Document Search Apparatus Using Phenomenon Expression Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In the design of long-service-life products like heat power plants and air-conditioning management systems, there is a challenge in preserving knowledge about past nonconforming cases and their solutions, as this knowledge is often dependent on experienced designers and not effectively transferred to novice designers, leading to potential repeated issues.
Innovation Solution
A document search system that acquires event texts, extracts phenomenon expressions, acquires synonymous and co-occurrence expressions, and uses these to search for related-event documents from a database, allowing users to find documents describing similar nonconforming cases, thereby facilitating the retrieval of knowledge on past issues and their solutions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If knowledge about past nonconforming cases is stored in a database and made accessible, then knowledge preservation and retrieval capability is improved, but the complexity of the system increases due to needing advanced search functionality
Solution Approach 1:
The patent introduces an expression conversion unit that acts as an intermediary between the user's natural language search queries and the database search system. This unit automatically converts search terms into synonymous expressions and co-occurrence expressions, bridging the gap without requiring complex user input or system configuration
Solution Approach 2:
The system creates copies of search terms through synonymous expressions and co-occurrence expressions. Instead of searching for the exact query term, the system generates multiple semantic variants and searches using these copies, improving knowledge retrieval without adding system complexity
2Device complexity
If simple keyword search is used, then system complexity is reduced, but search accuracy and ability to find relevant nonconforming cases deteriorates
Solution Approach 1:
The patent changes the search parameters by transforming single keyword searches into multi-expression searches. The expression conversion unit generates synonymous expressions and co-occurrence expressions, effectively changing the search from a simple parameter to a compound parameter set, improving accuracy while keeping the underlying search mechanism simple
Solution Approach 2:
The search process is segmented into distinct stages: original term input, expression conversion (generating synonyms and co-occurrences), and multi-term searching. This segmentation allows the system to achieve high search accuracy through a structured process without requiring a single complex search algorithm
3Loss of information
If experienced designers share knowledge through guidance, then knowledge transfer to novice designers is improved, but time consumption and operational complexity increase
Solution Approach 1:
The system enables novice designers to independently access and utilize accumulated knowledge from past nonconforming cases without requiring time-consuming guidance from experienced designers. The automated expression conversion and search functionality allows users to self-serve their knowledge needs
Solution Approach 2:
The system performs preliminary actions by pre-processing and storing synonymous expressions and co-occurrence expressions in the database before they are needed. When users search, the system can quickly retrieve relevant knowledge without performing complex real-time analysis, saving time
Data Source
AI summary
According to one embodiment, a document search apparatus includes processing circuitry. The processing circuitry acquires an event text relating to an event, acquire a phenomenon expression from the event text, the phenomenon expression indicating a phenomenon of the event, acquires a synonymous expression having a meaning similar to a meaning of the phenomenon expression, and acquires at least one co-occurrence expression co-occurring with an expression selected from among the phenomenon expression and the synonymous expression. The processing circuitry performs search using the selected expression and one of the at least one co-occurrence expression as search terms to acquire a related-event document relating to the event from document information.