Predictive dialing system and method

• US 20040170258A1
• Filed: 03/26/2004
• Published: 09/02/2004
• Est. Priority Date: 09/26/2001
• Status: Abandoned Application

• Current Assignee: HRJ Development Corporation
• Original Assignee: HRJ Development Corporation
• Inventors: Pathak, Sonal R, Levin, Mitch, Currie, Edward H, Pathak, Amarish M
• Application Number:
  US10/491,200
• Publication Number:
  US 20040170258A1
• Time in Patent Office: Days
• US Class Current: 379/88.01
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: H04M 1/2749 Automatic dialling or redialling systems, e.g. on off-hook or redial on busy
• CPC Class Codes: H04M 2201/40 using speech recognition speech recognition per se G10L15/00
• CPC Class Codes: H04M 3/42323 PBX's with CTI arrangements
• CPC Class Codes: H04M 3/424 Arrangements for automatic redialling at the subscriber's set H04M1/27
• CPC Class Codes: H04M 3/46 Arrangements for calling a number of substations in a predetermined sequence until an answer is obtained
• CPC Class Codes: H04M 3/465 Arrangements for simultaneously calling a number of substations until an answer is obtained
• CPC Class Codes: H04M 3/48 Arrangements for recalling a calling subscriber when the wanted subscriber ceases to be busy
• CPC Class Codes: H04M 3/5158 in combination with automated outdialling systems devices for signalling identity of wanted subscriber H04M1/26

Abstract

Predictive dialing system (100) and method is disclosed. Telephone numbers from a call list (FIG. 2) are automatically dialed and response received is analyzed. Message content in the response is detected using a speech recognizer and one or more follow-up actions based on the message content are performed. The follow-up actions may include rescheduling the call if the telephone number dialed is busy, not answering, or otherwise not working. Another follow-up action may include dynamically updating the list of telephone numbers in the call list, for example, if an operator message indicating that the phone number is no longer in service or that phone number has changed to a new number. Yet another follow-up action may include transferring the call to a live agent, for example, if a live greeting is received.

Claims

1. A method for predictive dialing, comprising:

receiving a telephone number to dial;

automatically dialing the telephone number; and

analyzing a response tone from a receiver side, including detecting whether speech is in the response tone, and determining message content of the speech if speech is detected.

2. The method of claim 1, further including:

determining one or more follow-up actions based on the analyzing.

3. The method of claim 2, wherein the one or more follow-up actions include:

determining whether the telephone number dialed is a valid number based on the analyzing of the response tone.

4. The method of claim 3, further including:

dynamically deleting the telephone number from a list of telephone numbers if the dialed telephone number is determined to be not valid.

5. The method of claim 3, wherein the telephone number is determined to be not valid, if the message content indicates that the telephone number has been disconnected.

6. The method of claim 3, wherein the telephone number is determined to be not valid, if the message content indicates that the telephone number has been changed.

7. The method of claim 1, wherein the analyzing step further includes using a statistical speech recognition model for determining the message content.

8. The method of claim 7, wherein the statistical speech recognition model includes a hidden Markov model.

9. The method of claim 1, wherein the analyzing step further includes using a plurality of speech recognition models, and the determining message content includes determining message content based on a result of a speech recognition model having a highest confidence level among the plurality of speech recognition models.

10. The method of claim 9, wherein the highest confidence level is determined as a function of one or more properties of the plurality of speech recognition models and one or more qualities of the response tone.

11. The method of claim 10, wherein the one or more properties of the plurality of speech recognition models include an ability to detect low audio quality and the one or more qualities of the response tone include a low audio quality.

12. The method of claim 9, the method further including:

determining one or more properties of the plurality of speech recognition models; and

determining one or more qualities of the response tone, to calculate the highest confidence level.

13. The method of claim 7, wherein the statistical speech recognition model includes audio data signature recognition.

14. The method of claim 7, wherein the statistical speech recognition model includes spectral analysis of voice data.

15. The method of claim 1, wherein the analyzing step further includes:

dividing the response tone into one or more sets of frequencies to separate into one or more components of sound; and

determining if a known pattern exists in the sets of frequencies.

16. The method of claim 15, further including:

computing a score based on whether a known pattern is recognized within the sets of frequencies.

17. The method of claim 1, further including:

echoing the message content, if it is determined that the message content includes a live greeting voice.

18. A system for predictive dialing, comprising:

one or more low level modules for dialing telephone numbers and receiving response signals from the dialed telephone numbers, the low level modules further recording the received response signals;

a speech recognition module for determining content of the received response signals; and

a high level module for providing the telephone numbers to the low level modules to dial, the high level module farther receiving the content of the received response signals and determining one or more actions to perform based on the content.

19. The system of claim 18, wherein the speech recognition module includes a signal processing module that divides the response tone into one or more sets of frequencies and separates components of sound from the one or more sets of frequencies to determine if one or more known patterns exist within the sets of frequencies, and computes a score based on whether one or more known patterns exist.

20. The system of claim 18, wherein the high level module and the one or more low level modules runs on multiple machines in a cluster-based computing model.

21. A program storage device readable by a machine, tangibly embodying a program of instructions executable by the machine to perform method steps of predictive dialing, comprising:

receiving a telephone number to dial;

automatically dialing the telephone number; and

analyzing a response tone from a receiver side, including detecting whether speech is in the response tone, and determining message content of the speech if speech is detected.

22. A method for predictive dialing, comprising:

receiving a list of telephone number to perform automatic dialing;

automatically dialing a telephone number in the list;

listening to a response tone from the telephone number dialed;

separating the response tone into one or more sets of frequencies;

extracting one or more sound components from the one or more sets of frequencies;

detecting a known pattern in the one or more sound components to recognize contents of the response tone; and

responding according to the recognized contents.