Retrieval Support Program for Incident Information
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In business systems, operators face difficulties in retrieving appropriate incident information due to variations in content across different operating systems, even when the underlying incident events have the same contents, leading to incomplete or inaccurate retrieval of solutions for new incidents.
Innovation Solution
A retrieval support program that analyzes frequency information across multiple categories to identify unique and common terms, generating correspondence information to associate first terms with second terms, allowing for the identification of correlativity between terms from different categories, thereby facilitating the retrieval of relevant incident information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If incident information is stored and retrieved using original terms from different operating system categories, then the information can be preserved in its original form, but retrieval accuracy decreases due to term variations across categories
Solution Approach 1:
The patent introduces correspondence information as an intermediary layer between terms from different operating system categories. This correspondence information stores relationships between terms (e.g., between first terms unique to specific OS categories and second terms common across categories), enabling accurate retrieval by translating category-specific terms into comparable forms without modifying the original incident information storage.
2Productivity
If incident information is retrieved using exact term matching, then retrieval speed is improved, but completeness of retrieval decreases due to synonym and category variation
Solution Approach 1:
The patent performs preliminary action by pre-establishing correspondence information between terms from different categories before the actual retrieval process. This correspondence information is stored in advance, allowing the retrieval process to quickly check pre-computed relationships rather than performing complex real-time analysis, thus maintaining high retrieval speed while improving completeness.
3Loss of information
If correspondence information is generated to associate terms across categories, then retrieval completeness is improved, but processing time and computational resources increase
Solution Approach 1:
The correspondence information is generated and stored in advance as a preliminary action, separating the computationally intensive term relationship analysis from the actual retrieval process. This allows the retrieval operation itself to be fast, as it only needs to query pre-computed correspondence relationships rather than performing complex analysis during retrieval.
Data Source
AI summary
A retrieval support method including acquiring, for each of a plurality of categories relating to computer systems, frequency information including a frequency of occurrence of each of a plurality of terms included in pieces of text information on the computer systems, specifying, a plurality of first terms and a plurality of second terms, the plurality of first terms being terms which is not used in common to the plurality of categories, the plurality of second terms being a terms which is used in common, generating correspondence information, for each of the plurality of first terms, that associates selected first term with at least one of the plurality of second terms which is included in a piece of text information including the selected first term, and generating specified information to identify each combination of first terms, belonging to different category among the plurality of categories, that have correlativity.