# US 2016/0259780 A1 · Natural language processing (nlp) interfacing with devices

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

- Number: US 2016/0259780 A1
- Document ID: US20160259780A1
- Status: Published application
- Kind: A1
- Application number: 14/641,148
- Filed: 2015-03-06
- Published: 2016-09-08
- Earliest claimed: 2015-03-06
- Assignees: Dell Products L.P.
- Applicants: Dell Products L.P.
- Inventors: Sachinrao Panemangalore; Vinay Sawal; Vivek Dharmadhikari; Kuntal Patel; Gavin Richard Cato
- Web page: https://patentgrep.com/patent/US20160259780A1

## CPC

- G06F 40/284: Lexical analysis, e.g. tokenisation or collocates
- G06F 9/451: Execution arrangements for user interfaces
- G06F 40/151: Transformation
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Aspects of the present invention provide a more universal, easy, natural, and vendor-agnostic interface to configure, manage, and/or monitor devices in networks. In embodiments, a user-friendly natural language interface, such as a chat or messaging interface, may be used to “live chat” with one or more devices. In embodiments, a natural language input from a user intended for a target device is received and converted into one or more properly formed commands that are target-specific for the target device and may be executed by the target device. In embodiments, results from the execution of the one or more commands may be appropriately formatted for presentation to the user.

## Claims

### Claim 1

A computer-implemented method for converting a natural language input into a command input for a target device comprising:
receiving a natural language input comprising free-form vocabulary, the natural language input being intended for a target device that requires commands to be properly formed for the target device to be executed;
performing natural language processing (NLP) input cleaning on the natural language input; and
performing input transformation on at least a portion of the cleaned natural language input to convert the cleaned natural language input one or more properly formed commands that are target-specific for the target device.

### Claim 2 (depends on 1)

The computer-implemented method of claim 1 wherein the step of performing natural language processing (NLP) input cleaning on the natural language input comprises performing one or more of the steps comprising:
responsive to the natural language input having plurals or reflexive pronouns, stemming the natural language input to remove any plurals and reflexive pronouns; and
responsive to the natural language input comprising compound statements, complex statements, or both, splitting the natural language input into two or more simple natural language inputs.

### Claim 3 (depends on 1)

The computer-implemented method of claim 1 wherein the step of performing input transformation on at least a portion of the cleaned natural language input to convert the cleaned natural language input one or more properly formed commands that are target-specific for the target device comprises:
performing a lookup in a lemmatization database to convert at least some of the free-form vocabulary of the cleaned natural language input to a target-specific input that comprises target-specific vocabulary that corresponds to the target device;
using the target-specific input to perform a lookup in a command template database to obtain a target-specific command template match; and
responsive to the target-specific command template match having one more empty values, filling the one or more empty values from data obtained from the natural language input to obtain a properly formed target-specific command.

### Claim 4 (depends on 1)

The computer-implemented method of claim 1 wherein the step of receiving a natural language input comprising free-form vocabulary comprises:
receiving the natural language input via a messaging application.

### Claim 5 (depends on 4)

The computer-implemented method of claim 4 further comprising the step of:
presenting to a user via the messaging application one or more avatars corresponding to one or more target devices, which can be communicated with via the messaging application.

### Claim 6 (depends on 1)

The computer-implemented method of claim 1 further comprising the step of:
submitting the one or more properly formed target-specific commands to the target device for execution.

### Claim 7 (depends on 6)

The computer-implemented method of claim 6 further comprising:
receiving a set of results from the target device after execution of the one or more properly formed target-specific commands; and
formatting at least some of the results from the set of results for presentation to a user.

### Claim 8 (depends on 7)

The computer-implemented method of claim 7 wherein the step of formatting at least some of the results from the set of results for presentation to a user comprises:
formatting at least of the results from the set of result for presentation to a user via a chat messaging application.

### Claim 9

A natural language processing (NLP) system for converting a natural language input into a target-specific command input for a target device, the system comprising:
an end-point interface system comprising a natural language processing (NLP) interaction end-point server that receives a natural language input from a user, the natural language input being intended for a target device;
a transformation system that cleans the natural language input and transforms at least a portion of the cleaned natural language input into one or more properly formed commands that are target-specific for the target device; and
a target-device interface module that communicates the one or more properly formed commands to the target device for execution by the target device.

### Claim 10 (depends on 9)

The natural language processing (NLP) system of claim 9 wherein the transformation system comprises:
an NLP input cleaning and transformation logic module, which is communicatively coupled to the end-point interface system to receive the natural language input and is communicatively coupled to a lemmatization database, that performs a lookup in the lemmatization database to convert at least some of the natural language input into a target-specific input corresponding to the target device; and
a target-specific NLP processing engine module, which is communicatively coupled to the NLP input cleaning and transformation logic module and to a command template database, that performs a lookup in the command template database using the target-specific input to obtain a target-specific command template match and that, responsive to the target-specific command template match having one more empty values, fills the one or more empty values from data obtained from the natural language input to obtain at least one of the one or more properly formed commands that are target-specific for the target device.

### Claim 11 (depends on 10)

The natural language processing (NLP) system of claim 10 wherein the lemmatization database comprises:
a collection of structured records having a list of commands and each command is associated with list of terms related to different devices, different vendors, or both that are considered synonyms.

### Claim 12 (depends on 10)

The natural language processing (NLP) system of claim 10 wherein the command template database comprises:
target-specific command templates, each target-specific command template being in a structured format and having an associated set of one or more tags related to usage of the target-specific command template; and
one or more indexes of the target-specific command templates.

### Claim 13 (depends on 9)

The natural language processing (NLP) system of claim 9 further comprising:
an NLP response engine that is communicatively coupled to the target device interface module and that receives results data related to at least one of the one or more properly formed commands and forwards at least some of that data to a user.

### Claim 14 (depends on 13)

The natural language processing (NLP) system of claim 13 wherein the NLP response engine is further configured to perform the step comprising:
formatting at least some of the results data for presentation to the user.

### Claim 15 (depends on 14)

The natural language processing (NLP) system of claim 14 wherein the step of formatting at least some of the results data for presentation to the user comprises:
formatting the at least some of the results data for presentation to the user via a chat messaging application.

### Claim 16 (depends on 10)

The natural language processing (NLP) system of claim 10 further comprising:
a training system that is communicatively coupled to the one or more sets of target-related documentation and that uses the target-related documentation to develop the lemmatization database and the command template database.

### Claim 17

A computer-implemented method for generating one or more databases for use in converting a natural language input into a command input for a target device, the method comprising:
generating a lemmatization database that comprise a collection of structured records having a list of commands and each command is associated with list of terms related to different devices, different vendors, or both that are considered synonyms; and
generating a command template database comprises:
target-specific command templates, each target-specific command template being in a structured format and having an associated set of one or more tags related to usage of the target-specific command template; and
one or more indexes of the target-specific command templates.

### Claim 18 (depends on 17)

The computer-implemented method of claim 17 wherein the step of generating a lemmatization database comprises the step of:
extracting command name and corresponding descriptions for a domain from a support corpus of documentation that include documentation across different devices, vendors, or both;
generating a set of word vectors from at least some of the support corpus documentation; and
clustering the set of word vectors into clusters in which a cluster represents word vectors that are deemed to be synonyms.

### Claim 19 (depends on 18)

The computer-implemented method of claim 18 further comprising the step of:
detecting and removing one or more outliers from one or more clusters.

### Claim 20 (depends on 17)

The computer-implemented method of claim 17 wherein the step of generating a command template database comprises the step of:
extracting target-specific command templates from a set of documentation;
generating a target-specific command templates corpus in a structured format and label each target-specific command template with an associated set of one or more tags related to usage of the target-specific command template; and
indexing the target-specific command templates corpus.
