Structured Document Heading Generation via Feature Word Extraction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing techniques for generating search results and answers from structured documents often rely on inadequate headings, which can be identical, similar, or lack meaningful content, leading to inappropriate search outputs.

Innovation Solution

An information processing device that acquires structured documents, extracts feature words from subordinate elements of headings, and generates new headings based on these feature words to improve heading appropriateness.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If headings are added by referring to tag information from structured documents, then the heading generation process is simple and fast, but the heading information becomes insufficient and may be identical to other headings

Engineering Contradiction:

Improveheading generation efficiencyVSAvoidheading information quality

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The patent segments the heading generation process into multiple components: extracting feature words from subordinate elements, analyzing relationships between headings and texts, and generating optimized headings. This segmentation allows the system to maintain efficiency while improving information quality by processing different aspects separately and systematically.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces feature words as an intermediary between the raw tag information and the final headings. These feature words are extracted from subordinate elements and serve as a bridge to generate more informative headings, preventing information loss while maintaining the simplicity of the overall process.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Device complexity

If simple tag information is used for headings, then the processing complexity is low, but the search result accuracy decreases

Engineering Contradiction:

Improveprocessing complexityVSAvoidsearch result accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent performs preliminary actions by extracting feature words from subordinate elements before generating the final headings. This preliminary processing enriches the heading information without significantly increasing overall system complexity, thereby improving search result accuracy while maintaining manageable processing complexity.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent changes the parameters of heading generation by transitioning from using only tag information to using extracted feature words from subordinate elements. This parameter change enhances the quality and discriminative power of headings, leading to improved search accuracy without requiring fundamental changes to the processing architecture.

Inventive Principle:

Principle #35Parameter changes

3Loss of time

If headings are generated without analyzing subordinate elements, then the processing time is short, but the heading appropriateness and relevance are poor

Engineering Contradiction:

Improveprocessing timeVSAvoidheading appropriateness

Core Design Contradiction:

Loss of timeVSReliability

Solution Approach 1:

The patent extracts essential feature words from subordinate elements such as texts and subordinate headings. This extraction process selectively takes out the most relevant information needed for generating appropriate headings, minimizing the time investment while maximizing the quality and appropriateness of the generated headings.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS12216987B2Generating heading based on extracted feature words

Publication Date: 2025.02.04 NEC CORP

AI summary

The information processing device generates headings from structured documents. The acquisition means acquires a structured document including headings and texts. The feature word extraction means extracts feature words from subordinate elements of the heading for the headings included in the structured document. The heading generation means generates a new heading corresponding to the subordinate elements based on the extracted feature words.