In the ongoing battle against robocalls, a new tool called SnorCall has emerged, aiming to revolutionize the fight against the rampant wave of unsolicited calls.
Designed to analyze the content of these calls, SnorCall seeks to provide insights into the issue's magnitude and the types of scams perpetrated by robocalls.
The creators of the tool foresee its application by regulators, phone carriers, and various stakeholders to enhance comprehension, monitoring, and mitigation of robocall trends. It aims to culminate in proactively addressing criminal activities linked to these calls.
Combatting Robocalls
Brad Reaves, assistant professor of Computer Science at North Carolina State University and corresponding author of a related paper, highlighted the challenge service providers and researchers face. While they can access call metadata, they lack the tools to scrutinize the actual content of robocalls on the necessary scale.
The concerns related to privacy issues have impeded such investigations. SnorCall offers a resolution that characterizes robocall content while upholding privacy, thanks to a partnership with telecommunications company Bandwidth.
The tool utilizes over 60,000 phone numbers dedicated solely to monitoring unsolicited robocalls, eliminating the need to use customers' phone numbers.
SnorCall operates by recording all incoming robocalls on monitored phone lines, grouping calls with similar audio to streamline content analysis. These recorded calls are then transcribed and assessed using Snorkel, a machine-learning framework, to identify the purpose of each call.
The framework assigns labels to indicate the call's focus, whether it pertains to a specific company or government program, solicits personal information, or requests funds. The insights derived from SnorCall can uncover trends and patterns associated with robocall scams.
"We've developed a tool that allows us to the characterize the content of robocalls," Reaves said in a statement.
"And we've done it without violating privacy concerns; in collaboration with a telecommunications company called Bandwidth, we operate more than 60,000 phone numbers that are used solely by us to monitor unsolicited robocalls. We did not use any phone numbers of actual customers," he added.
Validating SnorCall
For validation, SnorCall was employed to analyze 232,723 robocalls collected over 23 months across the dedicated phone lines. The tool's analysis delved into 26,791 unique audio files, offering valuable insights into scam operations and patterns.
A noteworthy discovery was the ability to extract call-back numbers that scammers encourage recipients to dial, potentially aiding authorities in tracing origins and perpetrators.
Furthermore, SnorCall's application provided insights into the dynamic nature of scam operations. For instance, patterns emerged in the frequency of Social Security scam robocalls during the pandemic.
Reaves noted the tool's adaptability, explaining that Snorkel allows customization to cater to stakeholders' specific needs. This versatility enables SnorCall to identify keywords associated with specific scam topics, making it adaptable for investigative purposes.
The research's implications extend to events shaping scam trends and uncovering new scam types. As SnorCall offers a potential solution to the persistent robocall challenge, researchers seek to collaborate further with industry partners to enhance its effectiveness and meaningful impact on countering robocall scams.
The findings of the study will be presented at the 31st USENIX Security Symposium.
Related Article: FTC Chair Claims AI Voice Cloners Supercharge Scams! How Can You Protect Yourself?
