System and method for implementing self learning corrections call monitoring

Patent metadata

Field	Value
Publication number	US-10404853-B2
Application number	US-201615256293-A
Country	US
Kind code	B2
Filing date	Sep 2, 2016
Priority date	Mar 9, 2016
Publication date	Sep 3, 2019
Grant date	Sep 3, 2019

Abstract

Novel tools and techniques might provide for implementing call monitoring, and, in particular embodiments, to methods, systems, apparatuses, and computer software for implementing self-learning corrections facility call monitoring. In some embodiments, a method might comprise an automated call monitor monitoring a voice or video call between a call participant and a prison inmate, and analyzing at least one of conversation content of the call, voice characteristics of the call participant and/or the prison inmate, and/or (for video calls) body language of the parties to the call, to determine whether the conversation content, voice characteristics, and/or body language are indicative of impermissible content and/or impermissible conduct (i.e., indicative of past, present, and/or future criminal behavior or acts, and the like). The automated call monitor might flag calls that present such indications of impermissible content and/or conduct, for review by corrections facility personnel, and might self-learn based on input by such personnel.

First claim

What is claimed is:

1. A method for implementing self-learning corrections facility call monitoring, comprising:
monitoring, with a computing system, a call between an original call participant and a prison inmate;
analyzing, with the computing system, conversation content of the call to determine whether the conversation content comprises impermissible content;
analyzing, with the computing system, voice characteristics of each of the original call participant and the prison inmate evident during the call to determine whether the voice characteristics of at least one of the original call participant or the prison inmate are indicative of impermissible conduct conducted by at least one of the original call participant or the prison inmate, wherein the voice characteristics of each of the original call participant and the prison inmate that are monitored for and are indicative of impermissible conduct comprise at least one of voice characteristics indicative of fear, voice characteristics indicative of anger, voice characteristics indicative of a threat, voice characteristics indicative of maliciousness, voice characteristics indicative of resignation, voice characteristics indicative of hesitation, vocal stutter, or growl;
identifying, with the computing system, one or more coded messages by analyzing the voice characteristics of each of the original call participant and the prison inmate in conjunction with key words used by at least one of the original call participant or the prison inmate across a plurality of monitored calls; and
based upon at least one of a determination that the conversation content comprises impermissible content or a determination that the voice characteristics of at least one of the original call participant or the prison inmate are indicative of impermissible conduct, conducted by at least one of the original call participant or the prison inmate, flagging, with the computing system, the call for review by corrections facility personnel.

2. The method of claim 1 , wherein analyzing, with the computing system, the conversation content of the call to determine whether the conversation content comprises impermissible content comprises analyzing, with the computing system, the conversation content of the call, by comparing the conversation content with at least one of a first set of conversation content that has previously been identified by corrections facility personnel as being impermissible content or a second set of conversation content that has previously been identified by corrections facility personnel as not being impermissible content.

3. The method of claim 2 , further comprising:
receiving, with the computing system, a first input from the corrections facility personnel regarding whether the conversation content constitutes impermissible content;
based on a determination that the first input indicates that the conversation content constitutes impermissible content, updating, with the computing system, the first set of conversation content on a database; and
based on a determination that the first input indicates that the conversation content does not constitute impermissible content, updating, with the computing system, the second set of conversation content on the database.

4. The method of claim 3 , further comprising:
analyzing, with the computing system, at least one of points of similarity and points of distinction between the conversation content and each of the first set of conversation content and the second set of conversation content, in view of the first input;
wherein updating the first set of conversation content on the database comprises updating, with the computing system, the first set of conversation content on the database, based at least in part on the first input and at least in part on the analysis of the at least one of points of similarity and points of distinction between the conversation content and each of the first set of conversation content and the second set of conversation content; and
wherein updating the second set of conversation content on the database comprises updating, with the computing system, the second set of conversation content on the database, based at least in part on the first input and at least in part on the analysis of the at least one of points of similarity and points of distinction between the conversation content and each of the first set of conversation content and the second set of conversation content.

5. The method of claim 1 , wherein analyzing, with the computing system, the voice characteristics of each of the original call participant and the prison inmate evident during the call to determine whether the voice characteristics of at least one of the original call participant or the prison inmate are indicative of impermissible conduct comprises analyzing, with the computing system, the voice characteristics of each of the original call participant and the prison inmate evident during the call, by comparing the voice characteristics of each of the original call participant and the prison inmate with at least one of a first set of voice characteristics that has previously been identified by corrections facility personnel as being indicative of impermissible conduct or a second set of voice characteristics that has previously been identified by corrections facility personnel as not being indicative of impermissible conduct.

6. The method of claim 5 , further comprising:
receiving, with the computing system, a second input from the corrections facility personnel regarding whether the voice characteristics of the at least one of the original call participant or the prison inmate constitutes indication of impermissible conduct;
based on a determination that the second input indicates that the voice characteristics of the at least one of the original call participant or the prison inmate constitutes indication of impermissible conduct, updating, with the computing system, the first set of voice characteristics on a database; and
based on a determination that the second input indicates that the voice characteristics of the at least one of the original call participant or the prison inmate does not constitute indication of impermissible conduct, updating, with the computing system, the second set of voice characteristics on the database.

7. The method of claim 6 , further comprising:
analyzing, with the computing system, at least one of points of similarity and points of distinction between the voice characteristics of the at least one of the original call participant or the prison inmate and each of the first set of voice characteristics and the second set of voice characteristics, in view of the second input;
wherein updating the first set of voice characteristics on the database comprises updating, with the computing system, the first set of voice characteristics on the database, based at least in part on the second input and at least in part on the analysis of the at least one of points of similarity and points of distinction between the voice characteristics of the at least one of the original call participant or the prison inmate and each of the first set of voice characteristics and the second set of voice characteristics; and
wherein updating the second set of voice characteristics on the database comprises updating, with the computing system, the second set of voice characteristics on the database, based at least in part on the second input and at least in part on the analysis of the at least one of points of similarity and points of distinction between the voice characteristics of the at least one of the original call participant or the prison inmate and each of the first set of voice characteristics and the second set of voice characteristics.

8. The method

Assignees

• Centurylink Ip Llc

Inventors

• Smith Stephen R

Classifications

• H04M2201/41

  using speaker recognition · CPC title

• H04M3/42221

  Conversation recording systems (at the subscriber's set H04M1/656) · CPC title

• G10L25/51

  for comparison or discrimination · CPC title

• H04M3/2281 Primary

  Call monitoring, e.g. for law enforcement purposes; Call tracing; Detection or prevention of malicious calls · CPC title