# US 2023/0409620 A1 · Non-transitory computer-readable recording medium storing information processing program, information processing method, information processing device, and information processing system

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

- Number: US 2023/0409620 A1
- Document ID: US20230409620A1
- Status: Published application
- Kind: A1
- Application number: 18/239,828
- Filed: 2023-08-30
- Published: 2023-12-21
- Earliest claimed: 2021-03-31
- Assignees: Fujitsu Limited
- Applicants: Fujitsu Limited
- Inventors: Nobuyuki Katae
- Web page: https://patentgrep.com/patent/US20230409620A1

## CPC

- G06F 16/3344: {using natural language analysis}
- G06F 40/166: Editing, e.g. inserting or deleting
- G06F 40/295: Named entity recognition
- G06F 40/30: Semantic analysis
- G06F 16/322: {Trees}
- G06F 16/3338: {Query expansion}
- G06F 16/36: Creation of semantic tools, e.g. ontology or thesauri
- G06F 40/237: Lexical tools
- 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 non-transitory computer-readable recording medium storing an information processing program for causing a computer to execute processing including: extracting named-entities that include a hypernym from a document, the document being any one of a search query or document data that is regarded as an analysis object; specifying a modifier character stung that modifies the extracted hypernym from the document; generating conditions to be applied when a knowledge graph is searched for a hyponym relevant to the extracted hypernym, based on a type and content of the named-entities included in the specified modifier character string; searching the knowledge graph for the hyponym in accordance with the generated conditions; and associating the extracted hypernym with the hyponym found in the search.

## Claims

### Claim 1

A non-transitory computer-readable recording medium storing an information processing program for causing a computer to execute processing comprising:
extracting named-entities that include a hypernym from a document, the document being any one of a search query or document data that is regarded as an analysis object;
specifying a modifier character string that modifies the extracted hypernym from the document;
generating conditions to be applied when a knowledge graph is searched for a hyponym relevant to the extracted hypernym, based on a type and content of the named-entities included in the specified modifier character string;
searching the knowledge graph for the hyponym in accordance with the generated conditions; and
associating the extracted hypernym with the hyponym found in the search.

### Claim 2 (depends on 1)

The non-transitory computer-readable recording medium according to claim 1, wherein
the associating includes:
performing a lookup in the document for the hyponym found in the search; and
associating the hypernym in the document with the hyponym found in the lookup in the document.

### Claim 3 (depends on 1)

The non-transitory computer-readable recording medium according to claim 1, wherein
the associating includes:
performing a lookup in another document different from the document for the hyponym found in the search; and
associating the hypernym in the document with the hyponym found in the lookup in the another document.

### Claim 4 (depends on 1)

The non-transitory computer-readable recording medium according to claim 1, wherein
the knowledge graph is information obtained by forming a directed graph with knowledge regarding compounds as nodes and relationships between the nodes as edges, and
the generating includes:
referring to a storage unit that stores information capable of specifying a search object node according to the type and the content of the named-entities included in a phrase or a clause that modifies generic names of the compounds; and generating the conditions, based on the type and the content of the named-entities included in the specified modifier character string.

### Claim 5 (depends on 1)

The non-transitory computer-readable recording medium according to claim 1, wherein
the generating includes
when a plurality of the named-entities is included in the modifier character string and the plurality of the named-entities involves an alternative conjunction, setting an OR condition in the conditions generated for each of the plurality of the named-entities.

### Claim 6 (depends on 1)

The non-transitory computer-readable recording medium according to claim 1, wherein
the generating includes
when a plurality of the named-entities is included in the modifier character string and the plurality of the named-entities involves a linking conjunction, setting an AND condition in the conditions generated for each of the plurality of the named-entities.

### Claim 7 (depends on 1)

The non-transitory computer-readable recording medium according to claim 1, wherein
the generating includes
when the named-entities that involve a negative word are included in the modifier character string, setting a NOT condition in the conditions for the named-entities.

### Claim 8 (depends on 1)

The non-transitory computer-readable recording medium according to claim 1, the processing further comprising:
searching the knowledge graph for the named-entities included in the modifier character string; and
changing the conditions generated for the named-entities, based on the hyponym of the named-entities when the knowledge graph includes the hyponym of the named-entities found in the search, wherein
the searching includes
searching the knowledge graph for the hyponym relevant to the hypernym in accordance with the changed conditions.

### Claim 9 (depends on 2)

The non-transitory computer-readable recording medium according to claim 2, the processing further comprising, when displaying the document, displaying the document such that association between the hypernym and the hyponym that have been associated in the document is specifiable.

### Claim 10 (depends on 3)

The non-transitory computer-readable recording medium according to claim 3, the processing further comprising, when displaying the document and the another document, displaying the document and the another document such that association between the hypernym in the document and the hyponym in the another document that have been associated is specifiable.

### Claim 11 (depends on 2)

The non-transitory computer-readable recording medium according to claim 2, wherein
the associating includes
outputting the hyponym found in the search in association with the hypernym in the document.

### Claim 12 (depends on 1)

The non-transitory computer-readable recording medium according to claim 1, wherein
the associating of the extracted hypernym with the hyponym found in the search includes setting the extracted hypernym and the hyponym found in the search as lookup keywords for when performing a lookup for a document according to the lookup query.

### Claim 13

An information processing method implemented by a computer, the information processing method comprising:
extracting named-entities that include a hypernym from a document, the document being any one of a search query or document data that is regarded as an analysis object;
specifying a modifier character string that modifies the extracted hypernym from the document;
generating conditions to be applied when a knowledge graph is searched for a hyponym relevant to the extracted hypernym, based on a type and content of the named-entities included in the specified modifier character string;
searching the knowledge graph for the hyponym in accordance with the generated conditions; and
associating the extracted hypernym with the hyponym found in the search.

### Claim 14 (depends on 13)

The information processing method according to claim 13, wherein
the associating of the extracted hypernym with the hyponym found in the search includes setting the extracted hypernym and the hyponym found in the search as lookup keywords for when performing a lookup for a document according to the lookup query.

### Claim 15

An information processing apparatus comprising:
memory; and
processor circuitry coupled to the memory, the processor circuitry being configured to perform processing including:
extracting named-entities that include a hypernym from a document, the document being any one of a search query or document data that is regarded as an analysis object;
specifying a modifier character string that modifies the extracted hypernym from the document;
generating conditions to be applied when a knowledge graph is searched for a hyponym relevant to the extracted hypernym, based on a type and content of the named-entities included in the specified modifier character string;
searching the knowledge graph for the hyponym in accordance with the generated conditions; and
associating the extracted hypernym with the hyponym found in the search.

### Claim 16 (depends on 15)

The information processing apparatus according to claim 15, wherein
the associating of the extracted hypernym with the hyponym found in the search includes setting the extracted hypernym and the hyponym found in the search as lookup keywords for when performing a lookup for a document according to the lookup query.
