US 12,112,145 B2
Artificial intelligence-based intelligent programming assistance
Kumar Abhinav, Hazaribag (IN); Alpana Dubey, Bangalore (IN); Hindnavis Vijaya Shravani, Hyderabad (IN); and Meenakshi D'Souza, Bangalore (IN)
Assigned to ACCENTURE GLOBAL SOLUTIONS LIMITED, Dublin (IE)
Filed by ACCENTURE GLOBAL SOLUTIONS LIMITED, Dublin (IE)
Filed on Jul. 8, 2021, as Appl. No. 17/370,670.
Claims priority of application No. 202011029184 (IN), filed on Jul. 9, 2020.
Prior Publication US 2022/0012021 A1, Jan. 13, 2022
Int. Cl. G06F 8/33 (2018.01); G06N 3/044 (2023.01); G06N 3/045 (2023.01)
CPC G06F 8/33 (2013.01) [G06N 3/044 (2023.01); G06N 3/045 (2023.01)] 20 Claims
OG exemplary drawing
 
1. An artificial intelligence-based intelligent programming assistance apparatus comprising:
at least one hardware processor; and
an error resolver, executed by the at least one hardware processor, to
ascertain, for a software program that is to be completed, code for the program,
identify, based on an analysis of the ascertained code and error message-based contextual information generated by a compiler associated with the ascertained code, at least one error in the ascertained code,
generate, based on the at least one identified error, a resolution to the at least one identified error, and
modify, based on the generated resolution to the at least one identified error, the ascertained code to resolve the at least one identified error;
run the modified ascertained code through the compiler,
wherein the compiler accepts the modified ascertained code on confirming absence of new errors in the modified ascertained code and that the modified ascertained code includes less number of errors than the ascertained code and
wherein the compiler rejects the modified ascertained code, wherein the modified ascertained code includes new errors or the modified ascertained code includes greater number of errors than the ascertained code;
a code generator, executed by the at least one hardware processor, to
generate, based on an analysis of the ascertained code and problem description-based contextual information associated with the ascertained code, a next token associated with the ascertained code,
generate, based on the next token associated with the ascertained code, further code for the program to be completed, and
add the further generated code to the code of the program to be completed;
a personalized feedback generator, executed by the at least one hardware processor, to
analyze, for the program, a performance of a user with respect to generation of the code for the program, and
generate, based on the analyzed performance of the user, feedback for the user; and
a response generator, executed by the at least one hardware processor, to
ascertain, from the user, a query associated with the program,
classify, by utilizing an ontology, the query into a specified concept of a plurality of concepts, and
generate, based on an analysis of the classified query, a response to the query.