# US 2019/0095524 A1 · Context-based virtual assistant implementation

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2019/0095524 A1
- Document ID: US20190095524A1
- Status: Published application
- Kind: A1
- Application number: 15/712,480
- Filed: 2017-09-22
- Published: 2019-03-28
- Earliest claimed: 2017-09-22
- Assignees: Oracle International Corporation
- Applicants: Oracle International Corporation
- Inventors: Michael Patrick Rodgers
- Web page: https://patentgrep.com/patent/US20190095524A1

## CPC

- G06F 16/3329: Natural language query formulation
- G06F 16/24575: {using context}
- G06F 16/334: Query execution (filtering based on additional data G06F16/335)
- G10L 15/22: Procedures used during a speech recognition process, e.g. man-machine dialogue
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING

## Abstract

Techniques for maintaining a dialogue between a virtual assistant and a user are disclosed. The system generates a context object. Contextual information, related to the dialogue, is stored to the context object. The contextual information may define a state. Based on the state, the virtual assistant may present information, present a query, execute a query, or receive information from a user. The virtual assistant determines context throughout the conversation via the context object. The virtual assistant may to determine a variable corresponding to received input using information stored to the context object.

## Claims

### Claim 1

A non-transitory computer readable medium comprising instructions which, when executed by one or more hardware processors, causes performance of operations comprising:
receiving, by a query system via a smart speaker, a first voice input from a user;
based on the first voice input: modifying, by the query system, a context object from a first state to a second state;
selecting, by the query system, information associated with the second state to be requested from the user;
requesting, by the query system via the smart speaker, the information associated with the second state;
receiving, by the query system via the smart speaker, a second voice input from the user;
mapping, based at least on the first voice input, at least a portion of the second voice input to a first variable corresponding to the information; and
storing at least the portion of the second voice input as a value for the first variable.

### Claim 2 (depends on 1)

The medium of claim 1, wherein the operations further comprise: prior to receiving the first voice input, configuring an initial state of the context object that is user-input agnostic.

### Claim 3 (depends on 1)

The medium of claim 1, wherein mapping based at least on the first voice input comprises mapping based on the second state that was determined based on the first voice input.

### Claim 4 (depends on 1)

The medium of claim 1, wherein executable code for the query system is computer-generated based on a user-defined state diagram comprising the first state and the second state.

### Claim 5 (depends on 4)

The medium of claim 4, wherein the user-defined state diagram is designed via a user interface comprising graphical components representing, at least, the first state and the second state in the user-defined state diagram.

### Claim 6 (depends on 1)

The medium of claim 1, wherein the operations further comprise:
based on the second voice input: modifying, by the query system, the context object from the second state to a third state;
selecting additional information, associated with the third state, to be requested from the user;
requesting, by the query system via the smart speaker, the additional information associated with the third state;
receiving, by the query system via the smart speaker, a third voice input from the user;
mapping, based at least on the second voice input, at least a portion of the third voice input to a second variable corresponding to the additional information; and
storing at least the portion of the third voice input as a value for the second variable.

### Claim 7 (depends on 6)

The medium of claim 6, wherein the third state is based on the query system transitioning to the second state from the first state.

### Claim 8 (depends on 1)

The medium of claim 1, wherein the operations further comprise discarding a second portion of the second voice input without mapping the second portion to any variable.

### Claim 9 (depends on 1)

The medium of claim 1, wherein modifying the context object from the first state to the second state is based on the first voice input being received while the context object is in the first state.

### Claim 10 (depends on 1)

The medium of claim 1, wherein the second state is the same as the first state.

### Claim 11

A non-transitory computer readable medium comprising instructions which, when executed by one or more hardware processors, causes performance of operations comprising:
receiving, by a query system via a smart speaker, a first voice input from a user;
based on the first voice input: modifying, by the query system, a context object from a first state to a second state;
selecting, by the query system, information associated with the second state to be requested from the user;
requesting, by the query system via the smart speaker, the information associated with the second state;
receiving, by the query system via the smart speaker, a second voice input from the user;
subsequent to receiving the second voice input: selecting, by the query system, additional information to be requested from the user based at least on both the first voice input and the second voice input; and
requesting, by the query system via the smart speaker, the additional information.

### Claim 12 (depends on 11)

The medium of claim 11, wherein the additional information to be requested from the user is selected based on an order of receipt of the first voice input and the second voice input.

### Claim 13 (depends on 11)

The medium of claim 11, wherein the additional information to be requested from the user is selected based on a time between the first voice input and the second voice input.

### Claim 14 (depends on 11)

The medium of claim 11, wherein the additional information to be requested from the user is selected based on a lack of any intervening user input, which changes the context object, between the first voice input and the second voice input.

### Claim 15 (depends on 11)

The medium of claim 11, wherein the operations further comprise: prior to receiving the first voice input, configuring an initial state of the context object that is user-input agnostic.

### Claim 16 (depends on 11)

The medium of claim 11, wherein the second state is the same as the first state.

### Claim 17 (depends on 11)

The medium of claim 11, wherein executable code for the query system is computer-generated based on a user-defined state diagram comprising the first state and the second state.

### Claim 18 (depends on 17)

The medium of claim 17, wherein the user-defined state diagram is designed via a user interface comprising graphical components representing a plurality of states of the user-defined state diagram.

### Claim 19 (depends on 11)

The medium of claim 11, wherein modifying the context object from the first state to the second state is based on the first voice input being received while the context object is in the first state.

### Claim 20

A system comprising:
at least one device including a hardware processor;
the system configured to perform operations comprising:
receiving, by a query system via a smart speaker, a first voice input from a user;
based on the first voice input: modifying, by the query system, a context object from a first state to a second state;
selecting, by the query system, information associated with the second state to be requested from the user;
requesting, by the query system via the smart speaker, the information associated with the second state;
receiving, by the query system via the smart speaker, a second voice input from the user;
mapping, based at least on the first voice input, at least a portion of the second voice input to a first variable corresponding to the information; and
storing at least the portion of the second voice input as a value for the first variable.
