File Search System with Relevance Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional file search systems face difficulties in efficiently retrieving desired files due to the extraction of excessive irrelevant files, requiring repeated keyword entries by users.
Innovation Solution
A file search system that includes an acquisition processing unit to acquire search keywords, a search processing unit to search files based on these keywords, and an output processing unit to output search results with a degree of relatedness represented by a score value corresponding to the appearance frequency of the keywords, improving search operability by prioritizing relevant files.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional full-text search is performed on multiple files, then all files containing the search keyword are extracted, but the number of extracted files becomes excessively large making it difficult for users to find desired files
Solution Approach 1:
The patent changes the parameter of search result presentation by introducing a relevance score based on keyword appearance frequency. Instead of simply listing all matching files, the system calculates and displays a score representing the degree of relatedness between the search keyword and each file, allowing users to prioritize results by relevance rather than treating all matches equally.
Solution Approach 2:
The patent segments the search results by introducing a relevance scoring mechanism that divides files into different levels of relevance. The output processing unit calculates a score for each file based on keyword frequency, effectively segmenting the result set into high-relevance and low-relevance files, helping users quickly identify the most relevant results without being overwhelmed by the total number of matches.
2Measurement precision
If users repeatedly enter different search keywords to find desired files, then search coverage increases, but operability and efficiency deteriorate
Solution Approach 1:
The patent implements feedback by providing users with relevance scores for each search result. This feedback mechanism allows users to immediately see which files are most relevant to their search query based on keyword frequency, eliminating the need for repeated search attempts with different keywords. The system feedback guides users directly to the most relevant results.
Solution Approach 2:
The patent performs preliminary action by pre-calculating and displaying the degree of relatedness (relevance score) for each file before users need to make selection decisions. The output processing unit prepares the ranked list of results with relevance indicators in advance, so users don't need to perform multiple search iterations to determine which files are most relevant.
3Quantity of substance
If all files containing search keywords are extracted without prioritization, then completeness of results is improved, but user ability to identify relevant files deteriorates
Solution Approach 1:
The patent introduces a new parameter (relevance score based on keyword frequency) to characterize the relationship between search keywords and files. This parameter transformation allows the system to maintain completeness of results while adding a dimension of relevance measurement, enabling users to identify the most relevant files among the complete result set.
Solution Approach 2:
The patent introduces an intermediary element (the relevance score calculated by the output processing unit) that mediates between the complete set of matching files and the user's need to identify relevant files. This intermediary provides quantitative information about the relationship between keywords and files, bridging the gap between comprehensive results and precise relevance identification.
Data Source
AI summary
A file search system includes an acquisition processing unit that acquires a search keyword for searching a predetermined file in a storage storing a plurality of files; a search processing unit that searches the predetermined file on the basis of the search keyword acquired by the acquisition processing unit; and an output processing circuit that outputs a search result of the search processing circuit and outputs a degree of relatedness representing a relationship between the search keyword and each of the files stored in the storage on a basis of a score value corresponding to appearance frequency of the search keyword, the score value corresponding to each of the files.