1. Abstract Privacy has always been a public concern. It can be dated back to “media privacy” of 1361 that was enacted to protect people from peeping toms in England. In 19th century, privacy became a hot issue with the arrival of modern photography and printing press that enabled easy reporting of personal information and pictures, automated data processing to catalog citizens during World War II, and record keeping Systems in the U.S. These actions led to the enactment of a number of privacy policies, specifically, focusing on data protection - mis-use of data and/or malicious access of data by unauthorized persons. In this era of the Internet, ubiquitous computing combined with the growth of the geo-spatial and the Information and Communication Technology industries has made it possible for anyone to access personal and location information of another person anytime and anywhere. For instance, GeoAPI of Twitter (a social media service) can help track the movement of a person over space and time. Despite having regulatory policies, it is possible to extract location information of a person by using VGI (Volunteered Geographic Information) and CGI (Contributed Geographic Information) available from social media sites. This study explores the extent to which location information obtained from social media sites are reliable and useful and are influenced by an individual's concern and knowledge of privacy.
The revolutionary changes in geo-spatial technologies (e.g. Global Positioning System (GPS), ultra-wide-band radio) have enabled the collection and generation of a large amount of geospatial data at different levels of accuracy, coverage and cost. The growth of the Information and Communication Technologies (ICT) (e.g. mobile devices, cell towers, the Internet) has led to the emergence of location-based services (LBS) (Google Latitude and Google Street View) and social media sites (e.g. Twitter and Facebook). Using spatial data, these services and sites provide information about an individual’s or a vehicle’s location accurately and precisely. Given the uncertainty about how this information can violate a person’s privacy, this study explores the extent to which people’s knowledge and concern of privacy influences their decision to share location information through VGI (Volunteered Geographic Information) and CGI (Contributed Geographic Information).
Location-based services and social media sites singularly do not violate personal information.
However, by coordinating location information with other types of information, such as an
individual’s address, these services can provide personal information to a third party, thereby leading
to location privacy violation. Location privacy though is an unclear concept, it is defined as “the
ability to prevent other parties from learning one’s current or past location” or “the ability of an
individual to move in public space with the expectation that under normal circumstances their location
will not be systematically and secretly recorded for later use”
Personal privacy - “the quality or state of being apart from company or observation” or
“freedom from unauthorized intrusion” (Webster 2010) - has received extensive legal attention with
the publication of the 1890 Harvard Law Review article “the right to be let alone”
Though the legal underpinning exists, it is possible to harvest spatial information from
crowdsourced geographic information (VGI and CGI)
Privacy has been a public concern since 1361 when “media privacy” was enacted to protect
people from peeping toms in England
Like privacy, surveillance is not a new concept. The first surveillance instrument - Bentham’s
Panopticon - a building within which all occupants could be tracked from one vista point - was
designed by Samuel Bentham in 1785
Location privacy will not be a problem if people were not tempted to share their location information on social media sites and LBS. Given our dependency on ICT and the recent popularity of location-based services (e.g., smart phones, Twitter’s location API, Google Latitude, etc.), in addition to the legal communities, actions must be taken by users, developers and providers of these technologies and services to protect privacy and location privacy. Therefore, additional to legal protection, a number of computational countermeasures have been taken to protect personal and location information of a user.
The users of location aware services may use self-regulatory techniques, such as, providing
limited personal information, adjusting privacy settings, limiting access to certain people, disabling or
refusing to use applications that may lead to losing privacy or provide limited privacy protection, and
not opting to share certain information, to protect their privacy
To protect users’ privacy, developers use a number of computation approaches: 1)
Anonymity: a user’s location information is same as other users; 2) Obfuscation: the accuracy of
location information is reduced by introducing noise; 3) Aggregation: a user’s location information is
aggregated with other users’ location data; 4) Encryption: the true location information is fudged
and/or encrypted
For these approaches to be successful users must be aware of their presence, and users must know that their privacy is/can be lost because third parties can access their personal and location information. A user must also take appropriate actions to prevent information sharing. In any case, there is still the possibility of extracting location information from social media sites that have significant potential for real life applications (e.g. crisis mapping). This raises the question, how much crowd-sourced geographic information is reliable and useful.
The study-site will be the University of Southern Mississippi, which is located in the City of
Hattiesburg, Mississippi. System architecture similar to the one presented in Stefanidis et al. (2011)
will be used to extract information from only one social media site: Twitter. Using twitter’s native
API and Python, both textual and graphic data will be extracted for two weeks duration for specific
hash tags (e.g., USM, Hattiesburg, and University). In addition to extracting real time data, data for
the same duration will be purchased from the third party vendor (Gnip) that provides access to twitter
feeds to increase data set size. By parsing the extracted data, the spatial information will be obtained
(this will include coordinates of the locations where the tweets have been posted). Likewise, from
textual and graphic data, geo-tagged information will be extracted for the study site and duration. By
geo-coding the location information, a point map will be created, which will be overlaid on the
University’s location information layer. Finally, the spatial distribution of error will be computed, and
total error and error variance based on spatial and temporal distribution will be computed to
understand the extent of data fudging done to protect location information. Finally, the survey used in
the study exploring knowledge of location privacy
We expect that the quality of the data is influenced by the situation rather than public’s concern of privacy. For instance, during a disaster, people are motivated to share accurate location information that is relevant to the disaster instead of their own personal space. We also expect that population of specific age group will be more concerned about their own privacy and protecting personal information. For instance, older generation (35+ years) will be more concerned about taking steps to protect their own privacy at any given time.
In the web 2.0 era, citizens a.k.a. “human sensors” provide valuable and timely information
about the post-disaster landscape that can be used in emergency response and crisis mapping