SYSTEM AND METHOD FOR ANALYZING AND CLASSIFYING CALLS WITHOUT TRANSCRIPTION

• US 20150365530A1
• Filed: 08/24/2015
• Published: 12/17/2015
• Est. Priority Date: 03/15/2013
• Status: Active Grant

• Current Assignee: Marchex Incorporated
• Original Assignee: Marchex Incorporated
• Inventors: Kolbegger, Chris, Flaks, Jason, Graber, Tim, Rieger, Bryan, Ismail, Ziad, Ramanathan, Govindaraj, Spehr, Darren, Berk, Matthew
• Granted Patent: US 9,614,962 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G10L 25/51 for comparison or discrimination
• CPC Class Codes: G10L 25/78 Detection of presence or absence of voice signals switching of direction of transmission by voice frequency in two-way loud-speaking telephone systems H04M9/10
• CPC Class Codes: G10L 25/93 Discriminating between voiced and unvoiced parts of speech signals G10L25/90 takes precedence
• CPC Class Codes: H04L 41/5067 Customer-centric QoS measurements
• CPC Class Codes: H04L 43/0852 Delays
• CPC Class Codes: H04M 2203/559 Sorting systems
• 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 3/42221 Conversation recording systems at the subscriber's set H04M1/656
• CPC Class Codes: H04M 3/5166 in combination with interactive voice response systems or voice portals, e.g. as front-ends
• CPC Class Codes: H04M 3/5175 Call or contact centers supervision arrangements

Abstract

A facility and method for analyzing and classifying calls without transcription. The facility analyzes individual frames of an audio to identify speech and measure the amount of time spent in speech for each channel (e.g., caller channel, agent channel). Additional telephony metrics such as R-factor or MOS score and other metadata may be factored in as audio analysis inputs. The facility then analyzes the frames together as a whole and formulates a clustered-frame representation of a conversation to further identify dialogue patterns and characterize call classification. Based on the data in the clustered-frame representation, the facility is able to make estimations of call classification. The correlation of dialogue patterns to call classification may be utilized to develop targeted solutions for call classification issues, target certain advertising channels over others, evaluate advertising placements at scale, score callers, and to identify spammers.

Claims

1. A computing system for analyzing a telephone call without transcription, the computing system comprising:

a memory for storing programmed instructions;

a processor configured to execute the programmed instructions to perform operations including;

receiving an audio signal of a telephone call to analyze, wherein the system does not transcribe the audio signal of the telephone call and wherein the audio signal comprises a first channel and a second channel;

monitoring a first channel time-in-activity by tracking a presence of activity on the first channel;

monitoring a second channel time-in-activity by tracking a presence of activity on the second channel; and

characterizing a portion of the telephone call based at least in part on the first channel time-in-activity and the second channel time-in-activity during the portion of the telephone call, wherein the activity comprises at least one of speech activity and non-speech activity.

2. The computing system of claim 1 wherein the operations further include:

depicting each of the first and second channels as a graphical waveform or clustered-frame sequence, wherein the graphical waveforms or clustered-frame sequences chart the first channel time-in-activity and the second channel time-in-activity as a function of time for the length of the telephone call.

3. The computing system of claim 2 wherein the operations further include:

identifying a pattern in the graphical waveforms or clustered-frame sequences, wherein identifying a pattern includes;

identifying a beginning, a middle, and an end of the telephone call;

evaluating an interaction between the first channel and the second channel, wherein evaluating includes considering a rate of interchange between the first channel time-in-activity and the second channel time-in-activity; and

identifying whether the interaction occurs at the beginning, middle, or end of the telephone call.

4. The computing system of claim 3 wherein the operations further include correlating the pattern in the graphical waveforms or clustered-frame sequences to a specific outcome of the telephone call to determine a call classification of the telephone call.

5. The computing system of claim 4 wherein the operations further include factoring in R-factor and MOS score metrics to determine a call classification of the telephone call.

6. The computing system of claim 1 wherein the portion of the telephone call is categorized according to a plurality of categorization options, the categorization options comprising:

a first channel monologue when there is predetermined amount of first channel time-in-activity that corresponds to speech and a relatively insubstantial second channel time-in-activity that corresponds to speech;

a conversation when there is a predetermined rate of interchange between the first channel time-in-activity that corresponds to speech and the second channel time-in-activity that corresponds to speech; and

a collision when there is coinciding first channel time-in-activity and the second channel time-in-activity.

7. The computing system of claim 1 wherein the portion of the telephone call is categorized according to a plurality of categorization options, the categorization options comprising:

an Interactive Voice Response, a wrong number, or an exchange of information.

8. The computing system of claim 1 wherein the audio signal of the telephone call is a live telephone call.

9. The computing system of claim 1 wherein the audio signal of the telephone call is a pre-recorded file.

10. The computing system of claim 1 wherein the operations further include applying:

at least one filter to the audio signal, wherein each of the at least one filter attenuates signals with frequencies above or below a predetermined cutoff frequency.

11. The computing system of claim 1 wherein the first and second channels comprise a client channel and an agent channel.

12. The computing system of claim 1 wherein the length of the telephone call comprises a plurality of frames, and the monitoring of the first channel time-in-activity and the second channel time-in-activity are each done in each frame of the plurality of frames.

13. The computing system of claim 1 wherein the non-speech activity is a ring tone, a dual-tone multi-frequency tone (DTMF), or music.

14. A method in a computing system for analyzing a telephone call without transcribing or saving to disk, the method comprising:

identifying a telephone call to monitor;

receiving an audio signal of the telephone call, wherein the audio signal of the telephone call that is received is not transcribed, and wherein the audio signal comprises a first channel and a second channel;

monitoring a first channel time-in-activity by tracking on the first channel when there is activity;

monitoring a second channel time-in-activity by tracking on the second channel when there is activity; and

characterizing a portion of the telephone call based at least in part on the first channel time-in-activity and the second channel time-in-activity during the portion of the telephone call, wherein the activity comprises at least one of speech activity and non-speech activity.

15. The method of claim 14 wherein the portion of the telephone call is categorized according to a plurality of categorization options, the categorization options comprising:

a first channel monologue when there is a predetermined amount of first channel time-in-activity corresponding to speech and a relatively insubstantial amount of second channel time-in-activity corresponding to speech;

a conversation when there is a predetermined rate of interchange between the first channel time-in-activity corresponding to speech and the second channel time-in-activity corresponding to speech; and

a collision when there is coinciding first channel time-in-activity and second channel time-in-activity.

16. The method of claim 14 wherein the portion of the telephone call is categorized according to a plurality of categorization options, the categorization options comprising:

an Interactive Voice Response when a predetermined amount of second channel time-in-activity corresponding to speech and a relatively insubstantial first channel time-in-activity corresponding to speech at the beginning of the telephone call are identified;

a wrong number when a predetermined amount of first channel time-in-activity and a relatively insubstantial second channel time-in-activity at the beginning of the telephone call are identified;

an exchange of first channel information when predetermined rate of interchange between the first channel time-in-activity corresponding to speech and the second channel time-in-activity corresponding to speech at the beginning of the telephone call are identified;

a reading of terms and conditions or a description of products or services when a predetermined amount of second channel time-in-activity corresponding to speech in the middle of the telephone call is identified; and

an end-of-call thrash when a predetermined rate of interchange between the first channel time-in-activity corresponding to speech and the second channel time-in-activity corresponding to speech at the end of the telephone call is identified.

17. The method of claim 14 further comprising:

depicting each of the first and second channels as a graphical waveform or clustered-frame sequence, wherein the graphical waveforms or clustered-frame sequences chart the first channel time-in-activity and the second channel time-in-activity as a function of time for the length of the telephone call.

18. The method of claim 17 further comprising:

identifying a pattern in the graphical waveforms or clustered-frame sequences, wherein identifying a pattern includes;

identifying a beginning, a middle, and an end of the telephone call;

evaluating an interaction between the first channel and the second channel, wherein evaluating includes considering a rate of interchange between the first channel time-in-activity and the second channel time-in-activity; and

identifying whether the interaction occurs at the beginning, middle, or end of the telephone call.

19. The method of claim 14 further comprising correlating the pattern in the graphical waveforms or clustered-frame sequences to a specific outcome of the telephone call to determine a call classification of the telephone call.

20. The method of claim 19 further comprising factoring in R-factor and MOS score metrics to determine the call classification of the telephone call.