US 11,735,165 B2
Task-oriented dialog system and method through feedback
Scot Fang, Los Angeles, CA (US)
Assigned to Beijing DiDi Infinity Technology and Development Co., Ltd., Beijing (CN)
Filed by Beijing DiDi Infinity Technology and Development Co., Ltd., Beijing (CN)
Filed on Dec. 11, 2020, as Appl. No. 17/118,845.
Prior Publication US 2022/0189460 A1, Jun. 16, 2022
Int. Cl. G10L 15/06 (2013.01); G10L 15/22 (2006.01); G06F 3/16 (2006.01); G10L 15/01 (2013.01); G06F 40/56 (2020.01); G06N 20/00 (2019.01); G06F 9/451 (2018.01)
CPC G10L 15/063 (2013.01) [G06F 9/453 (2018.02); G06F 40/56 (2020.01); G06N 20/00 (2019.01); G10L 15/01 (2013.01); G10L 2015/0638 (2013.01)] 17 Claims
OG exemplary drawing
 
1. A computer-implemented method for improving an automatic agent, comprising:
receiving a user input through a user interface;
obtaining a plurality of current utterance variables by tokenizing the user input;
generating a reply to the user input based on the plurality of current utterance variables and the automatic agent executing a machine learning policy, wherein the generating the reply comprises triggering one or more Application Programming Interface (API) calls according to the machine learning policy;
displaying the reply on the user interface;
in response to detecting a first user action applied to the generated reply, displaying the one or more triggered API calls;
in response to detecting a second user action applied to one of the one or more triggered API calls, displaying an alternative API call;
detecting a third user action applied to the alternative API call, and obtaining a different reply by triggering the alternative API call; and
adjusting the machine learning policy based on the plurality of current utterance variables, the alternative API call, and the different reply.