Detection and prevention of inmate to inmate message relay

• US 10,572,961 B2
• Filed: 03/15/2016
• Issued: 02/25/2020
• Est. Priority Date: 03/15/2016
• Status: Active Grant

• Current Assignee: Global Tel-Link Corporation
• Original Assignee: Global Tel-Link Corporation
• Inventors: Hodge, Stephen L.
• Primary Examiner(s): Parry, Chris
• Assistant Examiner(s): Abderrahmen, Chouat
• Application Number:
  US15/071,082
• Publication Number:
  US 20170270627A1
• Time in Patent Office: 1,442 Days
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/258 Data format conversion from or to a database
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 40/242 Dictionaries
• CPC Class Codes: G06Q 50/26 Government or public services business processes related to the transportation industry G06Q50/40
• CPC Class Codes: G10L 15/005 Language recognition
• CPC Class Codes: G10L 15/02 Feature extraction for speech recognition; Selection of recognition unit
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 2015/088 Word spotting
• CPC Class Codes: G10L 25/51 for comparison or discrimination
• CPC Class Codes: H04L 43/16 Threshold monitoring
• CPC Class Codes: H04L 51/046 Interoperability with other network applications or services
• CPC Class Codes: H04M 2201/18 Comparators
• CPC Class Codes: H04M 2201/40 using speech recognition
• CPC Class Codes: H04M 2203/2061 Language aspects
• 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 3/38 Graded-service arrangements, i.e. some subscribers prevented from establishing certain connections queuing arrangements H04Q3/64
• CPC Class Codes: H04M 3/42221 Conversation recording systems at the subscriber's set H04M1/656

Abstract

Secure system and method of detecting and preventing inmate to inmate message relays. A system and method which monitors inmate communications for similar phrases that occur as part of two or more separate inmate messages. These similar phrases may be overlapping in real time as in a conference call or can occur at separate times in separate messages. The communications that appear similar are assigned a score and the score is compared to a threshold. If the score is above a certain threshold, the communication is flagged and remedial actions are taken. If the flagged communication contains illegal matter then the communication can be disconnected or restricted in the future.

Claims

1. A communication monitoring method, comprising:

receiving a communication involving an inmate of a controlled-environment facility, the received communication including audio data of a recorded conversation and metadata;

extracting the metadata from the received communication, the metadata describing non-content information about the received communication and/or the conversation and including at least one timestamp;

comparing the extracted metadata to metadata of a plurality of previously-stored communications;

identifying a previously-stored second communication, from among the plurality of previously-stored communications that has metadata that is similar to the extracted metadata to within a predetermined degree;

retrieving the previously-stored second communication;

correlating spoken content of the received communication with spoken content of the previously-stored second communication, the correlating including;

first comparing spoken content included in the audio data of the received communication to spoken content of the previously-stored second communication;

second comparing non-spoken content included in the audio data of the received communication to non-spoken content of the previously-stored second communication;

identifying a plurality of shared words and/or phrases that were spoken in both the received communication and the previously-stored second communication;

for each of the plurality of shared words and/or phrases, determining a relevancy score indicative of a uniqueness and/or importance of the shared word and/or phrase; and

determining a frequency of each of the shared words and/or phrases that appear in both the received communication and the previously-stored second communication, each frequency indicating a number of times the corresponding shared word and/or phrase appears in both the received communication and the previously-stored second communication;

assigning a score to the received communication based on the correlation, the frequencies, and the relevancy scores of the shared words and/or phrases;

comparing the score with a predetermined threshold; and

determining the received communication to be an inmate-to-inmate communication based on the score comparison.

2. The communication monitoring method of claim 1, wherein the received communication is in an audio format, the communication monitoring method further comprising:

storing the received communication in the audio format; and

transcribing the received communication into a text format, wherein the text format reflects contents of the received communication.

3. The communication monitoring method of claim 1, further comprising:

determining that the received communication contains non-standard language; and

converting non-standard language portions of the received communication into a standard language, the converting including;

accessing a language dictionary to identify the non-standard language; and

translating the non-standard language into the standard language using the language dictionary and a plurality of conversion protocols.

4. The communication monitoring method of claim 1, further comprising:

determining that the received communication is a real-time communication;

detecting keywords and/or key phrases within the received communication in real-time; and

storing a recording of the received communication for performing the correlating at a later time.

5. The communication monitoring method of claim 1, further comprising:

determining that the received communication is not a real-time communication;

identifying a stored second communication having overlapping metadata with the communication in response to the determining; and

correlating contents of the received communication with contents of the identified stored second communication.

6. The communication monitoring method of claim 4, further comprising, in response to the determining, taking a remedial action, wherein the remedial action includes alerting a supervising personnel and disconnecting the communication.

7. The communication monitoring method of claim 5, further comprising, in response to the determining, taking a remedial action, wherein the remedial action includes withholding the received communication from being forwarded to an intended recipient until after the correlating.

8. A communication monitoring method, comprising:

receiving a first communication that involves an inmate of a controlled-environment facility, the first communication including audio data of a recorded conversation and metadata;

extracting the metadata from the first communication, the metadata describing non-content information about the first communication and/or the recorded conversation including at least one timestamp;

identifying keywords and/or keyphrases in the first communication;

comparing the extracted metadata to metadata of a plurality of previously-stored communications;

identifying a previously-stored second communication, from among the plurality of previously-stored communications, that has metadata that is similar to the extracted metadata to within a predetermined degree;

searching the previously-stored second communication for instances of the keywords and/or keyphrases identified in the first communication to determine a plurality of shared words and/or phrases;

determining, for each of the plurality of shared words and/or phrases, a relevancy score indicative of a uniqueness and/or importance of the corresponding shared word or phrase;

determining a frequency of appearance of each of the plurality of shared words and/or phrases that appear in both the first communication and the previously-stored second communication, each frequency indicating a number of times the corresponding shared word or phrase appears in both the first communication and the previously-stored second communication;

scoring the first communication based on the frequencies of appearance and the relevancy scores of the shared words and/or phrases;

comparing the score with a predetermined threshold; and

identifying the first communication as an inmate-to-inmate communication based on the score comparing.

9. The communication monitoring method of claim 8, further comprising:

determining that the first communication is a real-time communication;

forwarding the first communication to an intended recipient prior to completing the comparing; and

storing a copy of the first communication, wherein the comparing is performed using the stored copy of the first communication.

10. The communication monitoring method of claim 9, further comprising correlating contents of the first communication with contents of the previously-stored second communication, wherein the scoring of the first communication is based on the correlation.

11. The communication monitoring method of claim 8, further comprising:

determining that the first communication is not in real time;

correlating contents of the first communication with contents of the previously-scored second communication in response to the determination before forwarding the first communication to an intended recipient; and

scoring the first communication based on the correlation.

12. A communication monitoring system, the system comprising:

a database;

a transceiver, configured to receive a communication involving an inmate of a controlled-environment facility, the communication including audio data of a recorded conversation and metadata;

a processing subsystem configured to extract the metadata from the communication, the metadata describing non-content information about the communication and/or the conversation and including at least one timestamp; and

a detection and prevention analysis engine configured to;

compare the extracted metadata to metadata of a plurality of stored communications;

identify a stored communication from among the plurality of stored communications that has metadata that is similar to the extracted metadata to within a predetermined degree;

correlate spoken content of the communication with spoken content of the stored communication, the correlating including;

first comparing the spoken content included in the audio data of the communication to spoken content of the stored communication;

second comparing non-spoken content included in the audio data of the communication to non-spoken content included in the stored communication;

identifying a plurality of shared words and/or phrases that were spoken in both the communication and the stored communication;

for each of the plurality of shared words and/or phrases, determine a relevancy score indicative of a uniqueness and/or importance of the shared word and/or phrase; and

determining a frequency of each of the shared words and/or phrases that appear in both the communication and the stored communication, each frequency indicating a number of times the corresponding shared word and/or phrase appears in both the communication and the second communication;

assign a score to the communication based on the correlation, the frequencies, and the relevancy scores of the shared words and/or phrases;

compare the score with a predetermined threshold; and

determine that the communication is an inmate-to-inmate communication based on the score comparison.

13. The communication monitoring system of claim 12, further comprising a communication storage configured to store the communication in an appropriate format depending on a type of the communication.

14. The communication monitoring system of claim 12, further comprising a transcriber configured to transcribe the communication into a text format, wherein the text format reflects contents of the communication.

15. The communication monitoring system of claim 12, further comprising a language converter configured to:

determine that the communication contains non-standard language; and

convert the communication into a standard language, wherein to convert the communication, the language converter is configured to;

access a language dictionary to identify the non-standard language; and

translate nonstandard language portions of the communication into the standard language using the language dictionary and a plurality of conversion protocols.

16. The communication monitoring system of claim 12, wherein the detection and prevention analysis engine is further configured to:

determine that the communication is in real time; and

forward the communication to an intended recipient prior to performing the correlating.

17. The communication monitoring system of claim 12, wherein the detection and prevention analysis engine is further configured to:

determine that the communication is not in real time; and

perform the correlating prior to forwarding the communication to an intended recipient.

18. The communication monitoring system of claim 16, wherein the detection and prevention analysis engine is further configured to take remedial action in response to the determining that the communication is an inmate-to-inmate communication, wherein the remedial action includes alerting a supervising personnel and disconnecting the communication.

19. The communication monitoring system of claim 17, wherein the detection and prevention analysis engine is further configured to take remedial action in response to the determining that the communication is an inmate-to-inmate communication, wherein the remedial action includes withholding the communication and preventing future communication containing the same metadata.