System and method for prioritizing contacts

• US 7,013,005 B2
• Filed: 02/11/2004
• Issued: 03/14/2006
• Est. Priority Date: 02/11/2004
• Status: Expired due to Fees

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Simske, Steven J., Yacoub, Sherif, Lin, Xiaofan
• Primary Examiner(s): TIEU, BENNY QUOC
• Application Number:
  US10/776,569
• Publication Number:
  US 20050175167A1
• Time in Patent Office: 762 Days
• Field of Search: 379 8801- 8804, 379/265.02, 379/265.06, 379/265.09, 379/265.12, 379/266.02, 704/10, 704/231, 704/246
• US Class Current: 379/265.13
• CPC Class Codes: H04M 2201/40 using speech recognition speech recognition per se G10L15/00
• CPC Class Codes: H04M 3/523 with call distribution or queueing

Abstract

A system and method for prioritizing contacts is disclosed. The method discloses: receiving a set of requests from a set of contacts to be connected to a recipient; generating a first priority level score for each request based on an acoustical analysis of a contact speech signal within the request; generating a second priority level score for each request based on a keyword analysis of the request; combining the priority level scores for each request into a combined priority level score; prioritizing the requests within the set of requests based on their respective combined priority level scores; and routing to the recipient that request, from the set of requests, having a highest combined priority level score. The system of the present invention includes all means and mediums for implementing the method.

Claims

1. A method for prioritizing, comprising:

receiving a set of requests from a set of contacts to be connected to a recipient;

generating a first priority level score for each request based on an acoustical analysis of a contact speech signal within the request;

generating a second priority level score for each request based on a keyword analysis of the request;

combining the priority level scores for each request into a combined priority level score;

prioritizing the requests within the set of requests based on their respective combined priority level scores; and

routing to the recipient that request, from the set of requests, having a highest combined priority level score.

2. The method of claim 1, wherein the recipient is an operator.

3. The method of claim 1, wherein receiving includes:

recording a message from a contact.

4. The method of claim 3, wherein recording includes:

recording speech signals from the contact while the contact is on-hold.

5. The method of claim 1, wherein:

the requests are messages; and

routing includes sending the operator a message having a highest combined priority level score.

6. The method of claim 5, wherein:

the message is an e-mail message.

7. The method of claim 1, wherein:

the requests are in a form of a set of contacts waiting on-hold; and

routing includes connecting the operator to an on-hold contact having a highest combined priority level score.

8. The method of claim 1, wherein generating the first score includes:

identifying acoustic features within the speech signal corresponding to a set of predefined emotions;

setting the first score to a higher value if the acoustic features correspond to a first sub-set of the predefined emotions; and

setting the first score to a lower value if the acoustic features correspond to a second sub-set of the predefined emotions.

9. The method of claim 8, wherein:

the predefined emotions include anger and frustration.

10. The method of claim 1, wherein generating the second score includes:

searching the request for a set of predefined keywords;

setting the second score to a higher value if the request includes a predefined number of a first predetermined subset of the keywords; and

setting the second score to a lower value if the request includes a predefined number of a second predetermined subset of the keywords.

11. The method of claim 10, wherein:

the predefined keywords include “

urgent”

.

12. The method of claim 1, wherein generating the second score includes:

searching the request for a set of predefined business relevance keywords;

setting the second score to a higher value if the request includes a predefined number of a first predetermined subset of the business relevance keywords; and

setting the second score to a lower value if the request includes a predefined number of a second predetermined subset of the business relevance keywords.

13. The method of claim 12, wherein:

the predefined business relevance keywords include;

a service contract.

14. The method of claim 1, further comprising:

generating a third priority level score for the request based on a contact-assigned priority level for the request.

15. The method of claim 14, wherein generating the third score includes:

searching the request for a contact-assigned priority level selected from a set of predefined priority levels; and

setting the third score equal to one of a set of predetermined priority level scores based upon the contact-assigned priority level.

16. The method of claim 15, wherein:

the predefined priority levels include;

“

high” and

“

low”

.

17. The method of claim 1, wherein combining includes:

weighting the confidence scores using ground truth data.

18. The method of claim 1, wherein;

the scores are statistical confidence scores; and

combining includes, multiplying together the priority level confidence scores.

19. The method of claim 1, wherein:

the scores are statistical confidence scores; and

combining includes, combining the priority level confidence scores according to the following formula;

$S=\underset{j=1}{\overset{N}{\sum <br><br>}}<br><br><br><br>r_j<br><br>p_j$

 

(where N is a total number of classifiers, r is a weight assigned to classifier j, and Pj is a confidence score generated by Classifier j).

20. The method of claim 1, wherein:

the scores are statistical confidence scores; and

combining includes, combining the confidence scores from each classifier for each demographic characteristic according to the following formula;

$S=\underset{j=1}{\overset{N}{\prod <br><br>}}<br><br><br><br>p_j^{r_j}$

 

(where N is a total number of classifiers, r is a weight assigned to classifier j, and Pj is a confidence score generated by Classifier j).

21. The method of claim 1, further comprising:

skipping the generating the second priority level score and the combining the priority level scores, for a new request, if the first priority level score for the new request is already greater than a current highest combined priority level score; and

setting the combined priority level score for the new request equal to the first priority level score.

22. The method of claim 1, further comprising:

identifying computational resources required to calculate each of the priority level scores;

generating priority level scores for a new request hierarchically beginning with a priority level score requiring a least amount of computational resources; and

stopping generation of the priority level scores once the combined priority level score for a new request exceeds a current highest combined priority level score.

23. The method of claim 1, wherein combining includes:

using a neural net to combine the priority level scores.

24. The method of claim 1, wherein receiving includes:

receiving a set of requests from a set of customers to be connected to an operator at a call center.

25. A method for prioritizing, comprising:

receiving a set of requests from a set of contacts to be connected to an operator;

generating a first priority level score for each request based on an acoustical analysis of a contact speech signal within each request;

identifying acoustic features within the speech signal corresponding to a set of predefined emotions;

setting the first score for a request to a higher value if the acoustic features correspond to a first sub-set of the predefined emotions;

setting the first score for the request to a lower value if the acoustic features correspond to a second sub-set of the predefined emotions;

generating a second priority level score for each request based on a keyword analysis of each request;

searching each request for a set of predefined keywords;

setting the second score for the request to a higher value if the request includes a predefined number of a first predetermined subset of the keywords;

setting the second score for the request to a lower value if the request includes a predefined number of a second predetermined subset of the keywords;

generating a third priority level score for each request based on a contact-assigned priority level to the request;

searching each request for a contact-assigned priority level selected from a set of predefined priority levels;

setting the third score for the request equal to one of a set of predetermined priority level scores based upon the contact-assigned priority level;

combining the priority level scores for each request into a combined priority level score;

prioritizing the requests within the set of requests based on their respective combined priority level scores; and

routing to the operator that request, from the set of requests, having a highest combined priority level score.

26. A computer-usable medium embodying computer program code for commanding a computer to prioritize, comprising:

receiving a set of requests from a set of contacts to be connected to a recipient;

generating a first priority level score for each request based on an acoustical analysis of a contact speech signal within the request;

generating a second priority level score for each request based on a keyword analysis of the request;

combining the priority level scores for each request into a combined priority level score;

prioritizing the requests within the set of requests based on their respective combined priority level scores; and

routing to the recipient that request, from the set of requests, having a highest combined priority level score.

27. A system for prioritizing, comprising a:

means for receiving a set of requests from a set of contacts to be connected to a recipient;

means for generating a first priority level score for each request based on an acoustical analysis of a contact speech signal within the request;

means for generating a second priority level score for each request based on a keyword analysis of the request;

means for combining the priority level scores for each request into a combined priority level score;

means for prioritizing the requests within the set of requests based on their respective combined priority level scores; and

means for routing to the recipient that request, from the set of requests, having a highest combined priority level score.