Named Entity Extraction Grouping for Hierarchical File Naming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for automatically assigning file names from document images struggle with efficiently extracting and displaying named entities with hierarchical structures, requiring users to manually select hierarchical layers each time, which is cumbersome and inefficient.
Innovation Solution
An information processing apparatus that includes an extraction unit for recognizing named entities and a display control unit to group entities with hierarchical structures, allowing multiple related entities to be displayed as a single group on a UI screen, simplifying the selection and modification of hierarchical layers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If detailed category classification is performed for named entities with hierarchical structures, then extraction precision is improved, but device complexity increases and ease of operation deteriorates
Solution Approach 1:
The named entity recognition system is segmented into multiple hierarchical levels (coarse-grained and fine-grained categories). The extraction unit can operate at different granularity levels, allowing users to select the appropriate level of detail based on their needs, thus improving ease of operation while maintaining extraction precision when needed.
Solution Approach 2:
The system dynamically adjusts the classification granularity based on user input and context. Users can switch between viewing named entities at a high level (e.g., organization type) or drill down to specific hierarchical layers (e.g., department names), making the system adaptable to different operational requirements without sacrificing precision.
2Measurement precision
If detailed category classification is performed for named entities with hierarchical structures, then extraction precision is improved, but device complexity increases
Solution Approach 1:
The classification system is divided into modular hierarchical levels that can be independently managed. Each level handles a specific aspect of categorization, reducing the complexity burden on any single component while maintaining overall extraction precision through the combined hierarchical structure.
Solution Approach 2:
The system adds a hierarchical dimension to the classification structure, organizing named entities across multiple levels from general to specific. This dimensional organization reduces complexity by providing a structured framework that systematically manages detailed categorization without overwhelming the system architecture.
3Adaptability or versatility
If users manually select hierarchical layers for extraction each time, then adaptability is improved, but productivity deteriorates
Solution Approach 1:
The system performs preliminary classification of named entities into hierarchical categories during the extraction process. This preliminary organization allows users to later select from pre-structured options, maintaining adaptability while significantly improving productivity by eliminating the need for manual hierarchical layer selection during each processing task.
Solution Approach 2:
The extraction unit automatically performs hierarchical classification and organizes named entities into structured categories without requiring user intervention. This self-service capability maintains the system's adaptability to different entity types while boosting productivity by automating what would otherwise be a manual, time-consuming process.
Data Source
AI summary
The information processing apparatus includes: an extraction unit configured to extract a named entity by inputting document data to a named entity recognition model; and a display control unit configured to display a named entity extracted by the extraction unit on a UI screen, and in a case where a plurality of named entities extracted by the extraction unit is in a relationship between a generic concept and a specific concept, the display control unit displays the plurality of the named entities in one group.