Authors: Rui Zhu*, University of California - Santa Barbara
Topics: Geographic Information Science and Systems, Spatial Analysis & Modeling, Quantitative Methods
Keywords: Uncertainty, GIScience, Volunteered Geographic Information, Spatial Statistics, Semantics
Session Type: Paper
Start / End Time: 1:10 PM / 2:50 PM
Room: Coolidge, Marriott, Mezzanine Level
Presentation File: No File Uploaded
With the increasing popularity of volunteered geographic information (VGI), the quality of geospatial data currently attracts more attentions compared to the traditional issue of coverage in GIScience community. To understand the accuracy, or uncertainty, of geospatial data, many aspects have been applied, such as uncertainties in positions, attributes, and topological relations. However, there is few research on assessing the uncertainty of geospatial information from a semantic perspective, which plays an increasingly vital role in an era where place-based GIS emerges to corporate with space-based GIS. To address such as a gap, this work proposes an approach to take into account the semantic uncertainties of places to understand the quality of VGI. The semantics are specifically extracted from the feature type of geographic information (e.g., place types such as Restaurant, Hotel and Police Office) as it is semantically rich in terms of characterizing the functionality, popularity, and human perception of places. In order to assess semantic uncertainty of feature types, which are categorical compared to numeric values applied in most conventional uncertainty analysis, we design a set of spatial statistics, named as semantic signatures, to quantify the spatial characteristics of places. Having spatial signatures, we then evaluate the uncertainty of feature types by using traditional techniques of uncertainty analysis (e.g., variance and entropy). Our proposed approach can benefit to geospatial applications such as (1). recommending places for users; (2). cleaning up the data; (3). aligning place place across different geospatial data sources.