Quotation Indexing for Document Value Ranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing text-search-and-retrieval systems, particularly in legal research, fail to adequately rank documents by their relevance and value, as they rely on statistical methods rather than logical connotations, and lack tools to identify and highlight significant phrases or quotes from other documents.
Innovation Solution
A software system that identifies and highlights quoted text within documents, using quotation frequency and citations to determine the value of source documents, and displays a list of documents that have quoted the source, allowing for enhanced ranking and search result filtering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional statistical methods are used to rank search results, then the ranking process is simple and fast, but the relevance and value of documents cannot be adequately determined
Solution Approach 1:
The system pre-processes documents during indexing to identify and mark quoted text, storing quotation information in advance. This preliminary action enables the ranking system to access pre-computed quotation data during search operations, eliminating the need for complex real-time analysis while improving document value determination
Solution Approach 2:
The patent introduces quotation information as an intermediary metric between traditional citation counts and document relevance. By marking and tracking quoted text during indexing, the system creates a new layer of information that mediates between raw citation data and meaningful document valuation, enhancing relevance determination without proportionally increasing complexity
2Measurement precision
If quotation analysis is added to identify significant phrases, then document significance can be determined more accurately, but the system complexity and processing time increase
Solution Approach 1:
The system performs quotation identification and text marking during the document indexing phase rather than during search operations. By pre-processing documents to identify quoted passages and store this information in metadata, the system eliminates time-consuming analysis during user queries, achieving accurate phrase significance identification without increasing perceived processing time
Solution Approach 2:
The patent extracts quotation information as separate metadata from the main document content during indexing. By isolating quoted text identification into a distinct preprocessing step, the system can efficiently store and retrieve quotation data without repeatedly analyzing full document texts during search operations, reducing processing time while maintaining accuracy
3Measurement precision
If all matching documents are retrieved for a search query, then completeness of results is achieved, but the user cannot easily determine which documents are most valuable
Solution Approach 1:
The patent introduces quotation-based metrics as an intermediary evaluation layer that enriches search results with content-based value information. By tracking and weighting documents according to their quoted content during indexing, the system provides automated relevance signals that help users identify valuable documents within complete result sets, recovering the loss of content-based evaluation present in traditional Boolean search systems
Data Source
AI summary
A computerized system and method is presented for analyzing quotations made in a quoting document of text originally found in a source document. The present invention analyzes the quoting document for quoted text, searches the source document for that text, and stores the existence of the quotation in association with the source document. When displaying the source document, a list of text items that were quoted by another document is presented sorted by frequency of quotation. Quoted text is also highlighted when the original text of the source document is displayed. Hyperlinking is allowed between items of quoted text by the user and a list of documents that have quoted that text. From this list the full text of a quoting document may be displayed. The frequency of quotations taken from a source document can also be used to rank documents in search results.