# US 2020/0004822 A1 · Method and device for extracting attributes associated with centre of interest from natural language sentences

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

- Number: US 2020/0004822 A1
- Document ID: US20200004822A1
- Status: Published application
- Kind: A1
- Application number: 16/111,266
- Filed: 2018-08-24
- Published: 2020-01-02
- Earliest claimed: 2018-06-30
- Assignees: Wipro Limited
- Applicants: Wipro Limited
- Inventors: Arindam Chatterjee; Kartik Subodh Ballal
- Web page: https://patentgrep.com/patent/US20200004822A1

## CPC

- G06F 40/30: Semantic analysis
- G06F 40/205: Parsing
- G06F 40/253: Grammatical analysis; Style critique
- G06F 40/289: Phrasal analysis, e.g. finite state techniques or chunking
- G06N 3/042: Knowledge-based neural networks; Logical representations of neural networks
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- G06N 5/022: Knowledge engineering; Knowledge acquisition
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A method and device for extracting attributes associated with Center of Interest (COI) from natural language sentences is disclosed. The method includes creating an input vector comprising a plurality of parameters for each target word in a sentence inputted by a user. The method further includes processing for each target word, the input vector through a trained bidirectional GRU neural network, which is trained to identify attributes associated with COI from a plurality of sentences. The method includes associating COI attribute tags to each target word in the sentence based on processing of associated input vector through the trained bidirectional GRU neural network. The method further includes extracting attributes from the sentence based on the COI attribute tags associated with each target word in the sentence. The method further includes providing a response to the sentence inputted by the user based on the attributes extracted from the sentence.

## Claims

### Claim 1

A method for extracting attributes associated with Center of Interest (COI) from natural language sentences, the method comprising:
creating, by a COI attribute processing device, an input vector comprising a plurality of parameters for each target word in a sentence inputted by a user, wherein the plurality of parameters for each target word comprise a Part of Speech (POS) vector associated with the target word and at least two words preceding the target word, a word embedding for the target word, a word embedding for a head word of the target word in a dependency parse tree for the sentence, and a dependency label for the target word;
processing for each target word, by the COI attribute processing device, the input vector through a trained bidirectional Gated Recurrent Unit (GRU) neural network, wherein the trained bidirectional GRU neural network is trained to identify attributes associated with COI from a plurality of sentences, and wherein attributes associated with a COI in a sentence augment the context of the COI;
associating, by the COI attribute processing device, COI attribute tags to each target word in the sentence based on processing of associated input vector through the trained bidirectional GRU neural network;
extracting, by the COI attribute processing device, attributes from the sentence based on the COI attribute tags associated with each target word in the sentence; and
providing, by the COI attribute processing device, a response to the sentence inputted by the user based on the attributes extracted from the sentence.

### Claim 2 (depends on 1)

The method of claim 1, wherein the response comprises at least one of an answer to a query and an action corresponding to the query.

### Claim 3 (depends on 1)

The method of claim 1, wherein the dependency label for the target word indicates relation of the target word with the head word in the sentence.

### Claim 4 (depends on 1)

The method of claim 1 further comprising training the bidirectional GRU neural network to identify COI attribute tags for words within sentences.

### Claim 5 (depends on 4)

The method of claim 4, wherein training the bidirectional GRU neural network comprises:
assigning COI attribute tags to each word in a plurality natural language sentences retrieved from a data repository of natural language sentences comprising a plurality of COI attributes scenarios;
inputting, iteratively, the assigned COI attribute tag associated with each word in the plurality of natural language sentences to the bidirectional GRU neural network for training.

### Claim 6 (depends on 1)

The method of claim 1, wherein the COI attribute tags comprise Begin attribute tag, Inside attribute tag, and Others tag.

### Claim 7 (depends on 6)

The method of claim 6, wherein the associating the COI attribute tags to each target word comprises:
assigning a Begin attribute tag to a word in the sentence marking the beginning of the attributes;
assigning an Inside attribute tag to each word within the attributes succeeding the word marking the beginning of the attributes; and
assigning an Others tag to each remaining word in the sentence.

### Claim 8 (depends on 1)

The method of claim 1 further comprising determining the plurality of parameters for each target word in the sentence inputted by the user.

### Claim 9

A Center of Interest (COI) attribute processing device for extracting attributes associated with COI from natural language sentences, the COI attribute processing device comprising:
a processor; and
a memory communicatively coupled to the processor, wherein the memory stores processor instructions, which, on execution, causes the processor to:
create an input vector comprising a plurality of parameters for each target word in a sentence inputted by a user, wherein the plurality of parameters for each target word comprise a Part of Speech (POS) vector associated with the target word and at least two words preceding the target word, a word embedding for the target word, a word embedding for a head word of the target word in a dependency parse tree for the sentence, and a dependency label for the target word;
process for each target word, the input vector through a trained bidirectional Gated Recurrent Unit (GRU) neural network, wherein the trained bidirectional GRU neural network is trained to identify attributes associated with COI from a plurality of sentences, and wherein attributes associated with a COI in a sentence augment the context of the COI;
associate COI attribute tags to each target word in the sentence based on processing of associated input vector through the trained bidirectional GRU neural network;
extract attributes from the sentence based on the COI attribute tags associated with each target word in the sentence; and
provide a response to the sentence inputted by the user based on the attributes extracted from the sentence.

### Claim 10 (depends on 9)

The COI attribute processing device of claim 9, wherein the response comprises at least one of an answer to a query and an action corresponding to the query.

### Claim 11 (depends on 9)

The COI attribute processing device of claim 9, wherein the dependency label for the target word indicates relation of the target word with the head word in the sentence.

### Claim 12 (depends on 9)

The COI attribute processing device of claim 9, wherein the processor instructions further cause the processor to train the bidirectional GRU neural network to identify COI attribute tags for words within sentences.

### Claim 13 (depends on 12)

The COI attribute processing device of claim 12, wherein to train the bidirectional GRU neural network, the processor instructions further cause the processor to:
assign COI attribute tags to each word in a plurality natural language sentences retrieved from a data repository of natural language sentences comprising a plurality of COI attributes scenarios;
iteratively input the assigned COI attribute tag associated with each word in the plurality of natural language sentences to the bidirectional GRU neural network for training.

### Claim 14 (depends on 9)

The COI attribute processing device of claim 9, wherein the COI attribute tags comprise Begin attribute tag, Inside attribute tag, and Others tag.

### Claim 15 (depends on 14)

The COI attribute processing device of claim 14, wherein to associate the COI attribute tags to each target word, the processor instructions further cause the processor to:
assign a Begin attribute tag to a word in the sentence marking the beginning of the attributes;
assign an Inside attribute tag to each word within the attributes succeeding the word marking the beginning of the attributes; and
assign an Others tag to each remaining word in the sentence.

### Claim 16 (depends on 9)

The COI attribute processing device of claim 9, wherein the processor instructions further cause the processor to determine the plurality of parameters for each target word in the sentence inputted by the user.

### Claim 17

A non-transitory computer-readable storage medium having stored thereon, a set of computer-executable instructions causing a computer comprising one or more processors to perform steps comprising:
creating an input vector comprising a plurality of parameters for each target word in a sentence inputted by a user, wherein the plurality of parameters for each target word comprise a Part of Speech (POS) vector associated with the target word and at least two words preceding the target word, a word embedding for the target word, a word embedding for a head word of the target word in a dependency parse tree for the sentence, and a dependency label for the target word;
processing for each target word, the input vector through a trained bidirectional Gated Recurrent Unit (GRU) neural network, wherein the trained bidirectional GRU neural network is trained to identify attributes associated with COI from a plurality of sentences, and wherein attributes associated with a COI in a sentence augment the context of the COI;
associating COI attribute tags to each target word in the sentence based on processing of associated input vector through the trained bidirectional GRU neural network;
extracting attributes from the sentence based on the COI attribute tags associated with each target word in the sentence; and
providing a response to the sentence inputted by the user based on the attributes extracted from the sentence.
