Reading Comprehension Graphs for Shortest-Path Information Retrieval
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for document reading and comprehension, such as using tables of contents or keyword searches, are inefficient and fail to accurately and easily guide users to the necessary information, especially when keywords are not directly present or when document structures are complex.
Innovation Solution
A reading comprehension support system that creates a graph representing the structure of a document, allowing users to select keywords from output words or phrases, and searches for the shortest paths between these keywords, presenting related sentences and complementary information to enhance understanding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If keyword search is used to find desired information, then the reader can locate relevant sentences or paragraphs, but the search may return too many hits requiring excessive time to find the desired information
Solution Approach 1:
The patent segments the document into a hierarchical structure with sections and subsections, creating a segmented navigation system that allows readers to quickly locate information without searching the entire document. This segmentation reduces the number of hits from a global keyword search to a manageable subset within specific sections.
Solution Approach 2:
The patent introduces an intermediary index structure that mediates between the reader's information need and the document content. The index acts as a mediator by providing structured access points (sections and subsections) that guide readers directly to relevant information, reducing search time while maintaining accuracy.
2Ease of operation
If a single keyword is used for search, then the search process is simple, but a single keyword is unable to narrow down the desired information effectively
Solution Approach 1:
The patent segments the search process into multiple hierarchical levels (section level and subsection level). Readers can perform simple searches at the section level and then drill down to subsections, maintaining operational simplicity while improving retrieval accuracy through progressive narrowing.
Solution Approach 2:
The patent adds a hierarchical dimension to the search process by organizing content into sections and subsections. This dimensional structure allows readers to navigate from broad to specific levels, transforming a single-dimension keyword search into a multi-dimensional navigation process that improves precision without sacrificing simplicity.
3Reliability
If document structural analysis is performed in accordance with set rules, then the document can be systematically organized, but documents with various structures are difficult to handle
Solution Approach 1:
The patent creates a universal hierarchical structure (sections and subsections) that can accommodate various document types and structures. This multi-functional framework can adapt to different document formats while maintaining consistent organization principles, thereby achieving both reliability and versatility.
Solution Approach 2:
The patent allows flexibility in the hierarchical parameters by enabling documents to have different numbers of sections and subsections based on their specific needs. This parameter flexibility maintains the reliable hierarchical structure while adapting to various document types and structures.
4Ease of operation
If table of contents or index is used to find information, then the document structure can be navigated, but efficiency is low when the word to be found is not used directly in the table of contents or index
Solution Approach 1:
The patent segments the navigation system into hierarchical sections and subsections, allowing readers to navigate to relevant sections even when exact keywords are not present. This segmentation enables approximate matching and contextual navigation, improving retrieval accuracy while maintaining ease of operation.
Solution Approach 2:
The patent enhances the traditional index by introducing a hierarchical section-subsection structure as an intermediary between the reader's query and the document content. This intermediary structure provides multiple access points and contextual guidance, improving both navigation ease and retrieval accuracy when exact keywords are absent.
Data Source
AI summary
A document reading comprehension support system which presents information necessary for a user with high accuracy is provided. A reading comprehension support system which receives a designated document; creates a first graph representing a structure of the designated document with words or phrases contained in the designated document; outputs two or more words or phrases contained in the first graph; receives a plurality of designated words or phrases from the output words or phrases; searches the first graph with the plurality of designated words or phrases; and outputs a search result, is provided. As the search result, at least a second graph representing a shortest path between any two of the plurality of designated words or phrases in the first graph can be output. The shortest path is a path connecting any two of the plurality of designated words or phrases through at least one complementary word or phrase. The complementary word or phrase is a word or phrase different from the plurality of designated words or phrases.