Organized by Bryce C. Newell (University of Kentucky), Michael Katell (University of Washington), and Jeff Hemsley (Syracuse University)
We invite you to download and play with the data prior to the conference. For some sample analysis, see the images in the text below as well as the Electronic Frontier Foundation’s analysis of some of the Oakland data here.
We invite you to participate with us in this interactive session at iConference 2018 as we explore, analyze, and visualize automated license/number plate recognition databases from the Boston, Seattle, and Oakland police departments. Would you like access to the databases? Please download them here. We hope to have participants with expertise in (geospatial) data analysis, data visualization, and data ethics/law/policy, and to spark conversations about what researchers can (and should) do with this sort of data.
The growing ability of law enforcement to easily and affordably draw information-rich insights from the surveillance data they collect implicates important social, political, and ethical concerns. For example, many police departments have incorporated automatic license plate recognition (ALPR) into their surveillance portfolios by equipping squad cars with devices that capture the license plate number of every vehicle that comes into sensing range. The fact that some police agencies have been legally required to disclose ALPR databases to members of the public creates an opportunity for researchers seeking to better understand the impacts that the creation, analysis, and disclosure of these surveillance databases can have in society. The purpose of these sessions is to provide scholars access to ALPR data and to provide an interactive and collaborative opportunity for scholars from multiple disciplines to investigate the relevant social, technical, and ethical ramifications through data visualization, data analysis, and ethical/policy-based analysis.
The purpose of the proposed SIE is to provide an interactive and collaborative opportunity for scholars from multiple disciplines to investigate the social, technical, and ethical ramifications arising from the collection and public disclosure of large police surveillance databases. The intended audience includes scholars from a variety of disciplines within information science/studies, including data visualization, data science, social informatics, computational social science, geographic information systems (GIS), data curation, data and information ethics, and information policy. During the opening session, we plan to make space for scholars across these areas to join together to dive into the available data and make plans for visualizing and/or analyzing the data, and for unpacking the social and ethical implications that such analyses of ALPR data engender.
Prior to the conference, we will make the datasets available to potential participants here, both to allow additional time for those who would like to begin exploring the data early as well as to promote the session and generate interest among potential participants. Importantly, the license plate numbers in the scan data we provide has been altered to protect the privacy of those drivers whose plates have been captured in the datasets.
At the conference, we will hold an opening session near the beginning of the conference in which we will make multiple ALPR datasets available to participants (who haven’t already downloaded them), with the goal of spurring multiple, parallel efforts to dive into the data over the course of the conference, culminating in a closing session where participants come back together and present their analyses (empirical, technical, or ethical) and/or data visualizations and discuss the social and ethical implications that have arisen during their analysis. The data shared with participants will include multiple databases generated by the use of ALPR capture devices (cameras) by police departments in Washington, California, and Massachusetts, and that generally include precise latitudinal and longitudinal data from the scans of a vehicle’s license plate, actual license plate numbers, as well as date and timestamp information. The data was sourced through public records requests under each state’s access to information law and, as such, is considered part of the public record. (The use of ostensibly “public” but non-consensually sourced data for research provides a secondary opportunity for scholarly discourse in these sessions.)
Post-conference. After the conference, we plan to post the results of the SIE (e.g., visualizations, short summaries of analytic findings, etc.) to the website, and to organize these results into a collaboratively-written report.