Identification And Analysis Of Urban Functional Areas In Tianjin Using Multi-Source POI Data

With the continuous development of China's economy and society and the urbanization level,the problem of urban space development has become increasingly prominent.The development model needs to be changed from "incremental expansion" to "stock mining" of high quality.Therefore,the study of stock space is particularly important,and the study of urban functional areas is an important part of the study of stock space.The identification of the distribution status of urban functional zones,the analysis of the composite land functions and the optimization of urban land use play an important role in the design of urban planning.However,traditional methods of urban functional area identification are limited by objective factors such as difficult data acquisition,high cost and poor timeliness.Research often stays at the macro level and lacks elaborate analysis.With the emergence of open data on the Internet,it is very convenient for us to obtain urban details which are difficult to express by traditional planning data.In this context,this paper takes Tianjin as an example to identify and analyze the functional areas of the central city and Binhai New District of Tianjin,and extract the optimization suggestions based on the POI data of Baidu Map and Amap acquired by web crawler,combined with the traditional planning and current data.In this paper,two sources of POI data are selected to ensure a more comprehensive and detailed coverage of the experimental research area.In terms of data processing,this paper explored a multi-source POI data fusion method combining spatial location matching and name similarity matching.This method can solve the problem of common errors in longitude and latitude of data from different sources when using spatial location matching alone,and the problem of high requirement for non-spatial attributes and possible information missing and annotation errors.On this basis,the accuracy of POI fusion method is improved by using name similarity,and the POI fusion data set for urban functional area recognition is obtained by using this method.Based on the fused multi-source POI data set and the current data of urban land,the identification method of Tianjin urban functional area is studied.We used road network data to divide land into blocks as the smallest unit of urban functional area identification.This forms a scientific and reasonable basis for functional area identification.After identification,the accuracy of the recognition results was verified from multiple angles and channels.On the basis of comparing the results with the current urban map and Baidu map images,the recognition accuracy is quantified by using the confusion matrix.The overall accuracy is 78.33%,and the Kappa coefficient is 0.74.This recognition result can make up for the deficiency of current land status data in refining the distribution of urban functional areas.At the same time,on the basis of recognition results,the spatial distribution pattern of urban function is analyzed from the overall and the central urban area point of view.And it is found that there is still room for development of commercial function within the city area.The mixed degree of urban land in inner districts of the city and the main urban area of Binhai New District is better by using the mixed degree analysis model.Further more,we got the location entropy index by using the method of functional composite analysis.Based on the geographic information spatial statistical tools,the paper makes a multi-angle and multi-circle analysis of the city's residential,commercial,public service and industrial functions.Finally,combined with the analysis results,the paper summarizes the problems existing in Tianjin's urban functional structure.