Citation Network Analysis for Document Relevance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Keyword-based searches for relevant documents often fail to provide a complete representation due to varying terminologies and inconsistent citation practices, making it challenging and time-consuming to determine relevant documents through direct citations alone.
Innovation Solution
The development of automatic systems and methods that process citation information to determine relevant documents based on both direct and indirect citations, using relevance scores and indices to identify citationally relevant documents, thereby expanding the scope of search beyond direct citation networks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If keyword-based searches are used to find relevant documents, then the search process is simple and fast, but the completeness of relevant document representation is insufficient due to varying terminologies
Solution Approach 1:
The patent uses citation information as an intermediary to connect documents that share common references. By processing citation networks, the system identifies documents that are relevant through shared citations even when they don't share keywords, thus mediating between fast keyword search and comprehensive document discovery
Solution Approach 2:
The patent transitions from one-dimensional keyword matching to multi-dimensional document relationships by incorporating citation networks. It analyzes documents across multiple dimensions including direct citations, indirect citations, and co-citation patterns to comprehensively identify relevant documents
2Loss of information
If manual screening of citation information is performed to improve document relevance, then completeness of relevant documents improves, but time consumption increases significantly
Solution Approach 1:
The patent replaces manual mechanical screening with automated computer-based processing of citation information. The system automatically retrieves, processes, and analyzes citation networks to identify relevant documents, eliminating the time-consuming manual screening process while maintaining comprehensive document discovery
Solution Approach 2:
The system performs self-service by automatically processing its own citation analysis without human intervention. It autonomously retrieves citation data, processes the information through defined algorithms, and generates relevant document sets, making the comprehensive screening process efficient and scalable
3Reliability
If direct citation search is used to find relevant documents, then the search method is reliable and straightforward, but the scope of relevant documents is limited due to inconsistent citation practices
Solution Approach 1:
The patent implements a dynamic search approach that adapts the scope of citation analysis based on the query. It can adjust between direct citations only, indirect citations included, or multiple levels of citation networks, making the search both reliable and adaptable to different research needs and document coverage requirements
Solution Approach 2:
The system creates a universal search mechanism that works across different citation practices and document types. By processing multiple citation patterns (direct, indirect, co-citations) and applying normalization techniques, it handles inconsistent citation practices universally while expanding document coverage
Data Source
AI summary
Methods, systems and computer-readable storage media relate to automatically and efficiently determine a set of relevant document(s) relevant to an inputted set of document(s) based on at least indirect citations. The method may include determining a group of one or more citing documents from one or more citing documents for a queried set of one or more documents. The queried set may include the inputted set and/or one or more sets of relevant document(s). The method may also include determining a group of the one or more cited documents from the one or more citing documents for the group. The method may include determining relevance information for each cited document of the group. The method may further include determining a first set of one or more documents that are relevant to the inputted set of one or more documents based on the relevance information of the second group.