State Tracking Over Machine-Learned Relational Trees in a Dialog System

摘要

A method is provided for representing and updating the state of a dialog involving a series of queries and commands to an artificial intelligence system. Each statement within the dialogue may be modeled as a relational tree spanning nodes corresponding to named entities within the statement. A data structure may be used to store each of these trees and to modify them as the dialog progresses. A subsequent statement in the dialog may be parsed and its contents used to update an ongoing search initiated within that dialog. Statements may be used for the update process despite being fragmentary or not corresponding to any predetermined grammar. An algorithm is disclosed for updating the trees within the data structure after a new statement is parsed.

主权项

1. A method, comprising:
receiving data representing a first user intent statement; identifying a plurality of named entities within the first user intent statement; identifying one or more relationships between one or more pairs from the plurality of named entities; generating a graph comprising two or more of the plurality of named entities and one or more of the identified one or more relationships; receiving data representing a second user intent statement; performing an action based on an intent identified in the second user intent statement; and subsequent to performing the action based on the intent identified in the second user intent statement, performing an action based on an intent identified in the first user intent statement.