Identifying clusters for service management operations

Patent metadata

Field	Value
Publication number	US-10685359-B2
Application number	US-201715723731-A
Country	US
Kind code	B2
Filing date	Oct 3, 2017
Priority date	May 5, 2017
Publication date	Jun 16, 2020
Grant date	Jun 16, 2020

Abstract

Client instance data including a plurality of incidents is obtained, each incident including a plurality of fields. A target field and an evaluation field are selected from among the plural fields. The plurality of incidents are grouped into a plurality of clusters based on a degree of a natural language text similarity of respective target fields in the plurality of incidents. A quality value is determined for each of the plurality of clusters based on the degree of the natural language text similarity of respective target fields in grouped incidents of the cluster from among the plurality of incidents, and based on respective evaluation fields. Each of the plurality of clusters is ranked based on the respective quality value of the cluster and a number of the grouped incidents of the cluster. At least one of the ranked plurality of clusters is identified to perform a service management operation.

First claim

What is claimed is:

1. A cluster identification method, comprising:
extracting historical incident report data for a plurality of incidents from one or more databases related to usage of client instances, wherein the plurality of incidents are closed incidents that have gone through an incident management life cycle;
grouping the plurality of incidents into a plurality of clusters based on a degree of natural language text similarity between respective target fields in the plurality of incidents having been completed in natural language;
ranking each cluster of the plurality of clusters based at least on a respective number of grouped incidents in each cluster;
identifying at least one ranked cluster of the plurality of clusters to perform a service management operation associated with the at least one ranked cluster; and
deflecting a future incident associated with the at least one ranked cluster of the plurality of clusters to a virtual agent for automated resolution using the service management operation.

2. The method of claim 1 , wherein each incident of the plurality of incidents includes a plurality of fields, and the method comprising:
selecting an evaluation field from among the plurality of fields, wherein the evaluation field has been completed with a selected value from a pre-determined set of values; and
determining, for each cluster of the plurality of clusters, a quality value based on the degree of natural language text similarity between the respective target fields in the grouped incidents of each cluster, and based on a similarity between respective evaluation fields in the grouped incidents of each cluster.

3. The method of claim 2 , wherein ranking the plurality of clusters comprises:
ranking each cluster of the plurality of clusters based on the respective number of grouped incidents in each cluster and a respective quality value of each cluster.

4. The method of claim 3 , wherein the target field is a first target field and the quality value is a first quality value, and wherein the method comprises:
selecting a second target field from among the plurality of fields, the second target field being different from the first target field; and
determining, for each cluster of the plurality of clusters, a second quality value based on a degree of natural language text similarity between respective second target fields in the grouped incidents of each cluster, and based on the similarity between the respective evaluation fields in the grouped incidents of each cluster;
wherein the ranking of each cluster of the plurality of clusters is based on the respective first and second quality values of each cluster and the respective number of grouped incidents of each cluster.

5. The method of claim 4 , wherein:
the first target field is an incident short description text field;
the second target field is an incident resolution detail text field or an incident comments and work notes or close notes text field; and
the evaluation field is a validation field including an incident category or an incident assignment group.

6. The method of claim 4 , wherein the degree of the natural language text similarity between the respective second target fields in the grouped incidents of each cluster is determined by:
breaking down a natural language text of the respective second target fields in the grouped incidents of the cluster into single words or one or more contiguous sequences of plural words; and
determining the degree of the natural language text similarity based on a comparison of the broken down natural language text across the respective second target fields in the grouped incidents of the cluster.

7. The method of claim 4 , comprising:
displaying a ranked list of the plurality of clusters by displaying, for each cluster, a description of the respective first and second target fields and the respective number of grouped incidents included in the cluster; and
displaying a pareto chart that is associated with the description of the respective first target fields of the plurality of clusters and that is based on the respective number of grouped incidents in each cluster.

8. The method of claim 1 , comprising:
determining respective quality values for each cluster of the plurality of clusters; and
determining whether or not to regroup the plurality of incidents into a different plurality of clusters based on the determined respective quality values.

9. The method of claim 1 , comprising:
creating issue-specific templates; and
populating template text and one or more field values with information from the respective target fields of the at least one ranked cluster of the plurality of clusters.

10. A non-transitory computer-readable recording medium having stored thereon a program for a computer of a cluster identification system, the program comprising instructions that when executed, cause the computer to:
extract historical incident report data for a plurality of incidents from one or more databases related to usage of client instances, wherein the plurality of incidents are closed incidents that have gone through an incident management life cycle;
group the plurality of incidents into a plurality of clusters based on a degree of natural language text similarity between respective target fields in the plurality of incidents having been completed in natural language;
rank each cluster of the plurality of clusters based at least on a respective number of grouped incidents in each cluster;
identify at least one ranked cluster of the plurality of clusters to perform a service management operation associated with the at least one ranked cluster; and
deflect future incidents associated with the at least one of the ranked plurality of clusters to a virtual agent for automated resolution.

11. The computer-readable recording medium according to claim 10 , wherein each incident of the plurality of incidents includes a plurality of fields, and wherein the program includes instructions that when executed, cause the computer to:
select an evaluation field from among the plurality of fields, wherein the evaluation field has been completed with a selected value from a pre-determined set of values; and
determine, for each cluster of the plurality of clusters, a quality value based on the degree of natural language text similarity between the respective target fields in the grouped incidents of each cluster, and based on a similarity between respective evaluation fields in the grouped incidents of each cluster.

12. The computer-readable recording medium according to claim 11 , wherein the program includes instructions that when executed, cause the computer to:
rank each cluster of the plurality of clusters based on the respective number of grouped incidents in each cluster and a respective quality value of each cluster.

13. The computer-readable recording medium according to claim 11 , wherein the target field is a first target field and the quality value is a first quality value, and wherein the program includes instructions that when executed, cause the computer to:
select a second target field from among the plurality of fields, the second target field being different from the first target field; and
determine, for each cluster of the plurality of clusters, a second quality value based on a degree of natural language text similarity between respective second target fields in the grouped incidents of each cluster, and based on the similarity between the respective evaluation fields in the grouped incidents of each cluster;
wherein the ranking of each cluster of the plurality of clusters is based on the respective first and

Assignees

• Servicenow Inc

Inventors

• Walthers Bruce
• Kulkarni Abhay
• Ramachandran Mukund
• Koohmarey Darius

Classifications

• G06F40/186

  Templates · CPC title

• G06F40/279

  Recognition of textual entities · CPC title

• G06Q30/016 Primary

  After-sales · CPC title

• G06F16/285

  Clustering or classification · CPC title

• G06F16/24578

  using ranking · CPC title