US 12,455,734 B2
Deduplication of computer programs using similarity measures and artificial intelligence
Haiyang Xu, Shanghai (CN); Yuna Hou, Shanghai (CN); Fangling Liu, Shanghai (CN); Xiang Fan, Shanghai (CN); Xia Yu, Shanghai (CN); Wenle Zhao, Shanghai (CN); Jia Wang, Shanghai (CN); and You Li, Shanghai (CN)
Assigned to SAP SE, Walldorf (DE)
Filed by SAP SE, Walldorf (DE)
Filed on Aug. 3, 2023, as Appl. No. 18/229,915.
Prior Publication US 2025/0045043 A1, Feb. 6, 2025
Int. Cl. G06F 8/70 (2018.01)
CPC G06F 8/70 (2013.01) 20 Claims
OG exemplary drawing
 
1. A method, comprising:
receiving a request to compare a first computing program to a second computing program, wherein each computing program comprises an address, one or more tables accessed by a respective computing program, one or more input parameters, and one or more output parameters;
calculating an address similarity measure between the address of the first computing program and the address of the second computing program;
calculating a table similarity measure between the one or more tables accessed by the first computing program and the one or more tables accessed by the second computing program;
calculating an input similarity measure between the one or more input parameters of the first computing program and the one or more input parameters of the second computing program;
calculating an output similarity measure between the one or more output parameters of the first computing program and the one or more output parameters of the second computing program;
providing the address similarity measure, table similarity measure, input similarity measure, and output similarity measure to a trained artificial intelligence (AI) model;
receiving from the trained AI model a similarity determination between the first computing program and the second computing program; and
performing an action based on the similarity determination.