IP Library Granted Patent US 7,860,722
Granted Patent B1
US 7,860,722 · App. 11/334,522 · Granted Dec 28, 2010

System and method for keyword detection in a controlled-environment facility using a hybrid application

View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 7,860,722
App. No.
11/334,522
Granted
Dec 28, 2010
Kind
B1
Abstract

A method for keyword detection in a controlled-environment facility comprises capturing a conversation within the controlled-environment facility and executing a hybrid keyword detection application for detecting a keyword spoken during the conversation. A system comprises an audio capture system operable to capture an audio signal within a controlled-environment facility and a hybrid keyword detection system connected to the audio capture system, where the hybrid keyword detection system is operable to detect a keyword.

Claims (66)

1. A method comprising:

capturing audio at a telephone system switch or router for a controlled-environment facility, the audio comprising a voice signal from at least part of a conversation between at least one party within the controlled-environment facility and at least one other party; and

executing a hybrid keyword detection application on a hybrid keyword detection server, the hybrid keyword detection application receiving the captured audio from the telephone system switch or router and detecting whether at least one predefined keyword is present in the captured audio,

the hybrid keyword detection application first filtering the captured audio using a phonetic-based algorithm, the phonetic-based algorithm producing a first result comprising conversations potentially containing the predefined keywords, wherein the phonetic-based algorithm allows searches on how the keywords sound, and

then filtering only the captured audio from the first result using an acoustic-based algorithm to eliminate erroneous keyword detections by the phonetic-based algorithm, wherein the acoustic-based algorithm uses statistical methods to perform keyword detection the acoustic-based algorithm producing a second result; and

analyzing the first and second results to determine a keyword matching probability.

2. The method of claim 1 wherein said capturing the audio comprises:

recording the conversation.

3. The method of claim 1 wherein capturing the audio comprises:

intercepting an ongoing conversation.

4. The method of claim 1 wherein capturing the audio comprises:

capturing a telephone conversation.

5. The method of claim 1 wherein said capturing audio comprises:

capturing audio of a plurality of different conversations.

6. The method of claim 5 wherein said executing a hybrid keyword detection application comprises:

using a phonetic-based algorithm to determine ones of said plurality of different conversations in which said phonetic-based algorithm identifies said at least one predefined keyword as being present; and

using an acoustic-based algorithm to further analyze said determined ones of said plurality of different conversations to determine those of said determined ones of said plurality of different conversations in which said acoustic-based algorithm identifies said at least one predefined keyword as being present.

7. The method of claim 1 further comprising:

customizing at least one parameter of the hybrid keyword detection application for a particular controlled-environment facility.

8. The method of claim 1 wherein detecting the keyword comprises:

determining whether a keyword matching probability is greater than a threshold.

9. The method of claim 1 further comprising:

executing a predefined action associated with a detected keyword.

10. The method of claim 1 further comprising communicating said captured audio via a Voice over Internet Protocol (VoIP) network to said hybrid keyword detection application.

11. The method of claim 10 further comprising:

communicating corresponding metadata for said captured audio to said hybrid keyword detection application.

12. A system comprising:

an audio capture system in a telephone system switch or router operable to capture a voice signal for a conversation conducted between a resident of a controlled-environment facility and another party; and

a hybrid keyword detection system coupled to the telephone system switch or router and executing a hybrid keyword detection application that is operable to perform phonetic-based keyword detection and acoustic-based keyword detection on the voice signal captured from the telephone system switch or router, wherein the hybrid keyword detection system is operable to detect whether at least one predefined keyword is present in the received voice signal,

the hybrid keyword detection application first filtering the captured voice signal using a phonetic-based algorithm to identify conversations potentially containing the at least one predefined keyword, the phonetic-based algorithm producing a first result, and

then filtering only the captured voice signal from the first result using an acoustic-based algorithm to eliminate erroneous keyword detections by the phonetic-based algorithm, the acoustic-based algorithm producing a second result,

analyzing the first and second results to determine a keyword matching probability, and

determining whether the keyword match probability is greater than a threshold.

13. The system of claim 12 wherein the audio capture system comprises:

a controlled-environment facility's telephone system.

14. The system of claim 12 wherein the hybrid keyword detection system further comprises:

a hybrid keyword detection server; and

a database communicatively coupled to the hybrid keyword detection server.

15. The system of claim 12 wherein the hybrid keyword detection application is customized for a particular controlled-environment facility.

16. The system of claim 12 wherein detecting whether the at least one predefined keyword is present in the received audio comprises determining whether a keyword matching probability is greater than a threshold.

17. The system of claim 12 wherein said audio system captures audio signals for a plurality of conversations, and wherein said hybrid keyword detection application is operable to:

a) use a phonetic-based algorithm to determine ones of said plurality of conversations in which said phonetic-based algorithm identifies said at least one predefined keyword as being present; and

b) use an acoustic-based algorithm to further analyze said determined ones of said plurality of conversations to determine those of said determined ones of said plurality of conversations in which said acoustic-based algorithm identifies said at least one predefined keyword as being present.

18. The system of claim 12 further comprising a Voice over Internet Protocol (VoIP) network over which said captured audio signal is communicated to said hybrid keyword detection system.

19. The system of claim 18 wherein corresponding metadata for said captured audio signal is communicated to said hybrid keyword detection system.

20. The system of claim 12 wherein the audio signal is a recorded conversation.

21. The system of claim 12 wherein the audio signal is an ongoing conversation.

22. The system of claim 12 wherein the audio signal is a telephone conversation.

23. The system of claim 12 further comprising:

a client communicatively coupled to the hybrid keyword detection system, wherein the hybrid keyword detection system is configurable by said client.

24. The system of claim 12 wherein the hybrid keyword detection system executes an action associated with a detected keyword.

25. The system of claim 24 further comprising:

a client communicatively coupled to the hybrid keyword detection system, wherein the action associated with a corresponding keyword is configurable by said client.

26. A system comprising:

a plurality of telephone system switches or routers, wherein each telephone system switches or router is operable to capture voice signals from telephone communications within one of a plurality of geographically distributed controlled-environment facilities; and

a centralized hybrid keyword detection system connected to each of the plurality of telephone system switches or routers, where the hybrid keyword detection system is operable to use both phonetic-based keyword detection and acoustic-based keyword detection for detecting a keyword within the voice signals captured by the telephone system switches or routers,

the hybrid keyword detection system first filtering the captured voice signals using a phonetic-based algorithm to identify conversations potentially containing predefined keywords, the phonetic-based algorithm producing a first result,

then filtering only the captured voice signals within the first result using an acoustic-based algorithm to eliminate erroneous keyword detections by the phonetic-based algorithm, the acoustic-based algorithm producing a second result

analyzing the first and second result to determining a keyword matching probability, and

determining whether the keyword matching probability is greater than a threshold.

27. The system of claim 26 where the hybrid keyword detection system comprises:

a hybrid keyword detection server; and

a database connected to the hybrid keyword detection server.

28. The system of claim 27 where the hybrid keyword detection server is operable to execute a hybrid application for detecting the keyword.

29. The system of claim 26 where the hybrid keyword detection system is connected to each of the plurality of audio capture systems via a computer network.

30. The system of claim 26 wherein said hybrid keyword detection system uses said phonetic-based keyword detection to determine ones of a plurality of captured audio signals captured by said plurality of audio capture systems in which said phonetic-based keyword detection identifies said keyword as being present; and wherein said keyword detection system further uses said acoustic-based keyword detection to further analyze said determined ones of said plurality of captured audio signals to verify the presence of said keyword therein.

Assignments (30)
ASSIGNMENT OF SECURITY INTEREST IN PATENTS Recorded Apr 21, 2025
From: DEUTSCHE BANK AG NEW YORK BRANCH, AS EXISTING COLLATERAL AGENT
To: WILMINGTON SAVINGS FUND SOCIETY, FSB, AS SUCCESSOR COLLATERAL AGENT
Reel/Frame 070903/0303 →
RELEASE OF SECURITY INTEREST Recorded Mar 29, 2024
From: ALTER DOMUS (US) LLC
To: AVENTIV TECHNOLOGIES, LLC; SECURUS TECHNOLOGIES, LLC; ALLPAID, INC.
Reel/Frame 066951/0630 →
RELEASE OF SECURITY INTEREST Recorded Mar 29, 2024
From: ALTER DOMUS (US) LLC
To: AVENTIV TECHNOLOGIES, LLC; SECURUS TECHNOLOGIES, LLC; ALLPAID, INC.
Reel/Frame 066951/0385 →
RELEASE OF SECURITY INTEREST Recorded Mar 29, 2024
From: ALTER DOMUS (US) LLC
To: AVENTIV TECHNOLOGIES, LLC; SECURUS TECHNOLOGIES, LLC; ALLPAID, INC.
Reel/Frame 066951/0514 →
RELEASE OF SECURITY INTEREST Recorded Mar 29, 2024
From: ALTER DOMUS (US) LLC
To: AVENTIV TECHNOLOGIES, LLC; SECURUS TECHNOLOGIES, LLC; ALLPAID, INC.
Reel/Frame 066951/0561 →
RELEASE OF SECURITY INTEREST Recorded Mar 29, 2024
From: ALTER DOMUS (US) LLC
To: AVENTIV TECHNOLOGIES, LLC; SECURUS TECHNOLOGIES, LLC; ALLPAID, INC.
Reel/Frame 066952/0914 →
SECURITY INTEREST Recorded Mar 29, 2024
From: AVENTIV TECHNOLOGIES, LLC; SECURUS TECHNOLOGIES, LLC; SATELLITE TRACKING OF PEOPLE LLC; ALLPAID, INC.
To: WILMINGTON SAVINGS FUND SOCIETY, FSB
Reel/Frame 066951/0054 →
SUPER-PRIORITY FIRST LIEN PATENT SECURITY AGREEMENT Recorded Mar 28, 2024
From: SECURUS TECHNOLOGIES, LLC; SATELLITE TRACKING OF PEOPLE LLC; ALLPAID, INC.
To: DEUTSCHE BANK AG NEW YORK BRANCH AS COLLATERAL AGENT
Reel/Frame 066945/0310 →
PRIORITY SECOND LIEN PATENT SECURITY AGREEMENT Recorded Feb 1, 2024
From: SECURUS TECHNOLOGIES, LLC; SATELLITE TRACKING OF PEOPLE LLC; ALLPAID, INC.
To: WILMINGTON SAVINGS FUND SOCIETY, FSB AS COLLATERAL AGENT
Reel/Frame 066436/0420 →
PRIORITY FIRST LIEN PATENT SECURITY AGREEMENT Recorded Jan 2, 2024
From: SECURUS TECHNOLOGIES, LLC; ALLPAID, INC.; SATELLITE TRACKING OF PEOPLE LLC
To: DEUTSCHE BANK AG NEW YORK BRANCH, AS COLLATERAL AGENT
Reel/Frame 066567/0620 →
SECURITY INTEREST Recorded Aug 24, 2023
From: ALLPAID, INC.; SATELLITE TRACKING OF PEOPLE LLC; SECURUS TECHNOLOGIES, LLC
To: THE BANK OF NEW YORK MELLON TRUST COMPANY, N.A.
Reel/Frame 064699/0842 →
CHANGE OF NAME Recorded Aug 12, 2020
From: SECURUS TECHNOLOGIES, INC.
To: SECURUS TECHNOLOGIES, LLC
Reel/Frame 053476/0289 →
SECOND LIEN SECURITY AGREEMENT Recorded Nov 6, 2017
From: SECURUS TECHNOLOGIES HOLDINGS, INC.; SECURUS TECHNOLOGIES, INC.; SATELLITE TRACKING OF PEOPLE LLC
To: DEUTSCHE BANK AG NEW YORK BRANCH, AS COLLATERAL AGENT
Reel/Frame 044368/0969 →
FIRST LIEN SECURITY AGREEMENT Recorded Nov 3, 2017
From: SECURUS TECHNOLOGIES HOLDINGS, INC.; SECURUS TECHNOLOGIES, INC.; SATELLITE TRACKING OF PEOPLE LLC
To: DEUTSCHE BANK AG NEW YORK BRANCH, AS COLLATERAL AGENT
Reel/Frame 044779/0573 →
FIRST LIEN TERMINATION AND RELEASE OF SECURITY INTEREST IN PATENTS RECORDED AT REEL 030369, FRAME 0648 Recorded Nov 2, 2017
From: DEUTSCHE BANK TRUST COMPANY AMERICAS
To: SECURUS TECHNOLOGIES, INC.; SECURUS TECHNOLOGIES HOLDINGS
Reel/Frame 044681/0227 →
SECOND LIEN TERMINATION AND RELEASE OF SECURITY INTEREST IN PATENTS RECORDED AT REEL 030368, FRAME 0743 Recorded Nov 2, 2017
From: DEUTSCHE BANK TRUST COMPANY AMERICAS
To: SECURUS TECHNOLOGIES, INC.; SECURUS TECHNOLOGIES HOLDINGS
Reel/Frame 044361/0267 →
RELEASE OF 2ND LIEN SECURITY INTEREST Recorded Nov 1, 2017
From: DEUTSCHE BANK TRUST COMPANY AMERICAS
To: SECURUS TECHNOLOGIES, INC.; SECURUS TECHNOLOGIES HOLDINGS, INC.
Reel/Frame 044653/0030 →
RELEASE OF 1ST LIEN SECURITY INTEREST Recorded Nov 1, 2017
From: DEUTSCHE BANK TRUST COMPANY AMERICAS
To: SECURUS TECHNOLOGIES, INC.; SECURUS TECHNOLOGIES HOLDINGS, INC.
Reel/Frame 044640/0547 →
RELEASE OF SECURITY INTEREST SECOND LIEN Recorded May 8, 2013
From: BNP PARIBAS
To: SECURUS TECHNOLOGIES, INC.; T-NETIX, INC
Reel/Frame 030378/0433 →
RELEASE OF SECURITY INTEREST FIRST LIEN Recorded May 8, 2013
From: BNP PARIBAS
To: SECURUS TECHNOLOGIES, INC.; T-NETIX, INC
Reel/Frame 030378/0419 →
SECURITY AGREEMENT Recorded May 7, 2013
From: SECURUS TECHNOLOGIES, INC.; SECURUS TECHNOLOGIES HOLDINGS, INC.
To: DEUTSCHE BANK TRUST COMPANY AMERICAS, AS ADMINISTRATIVE AGENT
Reel/Frame 030369/0648 →
SECURITY AGREEMENT Recorded May 7, 2013
From: SECURUS TECHNOLOGIES, INC.; SECURUS TECHNOLOGIES HOLDINGS, INC.
To: DEUTSCHE BANK TRUST COMPANY AMERICAS, AS ADMINISTRATIVE AGENT
Reel/Frame 030369/0743 →
SECURITY AGREEMENT Recorded May 31, 2011
From: SECURUS TECHNOLOGIES, INC.; T-NETIX, INC.
To: BNP PARIBAS
Reel/Frame 026363/0253 →
SECURITY AGREEMENT Recorded May 31, 2011
From: SECURUS TECHNOLOGIES, INC.; T-NETIX, INC.
To: BNP
Reel/Frame 026363/0393 →
SUPPLEMENTAL NAME CHANGE RECORDING TO CONFIRM OWNERSHIP BY SECURUS TECHNOLOGIES, INC. Recorded May 24, 2011
From: SECURUS TECHNOLOGIES, INC.
To: SECURUS TECHNOLOGIES HOLDINGS, INC.
Reel/Frame 026328/0814 →
CHANGE OF NAME Recorded Nov 16, 2010
From: EVERCOM SYSTEMS, INC.
To: SECURUS TECHNOLOGIES, INC.
Reel/Frame 025372/0965 →
CHANGE OF NAME Recorded Nov 16, 2010
From: SECURUS TECHNOLOGIES, INC.
To: SECURUS TECHNOLOGIES HOLDINGS, INC.
Reel/Frame 025372/0983 →
ASSIGNMENT AND RELEASE OF SECURITY INTEREST Recorded May 4, 2010
From: WELLS FARGO FOOTHILL, LLC
To: T-NETIX, INC.; EVERCOM SYSTEMS, INC.; SECURUS TECHNOLOGIES, INC.
Reel/Frame 024329/0310 →
SECURITY AGREEMENT Recorded Oct 9, 2008
From: EVERCOM SYSTEMS, INC.
To: WELLS FARGO FOOTHILL, LLC
Reel/Frame 021651/0094 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jan 18, 2006
From: CHOW, CHARLES
To: EVERCOM SYSTEMS, INC.
Reel/Frame 017472/0165 →