SYSTEMS AND METHODS FOR GENERIC FLEXIBLE DIALOGUE MANAGEMENT

• US 20180114528A1
• Filed: 02/10/2017
• Published: 04/26/2018
• Est. Priority Date: 10/26/2016
• Status: Active Application

• Current Assignee: IPsoft Incorporated
• Original Assignee: IPsoft Incorporated
• Inventors: YASAVUR, Ugan, AMINI, Reza, TRAVIESO, Jorge
• Application Number:
  US15/429,987
• Publication Number:
  US 20180114528A1
• Time in Patent Office: Days
• CPC Class Codes: G06F 40/35 Discourse or dialogue representation
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/1822 Parsing for meaning understanding
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/223 Execution procedure of a spoken command
• CPC Class Codes: H04L 51/02 using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages

Abstract

A method for a virtual agent to process natural language utterances from a user is provided. The method can include receiving natural language utterance from the user, determining a type of the utterance, and based on the utterance type, determining an action for the virtual agent to take. The virtual agent can execute the action.

Claims

1. A method for a virtual agent to process natural language utterances from a user, the method comprising:

receiving a first natural language utterance from the user, wherein the first natural language utterance comprises a first goal or a first piece of information needed to satisfy the first goal, and the virtual agent has a first process in progress and wherein the natural language utterance does not satisfy one or more slots in the first process;

instantiating a second process from a plurality of processes for the virtual agent based on the first natural language utterance, wherein each process in the plurality of processes comprises a corresponding goal and a corresponding set of slots;

receiving a second natural language utterance from the user wherein the second natural language utterance comprises a second goal of the user, a second piece of information needed to satisfy the first goal or the second goal, or a first dialogue act;

determining which action of a plurality of actions for the virtual agent to take based on an utterance type of the second natural language utterance from the user, the plurality of actions comprising instantiate a third process, fill a slot within the first process, fill a slot within the second process, execute a behavior, ask for input from the user, and output a statement to the user; and

executing by the virtual agent the action.

2. The method of claim 1 wherein determining the action further comprises:

for the second natural language utterance of the second goal of the user, the utterance type is goal and;

if all information needed to satisfy the second goal has been received, then the action for the virtual agent is execute the behavior associated with the first goal, otherwise and

the action for the virtual agent is ask for input from the user.

3. The method of claim 1 wherein determining the action further comprises:

for the second natural language utterance of the second piece of information needed to satisfy the first goal or the second goal;

if the second natural language utterance is in response to the virtual agent asking for input from the user, then the action is fill one or more slots in the first process or the second process, otherwise

the action is instantiate one or more additional processes and fill a slot within each process of the one or more additional instantiated processes.

4. The method of claim 3 wherein instantiating the one or more additional processes further comprising identifying all processes from a plurality of processes that have a slot that matches the piece of information.

5. The method of claim 3 further comprising:

if a number of the one or more processes identified is less than or equal to three then transmitting a multi-choice question, otherwise, transmitting a request to clarify the piece of information to a user.

6. The method of claim 1 wherein determining the action further comprises:

for the second natural language utterance of the first dialogue act, the utterance type is dialogue act and the action is ask for input from the user, fill a slot, escalate to a human agent or output a statement to the user.

17. The method of claim 1, wherein the first natural language utterance further comprises one or more additional goals, one or more additional piece of information needed to satisfy the first goal, or any combination thereof.

18. The method of claim 17 wherein for the first natural language utterance of the first goal and the one or more additional goals, then instantiating the second process further comprises determining the second process for the first goal, and further comprising determining a corresponding additional process for each of the one or more additional goals.

19. The method of claim 17 wherein for the first natural language utterance of the first piece of information and the additional pieces of information, then instantiating the second process further comprises identifying one process of the plurality of processes that has a slot that the first piece of information can fill, and further comprising:

filing any slots in the first process or the second process that can be filled by the additional pieces of information, and for any of the additional pieces of information that cannot fill any slots in the first process or the second process, identifying all processes from the plurality of processes that have a corresponding slot for that the additional pieces of information that cannot fill any slots in the first process or the second process can fill, and instantiating all of the identified processes, such that all of instantiated processes operate in parallel.

7-16. -16. (canceled)

20. (canceled)