METHOD FOR AUTOMATING END-TO-END TESTING USING MACHINE LEARNING MODEL Russian patent published in 2025 - IPC G06F11/3668 G06N3/02 G06N20/00 

FIELD: computer engineering.

SUBSTANCE: invention relates to computer engineering and specifically to automated end-to-end software testing. Disclosed is a method for automating end-to-end testing using a machine learning model by machine translation, comprising steps of: collecting data from selected web service pages; processing the collected data, including cleaning from noise, in the form of unused layout elements, normalization and conversion into the corresponding processing format; machine learning model is trained on the basis of a predetermined data set; generating test cases based on model results; generating test scenarios, including a sequence of actions, expected results and checking conformity of actual and expected page state; executing the generated test cases and test scenarios by the automated testing system, wherein said test cases are additionally used for additional training of the machine learning model.

EFFECT: novel method of automating end-to-end testing.

5 cl, 10 dwg