# US 11,308,941 B2 · Natural language processing apparatus and program

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

- Number: US 11,308,941 B2
- Document ID: US11308941B2
- Status: Granted
- Kind: B2
- Application number: 16/829,508
- Filed: 2020-03-25
- Published: 2021-09-23
- Granted: 2022-04-19
- Earliest claimed: 2020-03-19
- Assignees: Nomura Research Institute, Ltd.
- Applicants: Nomura Research Institute, Ltd.
- Inventors: Junichiro Maki; Satoshi Tobita; Shuichi Watanabe; Yosuke Hori; Jun Eijima
- Web page: https://patentgrep.com/patent/US11308941B2

## CPC

- G06F 16/35: Clustering; Classification
- G10L 15/1815: {Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning}
- G06F 16/2237: {Vectors, bitmaps or matrices}
- G06F 16/3347: {using vector based model}
- G06F 40/30: Semantic analysis
- G06F 40/40: Processing or translation of natural language (natural language analysis G06F40/20; semantic analysis G06F40/30)
- G06F 16/00: Information retrieval; Database structures therefor; File system structures therefor
- G06F 16/24: Querying
- G06F 16/33: Querying
- G06F 16/3344: {using natural language analysis}
- G10L 2015/088: {Word spotting}
- 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

A natural language processing apparatus includes: a first calculation unit configured to calculate a distributed vector of a word included in a plurality of sentences based on a database that manages the plurality of sentences associated with a classification word; a second calculation unit configured to calculate a distributed vector of the sentence based on the distributed vector of the word included in each sentence; and a third calculation unit configured to calculate a distributed vector of the classification word based on the distributed vector of each sentence associated with the same classification word.

## Claims

### Claim 1

A natural language processing apparatus comprising:
a processor; and
a storage device coupled to the processor,
wherein the storage device comprises an enterprise database in which an enterprise, a business content sentence, a business characteristic word and a business type are associated with each other,
the processor:
generates an enterprise vector from a word vector of a word included in the business content sentence stored in the enterprise database;
generates a business characteristic word vector from the enterprise vector of the enterprise that is associated with the business characteristic word in the enterprise database; and
generates a business type vector from the enterprise vector of the enterprise that is associated with the business type in the enterprise database,
upon receiving a search query, the processor:
extracts a synonym having a word vector with a similarity, that is equal to or higher than a first reference value, with a word vector of the search query;
extracts the business characteristic word having the business characteristic word vector with a similarity, that is equal to or higher than a second reference value, with the word vector of the search query or the word vector of the synonym;
extracts the business type having the business type vector with a similarity, that is equal to or higher than a third reference value, with the word vector of the search query or the word vector of the synonym; and
extracts, from the enterprise database, enterprise information including at least one of the search query, the extracted synonym, the extracted business characteristic word or the extracted business type in at least one of the business content sentence, the business characteristic word or the business type.

### Claim 2 (depends on 1)

The natural language processing apparatus according to claim 1, wherein the processor further outputs at least one of the extracted synonym, the extracted business characteristic word or the extracted business type.

### Claim 3 (depends on 2)

The natural language processing apparatus according to claim 2, wherein the processor further extracts a plurality of synonyms, a plurality of business characteristic words or a plurality of business types, and
outputs the plurality of extracted synonyms, the plurality of extracted business characteristic words or the plurality of extracted business types arranged in order of the similarity.

### Claim 4 (depends on 3)

The natural language processing apparatus according to claim 3, wherein the processor outputs at least one of:
the similarity between the word vector of the synonym and the word vector of the search query;
the similarity between the business characteristic word vector of the business characteristic word and the word vector of the search query or the word vector of the synonym; or
the similarity between the business type vector of the business type and the word vector of the search query or the word vector of the synonym.

### Claim 5

A non-transitory medium comprising an enterprise database in which an enterprise, a business content sentence, a business characteristic word and a business type are associated with each other, and storing a program causing an information processing apparatus to execute:
generating an enterprise vector from a word vector of a word included in the business content sentence stored in the enterprise database;
generating a business characteristic word vector from the enterprise vector of the enterprise that is associated with the business characteristic word in the enterprise database; and
generating a business type vector from the enterprise vector of the enterprise that is associated with the business type in the enterprise database,
wherein, upon the information processing apparatus receives a search query, the program causes the information processing apparatus to execute:
extracting a synonym having a word vector with a similarity, that is equal to or higher than a first reference value, with a word vector of the search query;
extracting the business characteristic word having the business characteristic word vector with a similarity, that is equal to or higher than a second reference value, with the word vector of the search query or the word vector of the synonym;
extracting the business type having the business type vector with a similarity, that is equal to or higher than a third reference value, with the word vector of the search query or the word vector of the synonym; and
extracting, from the enterprise database, enterprise information including at least one of the search query, the extracted synonym, the extracted business characteristic word or the extracted business type in at least one of the business content sentence, the business characteristic word or the business type.