Guided FAQ Text Retrieval for Broad Keyword Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional FAQ search services often fail to provide desired search results due to high generality of user input keywords, leading to numerous irrelevant hits and difficulty in finding relevant information.
Innovation Solution
An information processing device that uses a goal sentence or similar, extracts guidance sentence candidates from a dictionary, and presents them to the user, allowing selection of a desired sentence or similar for easy retrieval.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional FAQ search service uses general keywords for search, then search coverage is improved, but search result precision deteriorates due to numerous irrelevant hits
Solution Approach 1:
The patent introduces an intermediary processing step between keyword input and search result display. The system extracts entities from the user's keyword input, finds related entities, and generates candidate search queries that combine these entities. This intermediary process transforms vague general keywords into more specific search queries, thereby maintaining search coverage while improving result precision.
Solution Approach 2:
The system performs preliminary entity extraction and relationship analysis on user input before executing the actual search. By pre-processing the query to identify key entities and their relationships, the system prepares optimized search queries in advance, which improves the precision of search results without sacrificing the versatility of user input.
2Quantity of substance
If conventional FAQ search service displays all search results, then information completeness is improved, but ease of operation deteriorates due to difficulty in finding desired information
Solution Approach 1:
The patent segments the search results into multiple ranked lists based on different entity relationships and query candidates. Instead of displaying all results in a single undifferentiated list, the system divides results into priority-ordered groups, allowing users to easily find desired information while maintaining access to complete information through the segmented structure.
Solution Approach 2:
The system applies local quality by providing different levels of result presentation - highly relevant results are prominently displayed with higher visibility, while less relevant results are available but less emphasized. This creates a quality gradient in the results display that helps users quickly locate desired information while preserving information completeness.
3Device complexity
If conventional FAQ search service processes simple keyword input, then device complexity is reduced, but productivity deteriorates due to time required to find desired information
Solution Approach 1:
The system implements self-service by automatically performing entity extraction, relationship analysis, and query optimization without requiring user intervention. The system autonomously processes the user's simple keyword input through multiple analytical steps and generates optimized search queries, thereby maintaining low processing complexity for the user while significantly improving productivity through automated intelligence.
Solution Approach 2:
The patent replaces manual mechanical search processes with automated information processing mechanisms. Instead of requiring users to manually refine queries or browse through results systematically, the system uses automated entity recognition, relationship mapping, and query generation to accelerate the search process, thereby improving productivity without substantially increasing perceived complexity.
Data Source
AI summary
An information processing device, method and program enabling a user to arrive easily at a desired target text is described. A keyword accepting unit accepts a keyword relating to a target text, where the target text is an answer to a FAQ or a product description. A guidance text candidate extracting unit extracts a guidance text candidate including the keyword from within a guidance text dictionary containing guidance text candidates intended for the user to reach a target text have been registered in advance in association with a plurality of target texts. A guidance text candidate presenting unit presents guidance text candidates based on the extracted guidance text candidates to the user. A target text presenting unit extracts from the guidance text dictionary a target text associated with a guidance text candidate selected by the user from the presented guidance text candidates and presents the target text to the user.