System and method for simultaneous recognition of intent and slot in voice assistant commands

In an embodiment of the present disclosure, a method of simultaneously recognizing intent and slot in a voice assistant command includes lexical-making a current utterance of a user of a device including the voice assistant command into a plurality of lexical-making, presetting the plurality of lexical-making with a previous utterance and a partition lexical-making, obtaining one or more predictions for the voice assistant command from a plurality of preset lexical elements using a transducer-based machine learning model, aligning at least one of a flag prediction, a target prediction, and a sub-target prediction according to one or more constraints, the identified intent and the identified slot are provided to the device based on the aligned at least one of the flag prediction, the target prediction, and the sub-target prediction and the intent prediction, such that the device executes the voice assistant command in accordance with the identified intent and the identified slot. 在本公开的实施例中，一种同时识别语音助理命令中的意图和槽位的