Information Processing System Using Metadata for Related Document Retrieval
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing document retrieval systems require manual user input of keywords for finding related documents, which may fail to display relevant documents even when they exist, especially in cases where appropriate keywords are not used.
Innovation Solution
A computer-implemented method and system that automatically extracts metadata from documents, creates retrieval phrases, and retrieves related documents based on metadata, displaying them as candidates for easy user association.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual keyword retrieval is used, then user control over retrieval is maintained, but retrieval accuracy decreases when appropriate keywords are not used
Solution Approach 1:
The system automatically extracts metadata from the current document and uses it to retrieve related documents without requiring manual keyword input from the user. The retrieval phrase creation unit generates search queries based on extracted metadata, enabling the system to serve itself in the retrieval process.
Solution Approach 2:
The metadata extraction unit pre-extracts relevant information from documents and stores it in the database. When a document is viewed, the retrieval phrase creation unit immediately uses the pre-extracted metadata to generate retrieval phrases, eliminating the need for users to manually think about and input appropriate keywords.
2Measurement precision
If automatic metadata extraction is implemented, then retrieval accuracy improves, but system complexity increases
Solution Approach 1:
The system is divided into distinct functional units: metadata extraction unit, database unit, retrieval phrase creation unit, and related document retrieval unit. Each unit performs a specific task, making the overall complex process manageable and maintainable through modular design.
Solution Approach 2:
The metadata acts as an intermediary between the document content and the retrieval process. Instead of directly analyzing full document content during retrieval, the system uses pre-extracted metadata as a bridge, simplifying the retrieval operation while maintaining accuracy.
3Loss of time
If manual keyword input is required, then user intent is precisely captured, but time consumption increases
Solution Approach 1:
The metadata extraction process runs continuously in the background as documents are stored or accessed, so that when retrieval is needed, the metadata is already prepared and ready to use, eliminating delays associated with on-demand keyword analysis.
Data Source
AI summary
A computer-implemented method comprises one or more processors configured to display a first document on a document browsing screen, retrieve, from a storage, a second document that is a document different from the first document and includes information corresponding to attribute information acquired from the first document, display the second document as a candidate for a related document related to the first document, and register the first document and the second document as related documents.