, and the consumption intensity of Japanese restaurants are not completely random, but have a certain spatial correlation. Therefore, the OLS regression results may be biased. It is necessary to use the spatial econometric model to for estimation.

From the perspective of variable coefficient and significance (Table 3), the results of the model calculation are consistently high, and the R² value increases after considering the spatial correlation (SLM = 0.92, 0.72, and 0.89 and SEM = 0.94, 0.72, 0.89, higher than the R² value of OLS). The SLM and SEM interpretation results are better than the OLS regression results. Combining the LM and R-LM values and their significance and comparing the fitting effects of SLM and SEM, the significance of SLM regression results and test results is better than SEM model results. Thus, this study selects SLM results for subsequent analysis.

Table 3 shows that urbanization level, population size, and economic scale are the key factors that lead to spatial distribution inequality and service imbalance of Japanese cuisine in China. Cities with a large economic scale can often gather more resources. This agglomeration enables economic activities to play a scale effect, which makes the price of products in big cities not only lower but also increases the variety available for selection (Handbury and Weinstein, 2015). Further research shows that in China, large cities have more products and lower prices (Feenstra et al., 2017). Therefore, larger economic-scale cities can gather more consumer demand and market of Japanese cuisine, and the scale of development of Japanese restaurants in these first-tier cities is significantly higher than in other cities.

Population size affects the quantity of food supply and the diversity of dishes. The main reason for this effect is that the size of a population increases the demand for food diversity at the consumption level, which leads to aggregation of various kinds of foods (including exotic foods) and finally results in greater diversity in a more densely populated city, giving the city stronger service ability and higher radiation field intensity. The larger the population, the more Japanese restaurants there are. For example, the top 25 Chinese cities for the number of restaurants accounts for 23% of the total population, but the number of restaurants accounts for 49% of the country (Figure 4). The number of restaurants in cities is basically consistent with the size of the urban population. In addition, the huge population size also entails a complex population structure and an ease in creating a complex heterogeneous demand set in a city, which produces high demand for food types in certain areas. The large coefficient also implies that there may be such a mechanism. The diversity of the population leads to integration of different cultures, making people more willing to try exotic foods (Glaeser et al., 2001).

Simultaneously, the degrees of effect of the elements in model 1 are the following: LNTV> LNTIP> LNUR> LNPOP> LNLR> LNPGDP> LNWAR. The degrees of the functions of the elements in model 2 are the following: LNUR> LNPOP> LNPGDP> LNFL. The degree of the functions of the elements in model 3 are the following: LNTV> LNUR> LNTIP> LNPOP> LNLR> LNPGDP> LNRND.

Consumer love for variety is a common assumption in modern economics (Glaeser et al., 2001). Due to the heterogeneity of cities and the non-equalization of service facilities, the spatial distribution of food and drink is often unbalanced and complex. Considering the necessity to promote food equality in different socioeconomic areas and the offsetting of the negative impact on the health of vulnerable groups, it is necessary to interpret the spatial equality of dietary distribution (Turner et al., 2018, Turner et al., 2020).

The global commodity network brings a systematic and interconnected consumption landscape. Considering that the socioeconomic environment of consumers is not consistent under different urbanization levels, the place effect should be considered when describing the characteristics of urban catering services; that is, different spatial locations of food, different degrees of realization of the same food, and different benefits. It is more comprehensive and objective to identify the service differences of Japanese cuisine in different regions and spaces with the help of a field intensity model.

This study also encourages and advocates the use of multiple geographic data, the use of data provided by other GeoWeb websites, and other Web2.0 online websites, actively exploring the connection and interaction between geosciences and socio-economic phenomena. In the era of mobile internet, the internet catering service platform has become an important channel for the majority of catering business operators to obtain business data and consumer evaluation information. This would largely change the impact of the competitive rent theory on the location choice of restaurants. These online catering websites can provide more positive and beneficial opportunities for researchers to pay attention to the urban food environment. Simultaneously, it is necessary to explore the application of big data in the field of geographical sustainability. The rise of big data provides new methods, new ideas, and new solutions to pressing urban issues (Xue et al., 2020; Liu et al., 2021; Yang et al., 2021). In contrast to questionnaire and interview approaches, we collected third-party big data provided by an online catering platform to analyze the spatial differences in the number of restaurants and dish styles of exotic food. The multifusion data and big data in this study also provided references for solutions to other urban problems (Shelestov et al., 2017; Haworth et al., 2018; Shirazi et al., 2021; Zhang et al., 2020).

Our results show that there are more Japanese restaurants and dish styles in Eastern China. This conclusion is consistent with the main findings of Tian et al. (2021) on the spatial distribution pattern of Chinese seafood restaurants. They found that seafood restaurants are concentrated in coastal cities. Our results also support the significant correlation between population size, economic development, and the distribution of different types of restaurants. However, this study does not confirm that the geographical distance proposed by Tian et al. (2022) has a significant impact on the distribution of restaurants. We believe that this may be related to consumers’ eating habits and the pricing of Japanese restaurants. In addition to the socioeconomic factors in the socioeconomic environment, food similarity and cultural similarity may also have a potential impact on the spatial inequality of dietary distribution (Figure 5). Rice is a staple of Japanese cuisine and so is very easily accepted by Chinese consumers. In addition, Japanese cuisine consists mainly of seafood, and the acceptability of seafood in China’s coastal regions is very high. It is not difficult to understand that Eastern coastal cities become hot spots and regions where Japanese restaurants are found in higher concentrations. In inland cities, local consumers are not accustomed to eating seafood, and so, the acceptability of this cuisine is relatively low, making these cities cold spots of distribution. This also demonstrates the objectivity of our findings.

A limitation of this study is that the topographic division of urban space, such as by landscape patterns, hinders the spread and diffusion of Japanese cuisine in China, and the research results may contain some errors as a result. In addition, the distance friction coefficient in the field intensity model is usually between 1.0 and 3.0 (the distance attenuation curve varies with the coefficient). The difference in parameter values may lead to deviation from the actual situation, which, in turn, may lead to less accurate calculation.

By describing the place effect of Japanese cuisine on different spaces in China, we found that the number of restaurants, dish styles, and field intensity shows a gradual decline from eastern cities to western cities. Compared to the Hu line, which reflects the spatial difference in population, the number of restaurants in eastern cities is 32 times that in western cities.

The influx of Japanese cuisine has not alleviated inequality in the urban food environment. The pricing strategies of different grades of dishes show that Japanese cuisine mainly serves the middle class, which means that rich cities often enjoy higher quality food. For cities in a weak position, this undoubtedly damages the wellbeing and the diversity of comfort that food can provide.

Based on the analysis of the local socioeconomic factors, urbanization level, population size, and economic scale are significantly correlated with the maturity of Japanese cuisine. However, the degree of responsiveness of service ability to the socioeconomic environment changes with the location. This also implies that the positioning demand of Japanese cuisine and urban socioeconomic environment is intrinsically related to its price characteristics. The choice of restaurant location and the strength of service ability are the result of the interaction of external socioeconomic environment and internal price characteristics.

We also find that sushi (in 328 cities) and other culturally distinctive dishes have become the most accessible and most likely to be sampled. For certain dishes, the spatial imbalance may not be significant. In the future, it will be necessary to analyze consumer personal values and the relationship between their purchasing lines and specific food choices.