Cyberattack Information Complementing via Dependency Parsing

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for structuring information on cyberattacks, such as those using named entity recognition, fail to provide specific details necessary for investment decisions and security measures, as they only extract general information like 'customer information' without specifying its contents.

Innovation Solution

An information complementing apparatus and method that includes named entity extraction, dependency parsing, and complementation processing units to extract and enhance named entities from news articles, providing detailed modifiers to specify the content of information like 'customer information' for better understanding and decision-making.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If named entity recognition is used to structure cyberattack information, then information extraction speed is improved, but information completeness deteriorates

Engineering Contradiction:

Improveinformation extraction speedVSAvoidinformation completeness

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The patent segments the information extraction process into two distinct stages: first, named entity recognition extracts basic entities quickly; second, dependency parsing separately extracts detailed modifiers and relationships. This segmentation allows each component to specialize, maintaining speed while improving completeness.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent merges named entity recognition and dependency parsing into a unified information structuring system. The output of NER feeds into dependency parsing, and both results are combined to produce comprehensive structured information including both entities and their detailed modifiers.

Inventive Principle:

Principle #5Merging (Combining)

2Measurement precision

If detailed information is extracted from news articles, then information quality is improved, but processing complexity increases

Engineering Contradiction:

Improveinformation qualityVSAvoidprocessing complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The processing system is segmented into specialized modules: named entity recognition handles entity identification, while dependency parsing handles relationship and modifier extraction. This segmentation reduces the complexity burden on any single component while maintaining high information quality.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Structured information serves as an intermediary representation between the raw news article and the final detailed extraction. The dependency parsing unit uses this structured format to efficiently extract modifiers and relationships, reducing overall processing complexity.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS20240176954A1Information complementing apparatus, information complementing method, and computer readable recording medium

Publication Date: 2024.05.30 NEC CORP

AI summary

An information complementing apparatus 10 includes: a named entity extraction unit 11 that extracts named entities from a news article about a cyberattack; a dependency parsing unit 12 that parse a dependency relation between words or clauses in the news article; and a complementation processing unit 13 that specifies a named entity satisfying a set condition from among the extracted named entities and complements the specified named entity with a corresponding modifier, based on a result of the dependency relation parsing.