Research On The Measurement Of Catering Vitality Based On Dianping Data And Its Relationship With Urban Landscap

The vitality of the catering industry can effectively reflect the intensity of residents’ dining activities within a city.Multi-dimensional perception of catering vitality can measure the industry’s attractiveness and diversity,and then calculate the spatial pattern and consumption structure of the catering industry,providing scientific decision-making for site selection,planning,and layout of the catering industry.Catering venues are mostly concentrated in areas where residents are active and where urban functions are centralized,and are closely related to the built environment within the city.Urban landscape is a new means of quantifying the urban built environment.Therefore,exploring the relationship between urban landscape and catering vitality helps to reveal the impact mechanism of the urban environment on catering vitality and provide a theoretical basis for improving catering vitality.With the advent of the big data era,catering service platforms represented by Dianping,Ele.me,and Meituan are increasingly favored by urban residents.Data generated by these platforms,such as catering review data,Weibo check-in data,and mobile signaling data,have the characteristics of wide coverage,free sharing,and quasi-real-time,and can effectively depict the pattern preferences and intensity characteristics of residents’ dining activities,providing important references for refined measurement of catering vitality.Therefore,this article chooses Guangzhou(Yuexiu District,Tianhe District,Haizhu District,and Liwan District),a city with a strong catering culture,as the research area.Firstly,a multi-dimensional topic information mining method was designed to calculate the vitality of urban catering in each dimension according to the multi-dimensional attribute information of Dianping data.Then,referring to the landscape index in landscape ecology,the measurement index of urban landscape pattern was constructed from three levels of single element,combined element and overall landscape,and the characteristics of urban landscape were quantitatively measured.Finally,a detailed comparative analysis of the association between catering vitality and urban landscape is performed by constructing various regression models.The study found that:1.Dianping data can effectively characterize the catering vitality,and the depicted multi-dimensional catering vitality has similarities and differences in spatial distribution.The high-value areas of catering vitality mainly appear in the west of Tianhe District,the south of Yuexiu District,the east and west/north of Liwan District,and the basic trend shows a decrease from the city center towards the outskirts.The spatial distribution of restaurant vitality in each dimension is different,and the vitality distribution based on the number of restaurants is relatively dispersed,showing a multi-center distribution.The vitality distribution based on the number of reviews is more concentrated.The spatial distribution characteristics of vitality based on taste,environment,and service are more similar,and the concentrated distribution area of hotspots is mainly in the southwest of Tianhe District.The restaurant vitality in each dimension has spatial autocorrelation,and the vitality of taste,environment,and service are positively correlated.2.The spatial distribution of multi-level urban landscape has its own characteristics.Different landscape elements showed different distribution patterns influenced by the land use function and development status of each block.Various POI facility densities,average building heights,plot ratios,road network densities,and ratios of roads to other areas showed consistent rules with the distribution of commercial districts and squares inside the city,while vegetation and water landscape were affected by the building landscape and showed a complex and irregular shape within the built-up area.3.There is a strong relationship between urban landscape and catering vitality,and there are commonalities and characteristics in the mechanism of influencing factors of multi-level urban landscape on multi-dimensional catering vitality.The correlation and relative importance of multi-level urban landscape to multi-dimensional restaurant vitality were different,and the vitality based on the number of restaurants had more consistency with urban landscape.The POI facility level and combined element level had a positive correlation with restaurant vitality,while in the single landscape element level,2D/3D building indicators and road network density had a positive correlation,and water and vegetation landscape element indicators had a negative correlation.The main urban landscape indicators affecting the vitality of the number of restaurants,the vitality of the number of reviews,and the vitality of restaurant evaluation were the density of commercial facilities,public service facilities,POI density,and building plot ratio.The relative importance of indicators in the single landscape element level was relatively small.The vitality of the number of restaurants and the vitality of the number of reviews were more suitable for the random forest regression model,while the vitality of restaurant evaluation was more suitable for the spatial autoregressive model.