System and method for assessing security threats and criminal proclivities

• US 10,074,362 B2
• Filed: 01/31/2017
• Issued: 09/11/2018
• Est. Priority Date: 01/31/2017
• Status: Active Grant

• Current Assignee: Global Tel-Link Corporation
• Original Assignee: Global Tel-Link Corporation
• Inventors: Volkart, Mitch
• Primary Examiner(s): Pullias, Jesse
• Application Number:
  US15/420,921
• Publication Number:
  US 20180218730A1
• Time in Patent Office: 588 Days
• Field of Search: 704231-257, 704270-275
• CPC Class Codes: G06F 40/279 Recognition of textual entities
• CPC Class Codes: H04M 2201/41 using speaker recognition
• CPC Class Codes: H04M 3/2281 Call monitoring, e.g. for law enforcement purposes; Call tracing; Detection or prevention of malicious calls
• CPC Class Codes: H04M 3/387 using subscriber identification cards

Abstract

A centralized and robust threat assessment tool is disclosed to perform comprehensive analysis of previously-stored and subsequent communication data, activity data, and other relevant information relating to inmates within a controlled environment facility. As part of the analysis, the system detects certain keywords and key interactions with the dataset in order to identify particular criminal proclivities of the inmate. Based on the identified proclivities, the system assigns threat scores to inmate that represents a relative likelihood that the inmate will carry out or be drawn to certain threats and/or criminal activities. This analysis provides a predictive tool for assessing an inmate'"'"'s ability to rehabilitate. Based on the analysis, remedial measures can be taken in order to correct an inmate'"'"'s trajectory within the controlled environment and increase the likelihood of successful rehabilitation, as well as to prevent potential criminal acts.

Claims

1. A system for calculating a threat score for an inmate of a controlled environment facility, the system comprising:

a monitoring subsystem configured to;

receive an inmate communication;

capture a voice sample of the inmate from the inmate communication;

generate an inmate voiceprint based on the voice sample;

compare the inmate voiceprint to previously-stored voiceprints in order to identify the inmate involved in the inmate communication;

detect topics in the inmate communication; and

assign weights to the detected topics; and

a threat assessment subsystem configured to;

receive the detected topics and the assigned weights from the monitoring subsystem;

scale the weights assigned to the detected topics by a number of data points examined for the inmate;

normalize the scaled weights based on scaled weights of other inmates;

retrieve previously-stored threat information relating to the inmate;

compare the received detected topics and scaled weights to the previously-stored threat information; and

assign a threat score to the inmate based on the comparison.

2. The system of claim 1, wherein the monitoring subsystem is further configured to detect keywords in the inmate communication, and

wherein the topics are detected based on the detected keywords.

3. The system of claim 2, further comprising a keyword database that stores a plurality of predefined keywords and a plurality of predefined topics, each of the plurality of predefined keywords being associated with at least one of the plurality of predefined topics.

4. The system of claim 3, wherein the monitoring subsystem is configured to detect the topics by identifying, for each of the detected keywords, the stored plurality of predefined topics to which the detected keywords correspond.

5. The system of claim 4, wherein the monitoring subsystem is configured to assign the weights to the detected topics based on a relative amount of detected keywords that correspond to the detected topics.

6. The system of claim 1, wherein the received inmate communication is a voice communication, and

wherein the monitoring subsystem is further configured to transcribe the voice communication to text.

7. The system of claim 6, wherein the monitoring subsystem is configured to identify the inmate by:

parsing a plurality of voices within the voice communication based on an audio analysis;

generating a voiceprint for a parsed voice from among the plurality of voices; and

comparing the generated voiceprint to previously-stored voiceprints associated with inmates of the controlled environment facility.

8. A system for calculating a threat score for an inmate of a controlled environment facility, the system comprising:

a monitoring subsystem configured to;

monitor inmate communication and activities;

generate an inmate voiceprint from an inmate voice sample extracted from the inmate communication; and

identify the inmate based on the inmate voiceprint; and

a threat assessment subsystem configured to;

access inmate information from one or more databases that store communication and activity data of the inmate;

identify a plurality of data relationships within the communication and activity data;

determine corresponding strengths of the data relationships;

assign weights to the relationships based on the strengths;

identify threats corresponding to the detected relationships;

assign a positive score to all threats whose corresponding relationship weight exceeds a predetermined minimum value;

identify all other threats as non-observed; and

calculate the threat score based on the strengths.

9. The system of claim 8, wherein the one or more databases include call records, visitation records, and incident histories of the inmate.

10. The system of claim 8, wherein the identifying and determining are performed as part of a virtual link analysis of the communication and activity data.

11. The system of claim 8, wherein the threat assessment subsystem is configured to determine the strengths of the data relationships based on a number of logical connections between data elements.

12. The system of claim 8, wherein the threat assessment subsystem is configured to calculate the threat score for the inmate by:

scaling the weights assigned to the relationships by a number of data points examined for the inmate; and

normalizing the scaled weights based on scaled weights of other inmates.

13. The system of claim 8, wherein the calculating of the threat score further includes assigning a base value to all non-observed threats.

14. A method for calculating a threat score for an inmate of a controlled environment facility, the method comprising:

receiving an inmate communication;

generating an inmate voiceprint based on an inmate voice sample extracted from the inmate communication;

identifying the inmate based on the inmate voiceprint;

detecting topics in the inmate communication;

assigning weights to the detected topics;

scaling the weights assigned to the detected topics by a number of data points examined for the inmate;

normalizing the scaled weights based on scaled weights of other inmates;

comparing scaled weighted detected topics to previously-stored threat information; and

assigning a threat score to the inmate based on the comparison.

15. The method of claim 14, further comprising detecting keywords in the inmate communication,

wherein the topics are detected based on the detected keywords.

16. The method of claim 15, further comprising associating a plurality of predefined keywords with at least one of the plurality of predefined topics within a keyword database.

17. The method of claim 14, wherein the received inmate communication is a voice communication, the method further comprising:

transcribing the voice communication to text.

18. The method of claim 17, further comprising:

parsing a plurality of voices within the voice communication based on an audio analysis;

generating a voiceprint for a parsed voice from among the plurality of voices; and

comparing the generated voiceprint to previously-stored voiceprints associated with inmates of the controlled environment facility in order to identify the inmate.

19. The method of claim 14, wherein the topics are detected by identifying, for each of the detected keywords, the stored plurality of predefined topics to which the detected keywords correspond.