Task assignment using ranking support vector machines

• US 9,466,039 B2
• Filed: 12/08/2014
• Issued: 10/11/2016
• Est. Priority Date: 10/26/2011
• Status: Expired due to Fees

• Current Assignee: Kyndryl Incorporated
• Original Assignee: International Business Machines Corporation
• Inventors: Gatti, Maira Athanazio de Cerqueira, Herrmann, Ricardo Guimaraes, Loewenstern, David, Pinel, Florian, Shwartz, Larisa
• Primary Examiner(s): NGUYEN, VAN H
• Application Number:
  US14/563,629
• Publication Number:
  US 20150095080A1
• Time in Patent Office: 673 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 20/10 using kernel methods, e.g. support vector machines [SVM]
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 10/063112 Skill-based matching of a person or a group to a task
• CPC Class Codes: H04L 41/5074 Handling of user complaints or trouble tickets

Abstract

A method of ranking workers for an incoming task includes recording a list of completed tasks in a computer data structure, extracting first attributes from the list for the tasks that were completed during a pre-determined period, generating a first feature vector for each task and worker from the first extracted attributes, training a Support Vector Machine (SVM) based on the feature vector to output a weight vector, extracting second attributes from an incoming task, generating a second feature vector for each worker based on the second extracted attributes, and ranking the workers using the second feature vectors and the weight vector. The first attributes may be updated during a subsequent period to re-train the SVM on updated first feature vectors to generate an updated weight vector. The workers may be re-ranked based on the second feature vectors and the updated weight vector. Accordingly, the feature vectors are dynamic.

Claims

1. A method of dispatching an incoming task, the method comprising:

extracting, by a processor, first attributes from a list of tasks that were completed during a pre-determined period by a plurality of workers;

generating, by a processor, a first feature vector for each task and worker from the first extracted attributes;

training, by a processor, a Support Vector Machine (SVM) based on the first feature vectors to output a weight vector;

extracting, by a processor, second attributes from an incoming task;

generating, by a processor, a second feature vector based on the second extracted attributes;

ranking, by the processor, the workers for performing the incoming task using the second feature vector and the weight vector, wherein each first feature vector indicates a percentage of each workers'"'"' workload spent on the completed tasks having same attributes as the incoming task during the pre-determined period; and

dispatching the incoming task to a selected worker among the workers with a highest rank determined by the ranking.

2. The method of claim 1, wherein at least one entry in the list comprises a worker identifier, at least one of the first attributes, and a completion time.

3. The method of claim 2, wherein one of the first attributes is a task outcome.

4. The method of claim 2, wherein the task outcome indicates completed with a breach of a Service Level Agreement (SLA) or completed without a breach of the SLA.

5. The method of claim 2, wherein the task outcome indicates completed on-time or completed late.

6. The method of claim 2, wherein the task outcomes indicates reassignment of the corresponding task to one of the workers from another one of the workers.

7. The method of claim 1, further comprising:

generating a score for each first feature vector based on whether the corresponding task was assigned to the corresponding worker; and

training the SVM based additionally on the scores.

8. The method of claim 1, wherein the attributes include at least one of (i) a priority of the task, (ii) a level of complexity of the task, (iii) a task type, (iv) a customer identifier, or a (v) description of the task.

9. The method of claim 1, further comprising:

extracting third attributes from the list for the tasks that were completed during a subsequent second pre-determined period;

generating a third feature vector from the third extracted attributes;

training the SVM based on the feature vector and the third feature vector to output a second weight vector; and

re-ranking the workers using the second feature vectors and the second weight vector.

10. The method of claim 9, further comprising:

generating a score for each first feature vector based on whether the corresponding task was assigned to the corresponding worker; and

training the SVM based additionally on the scores.

11. The method of claim 9, wherein the attributes include at least one of (i) a priority of the task, (ii) a level of complexity of the task, (iii) a task type, (iv) a customer identifier, or a (v) description of the task.

12. The method of claim 9, further comprising:

extracting third attributes from the list for the tasks that were completed during a subsequent second pre-determined period;

generating a third feature vector from the third extracted attributes;

training the SVM based on the feature vector and the third feature vector to output a second weight vector; and

re-ranking the workers using the second feature vectors and the second weight vector.

13. The method of claim 1, wherein the incoming task is a request for a computer related service task and the dispatching comprises a request fulfillment process forwarding the incoming task to the selected worker.

14. A method of dispatching an incoming task, the method comprising:

extracting, by a processor, first attributes from a list of tasks that were completed during a pre-determined period by a plurality of workers;

generating, by a processor, a first feature vector for each task and worker from the first extracted attributes;

training, by a processor, a Support Vector Machine (SVM) based on the first feature vectors to output a weight vector;

extracting, by a processor, second attributes from an incoming task;

generating, by a processor, a second feature vector based on the second extracted attributes;

ranking, by the processor, the workers for performing the incoming task using the second feature vector and the weight vector, wherein each first feature vector indicates a percentage of the tasks having same attributes as the incoming task that were assigned to a same one of the workers during the pre-determined period; and

dispatching the incoming task to a selected worker among the workers with a highest rank determined by the ranking.

15. The method of claim 14, wherein at least one entry in the list comprises a worker identifier, at least one of the first attributes, and a completion time.

16. The method of claim 15, wherein one of the first attributes is a task outcome.

17. The method of claim 15, wherein the task outcome indicates completed with a breach of a Service Level Agreement (SLA) or completed without a breach of the SLA.

18. The method of claim 15, wherein the task outcome indicates completed on-time or completed late.

19. The method of claim 15, wherein the task outcomes indicates reassignment of the corresponding task to one of the workers from another one of the workers.

20. The method of claim 14, wherein the incoming task is a request for a computer related service task and the dispatching comprises a request fulfillment process forwarding the incoming task to the selected worker.