Search Result Relevance via Concept Network Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current search engines struggle to accurately determine the relevance of search results to a user's intent due to the artificial nature of queries and the inability to differentiate between relevant and irrelevant terms, often returning irrelevant results ranked highly.
Innovation Solution
The technique involves parsing search queries into units and analyzing a concept network to select relevant concepts based on their relationships, performing individual searches for these concepts, and reranking results based on their relevance to the original query, using units, associated units, and extension units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If search engines use traditional query processing methods, then search results are returned quickly, but the relevance of results to user intent deteriorates
Solution Approach 1:
The search query is segmented into multiple units (e.g., 'skyscraper', 'building', 'Empire State Building') representing different levels of concept granularity. Each unit is then searched independently and results are merged, allowing the system to handle complex queries systematically while improving relevance through multi-level concept matching.
Solution Approach 2:
A concept network is introduced as an intermediary structure that maps queries to related concepts. This mediator layer translates user queries into concept relationships, enabling the system to retrieve relevant information through conceptual associations rather than direct keyword matching, thus improving relevance without requiring overly complex processing.
2Reliability
If search engines process queries literally, then processing is simple, but irrelevant terms cause irrelevant results to be ranked highly
Solution Approach 1:
The system changes the parameter of query interpretation from literal string matching to conceptual relationship matching. By transforming queries into units and searching through concept networks, the system adapts the meaning of terms dynamically, allowing it to distinguish between relevant and irrelevant terms even when they appear similar or related.
3Measurement precision
If search engines return all matching results, then coverage is complete, but the quality and relevance of results deteriorates
Solution Approach 1:
Different levels of concept units are assigned different weights and qualities in the ranking process. Extension units (specific concepts like 'Empire State Building') are prioritized over associated units (general concepts like 'building'), which are prioritized over related units (broader concepts like 'skyscraper'). This local differentiation of quality allows the system to return fewer, higher-quality results that better match user intent.
Data Source
AI summary
Techniques are provided for reranking and increasing the relevance of the results Internet searches. A search query is initially parsed into individual units. Each unit corresponds to one or more words that represents a natural concept. A concept network is analyzed to locate concepts that are related to the units in the search query. Particular concepts are selected from the concept network. Independent Internet searches are performed for each of the selected concepts. The search results from these searches are then compared to units in the original search query and ranked according to their relevance to the original search query.