GENERATION METHOD, GENERATION PROGRAM, AND INFORMATION PROCESSING DEVICE

Provided are a generation method, a generation program, and an information processing device that improve security of a machine learning system. Tree structure information representing a structure of an attack tree including a first node associated with information indicating an attack to be established and second nodes associated with conditions to establish the attack, difficulty level information indicating difficulty of response to make the conditions associated with the respective second nodes unsatisfied, and condition matching status information that indicates whether the second nodes are satisfied are received, and, for each non-establishment condition candidate being a combination by which the attack is not established among combinations of the conditions associated with the respective second nodes, provision information that indicates a priority of adoption of the non-establishment condition candidate is generated based on the tree structure information, the difficulty level information, and the condition matching status information.