A GloVe-Based POI Type Embedding Model for Extracting and Identifying Urban Functional Regions

Abstract

Points-of-interest (POIs) are an important carriers of location text information in smart cities and have been widely used to extract and identify urban functional regions. However, it is difficult to model the relationship between POIs and urban functional types using existing methods due to insufficient POIs information mining. In this study, we propose a Global Vectors (GloVe)-based, POI type embedding model (GPTEM) to extract and identify urban functional regions at the scale of traffic analysis zones (TAZs) by integrating the co-occurrence information and spatial context of POIs. This method has three main steps. First, we utilize buffer zones centered on each POI to construct the urban functional corpus. Second, we use the constructed corpus and GPTEM to train POI type vectors. Third, we cluster the TAZs and annotate the urban functional types in clustered regions by calculating enrichment factors. The results are evaluated by comparing them against manual annotations and food takeout delivery data, showing that the overall identification accuracy of the proposed method (78.44%) is significantly higher than that of a baseline method based on word2vec. Our work can assist urban planners to efficiently evaluate the development of and changes in the functions of various urban regions.

Research on Green Space Service Space Based on Crowd Aggregation and Activity Characteristics under Big Data—Take Tacheng City as an Example

“…The POI data contain the location and attribute information of various hot spots related to human life, and their plane aggregation degree is proportional to the heat of life in the region, which can be used for urban functional zoning and identification [ 30 , 31 , 32 ]. In this study, the POI data source is the Bige Map GIS Office downloader Gaode electronic navigation map data( , accessed on 1 April 2022), using interest point layer data.…”

“…The population size in the region directly raises the demand for green space, and the fineness of residents’ enjoyment of green space mainly depends on the resolution of population data [ 33 ]. At present, there are a variety of public large-scale population grid data, such as the GPW dataset produced by the area weighting method [ 32 ], China’ s population spatial distribution kilometer grid dataset produced by the multi-source geographic information fusion method and the WorldPop world population grid dataset (WorldPop data). WorldPop data are often used as a control group to verify product accuracy, which has high applicability and high spatial resolution, as well as population fitting accuracy in China [ 34 , 35 , 36 ].…”

Gender Identification Using Marginalised Stacked Denoising Autoencoders on Twitter Data

“…Tweets below 5 words were discarded with the tweets which are duplicated. The Glove model is an alternative word embedding method [17]. This method makes use of an unsupervised learning technique to construct word embeddings.…”

Accurate Identification of Micro-scale Urban Functional Zones from the Perspective of Dialectical Materialism

“…Additionally, Yi et.al 47 chose to isolate the hospital category separately as a representation category of POI data related to healthcare. Some POI data may have various functions within the city, which means they can have two or more possible meanings 48,49 . Although, it is true that different sources of POI data may employ different classification criteria, which provides various choices and perspectives for related research, it becomes challenging to determine the appropriate category if the data lacks specificity regarding the dominant functionality of the POI data.…”