Issue resolution utilizing feature mapping
AssigneeInternational Business Machines Corporation
InventorsGenady Grabarnik, Tao Li, Larisa Shwartz, Wubai Zhou
Life of this patent
Estimating the term…
Abstract
A method for providing a resolution to an open issue is described. The method includes utilizing structured corresponding learning to determine pivots from a database having historical data records to determine data records having similar open issues and providing their resolution. The method further includes providing a recommended resolution for the open issue based on the historical data records.
Claims20
A computer-implemented method for providing a resolution, the method comprising:
- providing, using a processing system, an original request having a criteria and requiring resolution;
- accessing, using a processing system, a database having a plurality of request records each including a request component and a resolution component;
- separating, using the processing system, the database into a source domain and a target domain, the source domain comprising request records created prior to a predetermined time period, the target domain comprising request records created after the predetermined time period;
- analyzing, using the processing system, the request records of the source domain and target domain to establish a combined request domain including data records having related requests and resolutions, the analysis including utilizing structured corresponding learning to define pivots which are utilized to determine data records having related requests and resolutions;
- analyzing, using the processing system, the original request and comparing the original request to data records of the combined request domain to determine similar data records and related resolutions;
- and
- providing a resolution recommendation for resolving the request of the original ticket.
A system for providing a resolution for a computer system ticket, the system comprising:
- a memory;
- and
- a processing system communicatively coupled to the memory,
- wherein the memory is configured to store a plurality of ticket records each having a symptom component and a resolution component;
- wherein the processing system is configured to divide the ticket records into a source domain and a target domain, and wherein the processing system utilizes structural correspondence learning to identify pivots associated with the ticket records for determining related ticket records based on either the symptom or resolution component and associating the related tickets in a common ticket database;
- and
- wherein the processing system is configured to analyze an original ticket record having a symptom and recommend a resolution based on a comparison with ticket records included in the common ticket database wherein a ticket record in the common ticket database has a symptom related to the symptom of the original ticket record and the recommended resolution is the resolution component of the ticket record in the common ticket database.
A computer program product for providing a resolution for a ticket, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, wherein the computer readable storage medium is not a transitory signal per se, the program instructions readable by a processing circuit to cause the processing circuit to perform a method comprising:
- providing an original ticket having a symptom requiring resolution;
- accessing a historical ticket database having a plurality of ticket each including a symptom component and a resolution component;
- separating the database into a source domain and a target domain, the source domain comprising ticket records created prior to a predetermined time period, the target domain comprising ticket records created after the predetermined time period;
- analyzing the ticket records of the source domain and target domain to establish a combined request domain including ticket records having related symptoms and resolutions, the analysis including utilizing structural correspondence learning to define pivots which are utilized to determine ticket records having related requests and resolutions;
- analyzing the original ticket and comparing the original ticket to ticket records of the combined request domain to determine similar ticket records and related resolutions;
- and
- providing a resolution recommendation for resolving the request of the original ticket.
Dependencies are inferred from the claim text.