Study on the Effects of Human Intention on Spatial Data Quality

Abstract

GIScience 2016 Short Paper Proceedings Study on the Effects of Human Intention on Spatial Data Quality Bo Zhao College of Earth, Ocean, and Atmospheric Sciences, Oregon State University, 101 SW 26th St, Corvallis, OR 97331 Email: zhao2@oregonstate.edu Abstract The GIScience community has devoted considerable research to locational uncertainty rather than other types of spatial data quality issues. Today, the flood of spatial big data has brought about new concerns, such as location spoofing, GPS jamming, or AIS (Automatic Identification System) hacking. Yet, the current data quality assessment framework falls short in defining, interpreting and analyzing these critical issues. By examining the reasons for measuring the geographic world, I suggest a modification of the hypothesis of the rational geographer in this paper, and further to analyze the distinctions among mistakes, spoofing and uncertainties, with the goal of placing the identified types of locational inconsistencies into a more holistic theoretical framework for spatial data quality. I call on GIScientists to pay more attention to the role of human intentions and advocate for a human-centric assessment of spatial data generation. Only then can we more effectively handle the emerging quality issues in the era of big data. 1. Introduction The GIScience community frequently focuses on uncertainty with regard to spatial data quality (Devillers et al., 2010). In the era of big data, the advent of mobile, social and geospatial technologies has created a considerable degree of heterogeneous, real-time and geo-referenced data. A large percentage of such data, especially VGI (Volunteered Geographic Information), geo- tagged social media, or location based service feeds, may be generated by ones’ mistakes and/or created deliberately rather than being merely affected by inherent uncertainties. Although geographers have recognized the significance of human intention, the motivations of the data generator are seldom examined, and the recent popularity of spoofings, especially those in the form of location spoofing (Zhao, 2015), GPS jamming (Grant et al., 2009) and AIS hacking (McCauley et al., 2016) are often dismissed by GIScientists. In order to more effectively address this critical issue, I will investigate the role of human intentions in the process of spatial data generation and clearly distinguish among uncertainty, mistakes and spoofing. In the remainder of this paper, I will review the concepts of error, accuracy, precision, uncertainty, mistakes, and spoofing in the context of geography. Then, I will examine the role of human motivation with regard to spatial data quality, and end with a brief concluding remark. 2. Uncertainty, mistakes, and spoofing Since human beings are forced to view the world through a fuzzy and distorting lens, the measured data are inevitably generalized, approximated, and subject to uncertainties (Zhang & Goodchild, 2002). In other words, the way we observe the world has invoked an inevitable locational (or positional) inconsistency between any observed object in the geographic world and the data that it produces (value of the object being measured). When referring to such locational inconsistencies, geographers usually consider them to be underlying uncertainties. One older and simpler term to describe uncertainty is error. By definition, error is the difference between the measured value and the “true” value of the object being measured. It is represented as an estimation of the range of values within which the true value is likely to be found. There are two types of errors: systematic