US 11,755,455 B1
Detecting and correcting user interface discrepancies
Calvin Yue-Ren Kuo, Mercer Island, WA (US); Zhaofeng Zhan, Belfair, WA (US); Stuart Olmstead-Wilcox, Richmond (CA); Tian Chen, Redmond, WA (US); Zheshen Wang, Bellevue, WA (US); Jingyu Dong, Kenmore, WA (US); and Dan Catalin Teodorescu, Redmond, WA (US)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on Jun. 29, 2021, as Appl. No. 17/362,842.
Int. Cl. G06F 16/958 (2019.01); G06N 20/00 (2019.01); G06F 16/957 (2019.01); G06F 11/36 (2006.01)
CPC G06F 11/3624 (2013.01) [G06F 11/3664 (2013.01); G06F 16/957 (2019.01); G06F 16/958 (2019.01); G06N 20/00 (2019.01)] 15 Claims
OG exemplary drawing
 
1. A method comprising:
simulating user interaction with an online shopping system user interface to generate requests for displayable webpage data for a plurality of webpages from a web server associated with the online shopping system and to generate a plurality of document object models (DOMs) associated with the webpage data;
applying the DOMs to a model to identify semantic blocks of the displayable webpage data for the plurality of webpages;
extracting information from the identified semantic blocks of data;
comparing the extracted information from the identified semantic blocks across the plurality of webpages;
detecting differences between the extracted information from a first identified semantic block of a first webpage with the extracted information from a second, corresponding semantic block of a second webpage;
analyzing source code associated with the first and second webpages to determine a root cause of the difference, wherein the source code is analyzed using a machine learning-assisted backend analysis unit comprising computer-executable instructions to generate call graphs for the source code and a dependency graph for the source code, and wherein the machine learning-assisted backend analysis unit further comprises computer-executable instructions to apply the call graphs and the dependency graph to a call pattern defect detection model to identify issues in the source code relating to calls between source code provided by different backend services; and
modifying the source code corresponding to at least one of the first and second webpages to correct the identified issues in the source code relating to calls between source code provided by different backend services.