US 11,947,519 B2
Assigning an anomaly level to a non-instrumented object
Yuk L. Chan, Rochester, NY (US); Anuja Deedwaniya, Poughkeepsie, NY (US); and Robert M. Abrams, Wappinger Falls, NY (US)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Dec. 14, 2020, as Appl. No. 17/120,338.
Prior Publication US 2022/0188290 A1, Jun. 16, 2022
Int. Cl. G06F 16/23 (2019.01); G06F 21/55 (2013.01); G06N 5/022 (2023.01); G06N 20/00 (2019.01); H04L 9/40 (2022.01)
CPC G06F 16/2365 (2019.01) [G06N 20/00 (2019.01); H04L 63/1425 (2013.01)] 14 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
defining a key performance indicator associated with a non-instrumented object of a processing system, wherein the non-instrumented object comprises an object of the processing system that does not have log data or metrics available for analysis;
determining, by a processing device, a current anomaly level for each of a plurality of key performance indicators for an instrumented object having a relationship with the non-instrumented object, wherein the instrumented object comprises another object of the processing system that has at least one of log data and metrics available for analysis;
assigning, by the processing device, an anomaly level to the non-instrumented object based on the current anomaly levels of the instrumented object, the anomaly level comprising a weighted function of the respective anomaly level of each of the plurality of key performance indicators, wherein each anomaly level is uniquely weighted;
for each type of non-instrumented object, normalizing the anomaly level for the respective type of non-instrumented object to a respective fixed scale, thus preventing key performance indicator-specific characteristics from influencing rankings of anomaly levels across non-instrumented components;
calculating, by the processing device, a confidence score for the anomaly level based at least in part on the relationship between the non-instrumented object and the instrumented object and based at least in part on a relationship between the non-instrumented object and other instrumented objects, wherein the confidence score is further based at least in part on a number of the relationships between the non-instrumented object and other instrumented objects; and
assigning the confidence score to the anomaly level assigned to the non-instrumented object.