Back to results
US 10,477,021 B1Granted
Systems for detecting harassing communication
AssigneeCapital One Services, LLC
InventorsMichael Mossoba, Abdelkader Benkreira, Joshua Edwards
Life of this patent
Estimating the term…
Abstract
A system includes one or more memory devices storing instructions, and one or more processors configured to execute the instructions to perform steps of a method for detecting harassment. The system may receive communication data from a first customer service representative terminal or a customer communication device and may determine, based at least in part on the communication data and stored communication data, that the communication data comprises harassing content. The system may, responsive to determining that the communication data comprises harassing content, transmit to a second customer service representative terminal, a notification indicating that the communication data comprises harassing content.
Claims10
A system for harassment detection comprising:
- one or more processors;
- and
- a memory coupled to the one or more processors and storing instructions that, when executed by the one or more processors, cause the one or processors to:
- receive voice data from at least one of a first customer service representative (CSR) terminal and a customer communication device, the voice data comprising data corresponding to spoken communication from a user of the first CSR terminal during an electronically facilitated conversation or a user of the customer communication device during the electronically facilitated conversation;
- determine, based at least in part on the voice data and at least one of stored semantics data, stored syntax data, stored intonation data, stored volume data, and stored non-word utterances data, that the voice data comprises harassing content by:
- deriving, from the voice data, at least three of derived semantics data, derived syntax data, derived intonation data, derived volume data, and derived non-word utterances data;
- determining, based on a predetermined list of harassment factors, one or more harassment scores for the at least three of the derived semantics data, derived syntax data, derived intonation data, derived volume data, and derived non-word utterances data;
- determining, at predetermined intervals during the electronically facilitated conversation, a conversation harassment score by:
- determining a level of similarity for each of the at least three of the derived semantics data, derived syntax data, derived intonation data, derived volume data, and derived non-word utterances data based on respective comparisons of at least three of the derived semantics data, derived syntax data, derived intonation data, derived volume data, and derived non-word utterances data to the stored semantics data, stored syntax data, stored intonation data, stored volume data, and stored non-word utterances data;
- and
- determining that the level of similarity for each of the at least three the derived semantics data, derived syntax data, derived intonation data, derived volume data, and derived non-word utterances data exceeds a predetermined similarity threshold;
- and
- determining that the conversation harassment score is above a predetermined harassment threshold;
- and
- responsive to determining that the voice data comprises harassing content based on the conversation harassment score being above the predetermined threshold, transmit, to a second CSR terminal, a notification indicating that the voice data comprises harassing content.
12 more elements
A system for harassment detection comprising:
- one or more processors;
- and
- a memory coupled to the one or more processors and storing instructions that, when executed by the one or more processors, cause the one or more processors to:
- receive customer service representative (CSR) voice data from a first CSR terminal and customer voice data from a customer communication device, the CSR voice data comprising data corresponding to spoken communication from a user of the first CSR terminal during an electronically facilitated conversation and the customer voice data comprising data corresponding to spoken communication from a user of the customer communication device during the electronically facilitated conversation;
- determine, based at least in part on a comparison of stored historical linguistics data to at least one of the received CSR voice data and the received customer voice data, that the CSR voice data or customer voice data comprises harassing content;
- transmit, to a second CSR terminal, a notification indicating that the CSR voice data or customer voice data comprises harassing content
- determine, based at least in part on at least one of the received CSR voice data and the received customer voice data, a communication topic associated with the spoken communication;
- determine, based at least in part on the stored historical linguistics data and at least one of the received CSR voice data and the received customer voice data, a likelihood that the user of the first CSR terminal or the user of the customer communication device will communicate harassing content,
- wherein determining the likelihood that the user of the first CSR terminal or the user of the customer communication device will communicate harassing content comprises determining that a number of previous electronically facilitated conversations associated with the user of the customer communication device and the communication topic exceeds a predetermined threshold;
- and
- responsive to determining that the likelihood that the user of the first CSR terminal or the user of the customer communication device will communicate harassing content is above a predetermined threshold, transmit an imminent harassment notification to the user of the second CSR terminal.
8 more elements
Dependencies are inferred from the claim text.