Competitive Location Of A Retail Agglomeration Using Social Media And Considering Spatial Effect

Recently,with the development of smart mobile devices and network technique,we can connect to the Internet anywhere at any time.In this background,the advent of the social media changed the conmmucation pattern between social media users.Now,the represented social media services are Twitter,Flicker,Sina Weibo and so on.Social media service has a large amount of users,and these users post information related to the real world in the social media platform by using words,video,and website links.When the the geographic information services were added to the social media,the spatial dimension were added to the multiple dimensions of social media data.The social media data mining attracts researchers’ attention and is applied in different fields(such as,geography,sociology,journalism and so on).The social media data mining produces a lot of research results and has important research applications.Social media data has been applied in the studies related to the competitive facility location.The aim of competitive facility location is the profit and market share maximization.In general,competitive facility location includes 3 steps:competitive location sampls collection,the evalution of the sensitivity parameters of influencing factors(such as,the distance between customer and the commercial facility,the commercial facility attraction),the calculation of a new commercial facilities’ market share.Sensitivity parameter refers to the sensitive degree of customers to influencing factors and the accuracy of parameter evaluation can be considered as the accuracy of competitive location.Among these steps,evaluating the sensitivity parameter of influencing factors is the core step.The accuracy of parameter evaluation can be treated as the accuracy of competitive location.Now,applying social media data in the competitive location with high accuracy faces some challenges:(1)social media data is disperse and can not reflect the complete activity of users.If we treat each social media user as the traditional respondent,the attribute of extracted sample will not be precise and the inaccuracy attribute can influence the accuracy of location results.(2)The spatial effects of samples can influence the the competitive location result.The attribute of samples extracted from social media data are related to each other and there is spatial effect between different samples.Owing to the spatial effect,the samples extracted by using different methods can produce different location results.(3)The sensitivity parameters are spatial heterogeneity.Most studies use the stationary sensititive parameters to locate a new commercial facility.However,because of the influence of environment,the sensitivity parameters in different areas are different in the real world.Aiming at these challenges above,this paper proposes a competitive location method which considers about spatial effect based on social media data.This method can use the social media data to locate the commercial facility with high accuracy.This study includes three aspects:(1)we study how to extract the competitive location samples from social media data by using spatial aggregation.This paper constructs spatial units to divide the research area and reflects the sample attributes of social media users to the spatial units.This method reveals how the spatial aggregation influences the results of competitive location and extract the competitive location samples effectively.(2)We study how the spatial effect influence the location result.This paper explores the influence of spatial effect from three aspects:spatial aggregation effect,spatial unit hierarchy effect and MAUP(Modifiable Areal Unit Problem)effect.Based on the analysis of spatial effect influence,we verify the effectiveness of the samples extraction method,select the suitable hierarchy,size and shape of spatial unit.(3)Based on the analysis of spatial effect influence,we proposed the competitive location method by considering sensititive parameter spatial heterogeneity.This method compares the spatial homogeneity results with spatial heterogeneity results,chose the suitable sensitive parameters,reveal how the parameter spatial heterogeneity influences the results of competitive location.By using this method,we can chose the best location of new commercial facility with high accuracy.To verify the effectiveness of the propsed method,this paper takes 5 retail agglomerations in Beijing as an example.The research area is the area surrounded by Sixth Ring Road and we collected the geo-ta.gged Sina Weibo data in Beijing.The experiment results indicate that the method proposed in this paper can extract samples from social media data,explore how the spatial effect influence the location results,and chose the best location for a new retail agglomeration with high accuracy.