Akin to Junk mail or user solicitation, robo-calls rely on a common methodology. This methodology works by accosting individuals and inveigling a user reaction to the soliciting parties’ targeted interests. In the same light, any solution to identify or prevent robo-calling requires a framework that is dependent on the underlying behavioral responses that are observed with unsolicited call interactions, hence my proposed solution provides an end-user complementary service which offers options that would -classify calls with specific messages on the caller ID(Crowd sourced call data collation), offer pre-answer voice analysis for known automated recordings or provide a transparent caller pre acknowledgement voice ringtone itself (This allows for human heuristic feedback/ data points). Subsequently, based on the confidence level from the call markers, blocks for future calls can be applied or users can simply decline calls with a higher confidence level. This solution requires a user buy-in (or free subscription) and data collation and aggregation from the phone company or a vendor. The solution works as follows. *Regulatory requirement: Caller ID spoofing should be illegal 1) Phone users agree to provide call statistical information that is aggregated and feed into a collation engine.

Call Markers and Indices • Time to user hang-ups (From crowd sourced data) • Commonly Unanswered calls (From crowd sourced data) • Originating call volume data (From Telco call aggregation or Third party/crowd sourced user call history) • FCC Data (Used as whitelist or blacklist –Vetted organization, Government, Emergency Services, etc ) • User feedback from a voicemail options(Telcos provides an option to mark a received call voicemail as robo-calls) – Heuristic data • User feedback from pre caller voice acknowledgement ringer • Voice analysis from a spoofed call answer that tricks automated robot systems to trigger • Honey pot/Trap phone numbers donated by users (Used for as a tertiary data collection source since the numbers are not primary user numbers )

2) The engines algorithm defines a policy for classifying calls. 3) Additional caller ID information based on the classification is provided to the user 4) Numbers with high confidence intervals of being robo-calls are dropped.

Implementation Options 1) Smart Phone applications 2) Telco option 3) IP phone company rediretion

Share this project:

Updates