US 12,242,524 B2
Method and system for performing a search in a dataset in a mainframe session
Manas Kumar Behera, Hyderabad (IN); and Velkumar Sashikumar, Salem (IN)
Assigned to JPMORGAN CHASE BANK, N.A., New York, NY (US)
Filed by JPMorgan Chase Bank, N.A., New York, NY (US)
Filed on May 1, 2023, as Appl. No. 18/141,708.
Claims priority of application No. 202311005504 (IN), filed on Jan. 27, 2023.
Prior Publication US 2024/0256585 A1, Aug. 1, 2024
Int. Cl. G06F 16/00 (2019.01); G06F 16/3332 (2025.01); G06F 16/338 (2019.01)
CPC G06F 16/3334 (2019.01) [G06F 16/338 (2019.01)] 12 Claims
OG exemplary drawing
 
1. A method for facilitating resolution of a mainframe production problem experienced by a mainframe and increasing the speed at which a mainframe resolves the mainframe production problem, the method being implemented by at least one processor, the method comprising:
receiving, by the at least one processor via a communication interface, at least one keyword associated with at least one query for resolving the mainframe production problem;
compressing, with at least one processor, a dataset stored in the mainframe to be searched using the at least one keyword to reduce a retrieval time for retrieving at least part of the dataset stored in the mainframe during a search for the received at least one keyword in the dataset;
identifying during a mainframe session, by the at least one processor, in the dataset at least one step or paragraph containing the at least one keyword based on the at least one query using a trained artificial-intelligence, machine-learning model
trained on training data of a labeled dataset identifying how its data should be classified to learn from the labeled dataset of the training data how to classify data in the dataset stored in the mainframe, and
using one or more of a k-method analysis, a regression analysis, a decision tree analysis, a random forest analysis, a k-nearest neighbor's analysis, a logistic regression analysis, a k-fold cross-validation analysis, and a balanced class weight analysis;
tagging, by the at least one processor, the at least one identified step with a composite identifier or tagging the at least one identified paragraph with the composite identifier, the composite identifier including a query identifier, a step identifier, and a paragraph identifier;
verifying, by the at least one processor, the at least one tagged identifier by matching the at least one tagged identifier with a corresponding parameter a) associated with the compressed dataset in the mainframe session and b) stored in a database associated with the mainframe session, thereby ensuring the accuracy of the search, wherein the verifying of the at least one tagged identifier results in one from among a) a successful verification of the at least one tagged identifier and b) an unsuccessful verification of the at least one tagged identifier;
using, by the at least one processor, the trained artificial-intelligence, machine-learning model to generate in the database a set of data logs related to the at least one query and resolution of the mainframe production problem and to provide a recommendation to resolve the mainframe production problem using the generated set of data logs;
populating, by the at least one processor, an output file with the retrieved at least one identified step or paragraph;
using the trained artificial-intelligence, machine-learning model to automatically format the output file, by the at least one processor, in response to a successful verification of the at least one tagged identifier; and
controlling, by the at least one processor, a liquid crystal display, an organic light emitting diode, a flat panel display, a solid-state display, a cathode ray tube display, or a plasma display to display a visual representation of the formatted output file containing the at least one identified step or paragraph for facilitating the resolving of the mainframe production problem in response to the successful verification of the at least one tagged identifier.