talk: Evaluating privacy for smart space occupancy data
The UMBC Cyber Defense Lab presents
Empirical Evaluation of Diverse PETs to Publish Smart Space Occupancy Data
Roberto Yus, UMBC
Joint work with Alisa Pankova, Peeter Laud, Sharad Mehrotra, and Sameera Ghayyur
12-1pm ET, Friday, 28 April 2023 via WebEx
We experimentally evaluate diverse privacy-enhancing technologies (PETs) for publishing occupancy data derived from continuous sensor streams in emerging smart buildings. Ensuring an individual's privacy in such contexts, especially with formal privacy guarantees, is a hard challenge that has attracted significant research interest. Various techniques that offer different levels of privacy guarantees have been proposed. Techniques derived from differential privacy define privacy goals (such as protecting published data from revealing a user's precise location or participation for an event during any small interval of time) and offer formal privacy guarantees (for the defined goals). Other techniques motivated by de-linking strategies, offer practical privacy but do not offer formal privacy guarantees.
Our study evaluates the practical implications to an individual's privacy, defined as the certainty at which the location of an individual can be determined at a given time, from publishing a continuous occupancy map generated from streaming sensor data. We present our methodology for computing such adversarial guesses based on aggregated data published while using three different PETs. Additionally, we present realistic adversaries in our context, including a rogue building administrator and a group of students trying to stalk others. We empirically evaluate our methodology on a real dataset containing three months of occupancy levels of a university building derived from connectivity events captured by the 64 WiFi access points within the building.
Roberto Yus (ryus@umbc.edu) is an assistant professor in the Department of Computer Science and Electrical Engineering at UMBC, where he leads the DAMS (DAta Management & Semantics) Research Group. His research interests concern data management, knowledge representation, data privacy, and the Internet of Things. Specifically, he focuses on semantic- and privacy-preserving IoT data management.
Host: Alan T. Sherman, sherman@umbc.edu. Support for this event was provided in part by the National Science Foundation under SFS grant DGE-1753681. The UMBC Cyber Defense Lab meets biweekly Fridays 12-1pm. All meetings are open to the public. Upcoming CDL meetings: May 5, UMBC CSEE Research Day, May 12, Kia-Won-Tia, UMBC Cyberdawgs
Empirical Evaluation of Diverse PETs to Publish Smart Space Occupancy Data
Roberto Yus, UMBC
Joint work with Alisa Pankova, Peeter Laud, Sharad Mehrotra, and Sameera Ghayyur
12-1pm ET, Friday, 28 April 2023 via WebEx
We experimentally evaluate diverse privacy-enhancing technologies (PETs) for publishing occupancy data derived from continuous sensor streams in emerging smart buildings. Ensuring an individual's privacy in such contexts, especially with formal privacy guarantees, is a hard challenge that has attracted significant research interest. Various techniques that offer different levels of privacy guarantees have been proposed. Techniques derived from differential privacy define privacy goals (such as protecting published data from revealing a user's precise location or participation for an event during any small interval of time) and offer formal privacy guarantees (for the defined goals). Other techniques motivated by de-linking strategies, offer practical privacy but do not offer formal privacy guarantees.
Our study evaluates the practical implications to an individual's privacy, defined as the certainty at which the location of an individual can be determined at a given time, from publishing a continuous occupancy map generated from streaming sensor data. We present our methodology for computing such adversarial guesses based on aggregated data published while using three different PETs. Additionally, we present realistic adversaries in our context, including a rogue building administrator and a group of students trying to stalk others. We empirically evaluate our methodology on a real dataset containing three months of occupancy levels of a university building derived from connectivity events captured by the 64 WiFi access points within the building.
Roberto Yus (ryus@umbc.edu) is an assistant professor in the Department of Computer Science and Electrical Engineering at UMBC, where he leads the DAMS (DAta Management & Semantics) Research Group. His research interests concern data management, knowledge representation, data privacy, and the Internet of Things. Specifically, he focuses on semantic- and privacy-preserving IoT data management.
Host: Alan T. Sherman, sherman@umbc.edu. Support for this event was provided in part by the National Science Foundation under SFS grant DGE-1753681. The UMBC Cyber Defense Lab meets biweekly Fridays 12-1pm. All meetings are open to the public. Upcoming CDL meetings: May 5, UMBC CSEE Research Day, May 12, Kia-Won-Tia, UMBC Cyberdawgs
Posted: April 24, 2023, 8:40 AM