*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Title: Combating Robocalls to Enhance Trust in Converged Telephony
Sharbani Pandit
School of Computer Science
Georgia Institute of Technology
Date: Wednesday, October 7th, 2020
Time: 10:00 AM -12:00 PM (EDT)
Location: https://bluejeans.com/336575503
**Note: this proposal is remote-only**
Committee:
Dr. Mustaque Ahamad (Advisor), School of Computer Science, Georgia Institute of Technology
Dr. Roberto Perdisci, School of Computer Science, University of Georgia, Georgia Institute of Technology
Dr. Diyi Yang, School of Interactive Computing, Georgia Institute of Technology
Dr. Mostafa Ammar, School of Computer Science, Georgia Institute of Technology
Dr. Lillian Lo, Principal Data Scientist, AT&T
Abstract:
Telephone scams are now on the rise and without effective countermeasures the number of scam/spam call people receive will continue to increase. Voice scams have become such a serious problem that people often no longer pick up calls from unknown callers. The vision of this research is to bring trust back to the telephony channel. We believe this can be done by stopping unwanted and fraud calls and offering a novel interaction model that can help enhance the trust and effectiveness of voice interactions.
We study how to automatically build phone blacklists from multiple data sources and evaluate the effectiveness of such blacklists in the stopping current robocalls. To address the threat model where caller ID is spoofed, we introduce the notion of a virtual assistant aimed at filtering out unwanted calls. To this end, we developed a Smartphone based app named RobocallGuard which can pick up calls from unknown callers on behalf of the callee and filter out non-targeted calls. We conduct a user study that shows that users are comfortable with a virtual assistant filtering out unwanted calls on their behalf. Finally, we introduce SmartVA which can effectively block targeted robocalls. SmartVA uses a combination of NLP based machine learning models to determine if the caller is a human or a robocaller. To the best of our knowledge, we are the first to develop such a defense system that can interact with the caller and detect robocalls where robocallers utilize caller ID spoofing and voice activity detection to bypass the defense mechanism. By making these contributions, we aim at bringing trust back to the telephony channel and making a better telephony experience for everyone.
