# US 12,112,128 B2 · Apparatus and method for generating word embedding library

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

- Number: US 12,112,128 B2
- Document ID: US12112128B2
- Status: Granted
- Kind: B2
- Application number: 17/955,273
- Filed: 2022-09-28
- Published: 2024-03-28
- Granted: 2024-10-08
- Earliest claimed: 2022-09-28
- Assignees: Korea Electric Power Corporation
- Applicants: Korea Electric Power Corporation
- Inventors: Sung Min Kim; Hye Won Lim; Yoon Bo Shim; Yui Ha; Yoon Seok Choi
- Web page: https://patentgrep.com/patent/US12112128B2

## CPC

- G06F 40/211: Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
- G06F 40/129: Handling non-Latin characters, e.g. kana-to-kanji conversion
- G06F 40/268: Morphological analysis
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

The present invention relates to a method of generating a word embedding library, including: receiving, by a processor, original text composed of Hangul through an input interface; segmenting, by the processor, the original text by morpheme, combining segmented morphemes step by step according to a preset rule, and matching a tag to a combination of step-by-step morphemes according to a morphological attribute or a syntactic attribute of the combination of step-by-step morphemes; and generating, by the processor, a word embedding library by classifying the morphemes included in the original text based on the tag matched to the combination of step-by-step morphemes.

## Claims

### Claim 1

An apparatus and a method for generating a word embedding library, comprising:
an input interface; and
a processor,
wherein the processor is configured to receive original text composed of Hangul through the input interface,
segment the original text by morpheme and combine segmented morphemes step-by-step according to a preset rule,
match a tag to a combination of step-by-step morphemes according to a morphological attribute or a syntactic attribute of the combination of step-by-step morphemes, and
generate a word embedding library by classifying the morphemes included in the original text based on the tag matched to the combination of step-by-step morphemes,
a display configured to output various results calculated in a process of generating the word embedding library, and
a memory configured to pre-store various data necessary for generating the word embedding library.

### Claim 2 (depends on 1)

The apparatus of claim 1, wherein the processor is configured to:
match a part-of-speech tag to the segmented morpheme according to a part of speech when matching the tag to the combination of step-by-step morphemes;
identify, from the original text, a noun type word defined by a first morpheme whose part of speech is a first type or a combination of the first morpheme and another morpheme based on the part-of-speech tag;
determine a type of the noun type word based on the part-of-speech tag matched to the noun type word; and
match a noun type tag to the noun type word according to the type of the noun type word.

### Claim 3 (depends on 2)

The apparatus of claim 2, wherein the processor is configured to:
identify, from the original text, a compound noun type word defined by a combination of a noun type word and a first type noun type word based on the noun type tag; and
match a compound noun type tag to the compound noun type word when matching the tag to the combination of step-by-step morphemes.

### Claim 4 (depends on 3)

The apparatus of claim 3, wherein the processor is configured to:
identify, from the original text, a basic descriptive word defined by a second morpheme whose part of speech is a second type or a combination of the second morpheme and another morpheme based on the part-of-speech tag when matching the tag to the combination of step-by-step morphemes;
determine a type of the basic descriptive word based on the part-of-speech tag matched to the second morpheme;
match the basic descriptive tag to the basic descriptive word according to the type of the basic descriptive word;
identify a supplementary descriptive word defined by a basic descriptive word located in front of the basic descriptive word from the original text based on the basic descriptive tag;
match a supplementary descriptive tag to the supplementary descriptive word;
identify a descriptive word defined by a combination of the basic descriptive word and the supplementary descriptive word from the original text based on the basic descriptive tag and the supplementary descriptive tag; and
match a descriptive tag to the descriptive word according to the type of the basic descriptive word included in the descriptive word.

### Claim 5 (depends on 4)

The apparatus of claim 4, wherein the processor is configured to:
identify, from the original text, a noun phrase defined by a combination of a descriptive word and a third morpheme whose part of speech is a third type based on the part-of-speech tag and the descriptive tag when matching the tag to the combination of step-by-step morphemes;
determine a type of the noun phrase based on the part-of-speech tag matched to the descriptive tag and the third morpheme; and
match a noun phrase tag to the noun phrase according to the type of the noun phrase.

### Claim 6 (depends on 5)

The apparatus of claim 5, wherein the processor is configured to:
identify a compound noun phrase defined by a combination of a noun phrase and another noun phrase from the original text based on the noun phrase tag when matching the tag to the combination of step-by-step morphemes; and
match a compound noun phrase tag to the compound noun phrase.

### Claim 7 (depends on 6)

The apparatus of claim 6, wherein the processor is configured to:
identify, from the original text, a subordinate component of a noun phrase defined by a combination of a compound noun type word, a noun type word, or a noun phrase and a fourth morpheme whose part of speech is a fourth type based on the noun phrase tag and the part-of-speech tag when matching the tag to the combination of step-by-step morphemes;
determine the type of the basic descriptive word based on the part-of-speech tag matched to the second morpheme;
match a subordinate component tag to the subordinate component according to a type of the subordinate component;
identify a noun clause defined by a combination of a subordinate component and a noun phrase from the original text based on the noun phrase tag and the subordinate component tag;
determine a type of the noun clause based on the noun phrase tag; and
match a noun clause tag to the noun clause according to the type of the noun clause.

### Claim 8 (depends on 7)

The apparatus of claim 7, wherein the processor is configured to:
identify, from the original text, a compound noun phrase defined by a combination of a noun phrase and a noun clause or a combination of the noun clause and another noun clause based on the noun phrase tag and the noun clause tag when matching the tag to the combination of the step-by-step morphemes; and
match the compound noun clause tag to the compound noun clause.

### Claim 9 (depends on 8)

The apparatus of claim 8, wherein the processor is configured to:
identify a noun defined by a compound noun clause, a noun clause, a compound noun phrase, a noun phrase, a compound noun type word, a noun type word, or a morpheme capable of independently performing a nominal role in a sentence when matching the tag to the combination of the step-by-step morphemes; and
match a noun tag to the noun.

### Claim 10 (depends on 9)

The apparatus of claim 9, wherein the processor is configured to:
identify a predicate defined by a combination of a descriptive word and a fifth morpheme whose part of speech is a fifth type from the original text based on the descriptive tag and the part-of-speech tag and the descriptive tag when matching the tag to the combination of step-by-step morphemes;
determine a type of the predicate; and
match a predicate tag to the predicate according to the type of the predicate.

### Claim 11

A method of generating a word embedding library, comprising:
receiving, by a processor, original text composed of Hangul through an input interface;
segmenting, by the processor, the original text by morpheme, combining segmented morphemes step by step according to a preset rule, and matching a tag to a combination of step-by-step morphemes according to a morphological attribute or a syntactic attribute of the combination of step-by-step morphemes; and
generating, by the processor, a word embedding library by classifying the morphemes included in the original text based on the tag matched to the combination of step-by-step morphemes.

### Claim 12 (depends on 11)

The method of claim 11, wherein the matching of the combination of the step-by-step morphemes includes:
matching a part-of-speech tag to the segmented morpheme according to a part of speech;
identifying, from the original text, a noun type word defined by a first morpheme whose part of speech is a first type or a combination of the first morpheme and another morpheme based on the part-of-speech tag;
determining a type of the noun type word based on the part-of-speech tag matched to the noun type word; and
matching a noun type tag to the noun type word according to the type of the noun type word.

### Claim 13 (depends on 12)

The method of claim 12, wherein the matching of the combination of the step-by-step morphemes includes:
identifying a compound noun type word defined by a combination of a noun type word and a first type noun type word from the original text based on the noun type tag; and
matching a compound noun type tag to the compound noun type word.

### Claim 14 (depends on 13)

The method of claim 13, wherein the matching of the combination of the step-by-step morphemes includes:
identifying, from the original text, a basic descriptive word defined by a second morpheme whose part of speech is a second type or a combination of the second morpheme and another morpheme based on the part-of-speech tag;
determining the type of the basic descriptive word based on the part-of-speech tag matched to the second morpheme;
matching the basic descriptive tag to the basic descriptive word according to the type of the basic descriptive word;
identifying a supplementary descriptive word defined by a basic descriptive word located in front of the basic descriptive word from the original text based on the basic descriptive tag;
matching a supplementary descriptive tag to the supplementary descriptive word;
identifying a descriptive word defined by a combination of the basic descriptive word and the supplementary descriptive word from the original text based on the basic descriptive tag and the supplementary descriptive tag; and
matching a descriptive tag to the descriptive word according to the type of the basic descriptive word included in the descriptive word.

### Claim 15 (depends on 14)

The method of claim 14, wherein the matching of the combination of the step-by-step morphemes includes:
identifying a noun phrase defined by a combination of a descriptive word and a third morpheme whose part of speech is a third type from the original text based on the part-of-speech tag and the descriptive tag;
determining a type of the noun phrase based on the part-of-speech tag matched to the descriptive tag and the third morpheme; and
matching a noun phrase tag to the noun phrase according to the type of the noun phrase.

### Claim 16 (depends on 15)

The method of claim 15, wherein the matching of the combination of the step-by-step morphemes includes:
identifying a compound noun phrase defined by a combination of a noun phrase and another noun phrase from the original text based on the noun phrase tag; and
matching a compound noun phrase tag to the compound noun phrase.

### Claim 17 (depends on 16)

The method of claim 16, wherein the matching of the combination of the step-by-step morphemes includes:
identifying, from the original text, a subordinate component of a noun phrase defined by a combination of a compound noun type word, a noun type word, or a noun phrase and a fourth morpheme whose part of speech is a fourth type based on the noun phrase tag and the part-of-speech tag;
determining the type of the basic descriptive word based on the part-of-speech tag matched to the second morpheme;
matching a subordinate component tag to the subordinate component according to a type of the subordinate component;
identifying a noun clause defined by a combination of a subordinate component and a noun phrase from the original text based on the noun phrase tag and the subordinate component tag;
determining a type of the noun clause based on the noun phrase tag; and
matching a noun clause tag to the noun clause according to the type of the noun clause.

### Claim 18 (depends on 17)

The method of claim 17, wherein the matching of the combination of the step-by-step morphemes includes:
identifying a compound noun phrase defined by a combination of a noun phrase and a noun clause or a combination of the noun clause and another noun clause from the original text based on the noun phrase tag and the noun clause tag; and
matching the compound noun clause tag to the compound noun clause.

### Claim 19 (depends on 18)

The method of claim 18, wherein the matching of the combination of the step-by-step morphemes includes:
identifying a noun defined by a compound noun clause, a noun clause, a compound noun phrase, a noun phrase, a compound noun type word, a noun type word, or a morpheme capable of independently performing a nominal role in a sentence; and
matching a noun tag to the noun.

### Claim 20 (depends on 19)

The method of claim 19, wherein the matching of the combination of the step-by-step morphemes includes:
identifying, from the original text, a predicate defined by a combination of a descriptive word and a fifth morpheme whose part of speech is a fifth type based on the descriptive tag and the part-of-speech tag and the descriptive tag;
determining a type of the predicate; and
matching a predicate tag to the predicate according to the type of the predicate.