Graph-Based Keyword Expansion for Information Retrieval
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing keyword expansion methods in information retrieval technology rely on external search engines, which are quality and performance dependent, and lack effective algorithms for generating relevant related keywords.
Innovation Solution
A standalone keyword expansion mechanism processes terms from documents to build a graph data structure where keywords and their relationships are represented as nodes and edges, with weights indicating relevance, allowing for the determination of expanded keywords based on proximity and association weights.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If external search engines are used for keyword expansion, then keyword expansion can be achieved, but the solution depends on external search engine quality and availability
Solution Approach 1:
The patent extracts the keyword expansion functionality from external search engines and implements it as a standalone system. The core idea is to take out the dependency on external services by building an independent graph-based expansion mechanism that processes documents locally without requiring external search engine APIs or services.
Solution Approach 2:
The system performs self-service by using its own internally built graph data structure to conduct keyword expansion. Instead of relying on external search engines, the system uses its accumulated knowledge base and graph relationships to autonomously generate expanded keywords, making the expansion process self-contained and independent.
2Measurement precision
If search engine snippets are used for keyword expansion, then related keywords can be extracted, but the expansion quality depends on snippet generation algorithms
Solution Approach 1:
The patent extracts the keyword expansion logic from external search engine snippet generation algorithms and implements it through a dedicated graph-based approach. By taking out the dependency on external algorithms, the system achieves independent control over expansion quality through its own graph traversal and weighting mechanisms.
Solution Approach 2:
The system changes the fundamental parameters of keyword expansion by using graph theory concepts (nodes, edges, weights, traversal algorithms) instead of traditional search engine snippet generation parameters. This parameter transformation enables superior expansion quality through graph-based relationship analysis and weight accumulation mechanisms.
3Adaptability or versatility
If a graph data structure is built from document terms, then standalone keyword expansion is achieved, but document processing complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-processing documents to extract terms and build the graph data structure before actual keyword expansion queries are executed. This preliminary graph construction captures relationships between terms across documents, enabling fast and accurate expansion queries without re-processing the entire document collection during each search operation.
Solution Approach 2:
The graph data structure serves as an intermediary between the raw document collection and the keyword expansion output. Instead of directly processing documents during expansion queries, the system uses the pre-built graph as an intermediate representation that encapsulates term relationships, simplifying the expansion process while maintaining standalone capability.
Data Source
AI summary
A keyword may be expanded into related words, such as for use in information retrieval. The terms comprising words and/or phrases of a large number of documents (e.g., web pages) are processed into a graph data structure, in which the terms are represented as nodes and edges represent the relationships between the nodes, with weights for each edge representing the relevance of the relationship. The graph may be built by selecting each term of a document and considering the terms within a certain number of words to be associated with the selected term; for each such association the weight indicative of the relevance is increased. When the graph is accessed with a keyword, the edges from that keyword's node and their respective weights indicate which other nodes are most relevant to the keyword, thereby providing the corresponding expanded terms.