| CPC G06Q 30/0631 (2013.01) [G06N 5/022 (2013.01); G06N 5/025 (2013.01); G06N 20/00 (2019.01); G06Q 10/0637 (2013.01)] | 18 Claims |
|
1. A non-transitory computer-readable medium to store instructions, which when executed by a computer, cause the computer to perform operations comprising:
receiving, via a graphical user interface (GUI), a first request to perform an exploration in a predictive and maintenance service application;
in response to the receiving the first request:
creating the exploration;
recording a series of actions performed during the exploration in a storage using a processor of the computer, wherein the recorded series of actions performed during the exploration tracks at least the actions performed, inputs provided for analysis, and micro services selected to perform root cause analysis of an issue during the exploration; and
saving the exploration in an evidence package in the storage, wherein the evidence package comprises a collection of explorations including the current exploration;
determining a predicted list of recommended micro services for the exploration using the processor that executes a machine learning algorithm to perform an analysis of the evidence package comprising the collection of explorations in the storage;
receiving a second request to perform the exploration in the predictive and maintenance service application, the second request occurring after the first request;
in response to the receiving the second request, in a recommendation engine, automatically identifying a list of micro services based on the determined predicted list of recommended micro services for the exploration, wherein automatically identifying the list of micro services further comprises:
learning, by the machine learning algorithm, a corresponding model as a corresponding learned model by computing a corresponding distance of the evidence package with each successful evidence package of successful evidence packages previously saved in time series storage; and
ranking each corresponding learned model of the learned models based on the computed corresponding distance, wherein the identified list of micro services further includes individual micro services in one or more successful evidence packages that correspond to the learned model of the learned models having the highest rank above a pre-defined threshold; and
displaying, via the GUI, the list of micro services as recommendations for performing the exploration in the predictive and maintenance service application.
|