Concept Disambiguation Module for Search Relevance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current information retrieval systems struggle to accurately determine user intent from ambiguous keywords, leading to irrelevant search results and inefficient targeted advertisement.
Innovation Solution
The system expands content nodes into groupings of concepts and phrases, analyzing these groupings to provide relevant content by projecting keywords and phrases onto a conceptual map, thereby refining user intentions and identifying relevant content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If keyword expansion is performed using various resources, then the quantity of search results increases, but the relevance of search results deteriorates due to ambiguous keyword meanings
Solution Approach 1:
The patent introduces an intermediary component called a 'concept disambiguation module' that acts as a mediator between keyword expansion and search result generation. This module resolves ambiguous keyword meanings by analyzing contextual signals from multiple resources (user profile, browsing history, current page content) and selecting the most appropriate concept interpretation before generating search results, thereby maintaining relevance while enabling comprehensive keyword expansion
Solution Approach 2:
The system dynamically changes the interpretation parameter of keywords based on contextual analysis. By evaluating multiple contextual signals (user profile attributes, browsing history patterns, current page content), the system adjusts which concept expansion path to follow for ambiguous keywords, transforming a static keyword matching process into a dynamic, context-aware parameter selection process
2Device complexity
If multiple databases are processed separately for search results and advertisements, then the complexity of processing is reduced, but the accuracy of targeted advertisement deteriorates
Solution Approach 1:
The patent merges the processing of search results and targeted advertisements by introducing a unified 'concept extraction and matching' stage that processes both information types simultaneously from the same contextual signals. The concept disambiguation module generates a single set of resolved concepts that feed into both search result generation and advertisement targeting, eliminating redundant processing while improving advertisement accuracy through better concept understanding
Solution Approach 2:
The concept disambiguation module serves multiple functions: it disambiguates keywords for search results, extracts relevant concepts for advertisement targeting, and provides contextual understanding for both processes. This multi-functional component reduces overall system complexity by consolidating what would otherwise require separate processing pipelines
3Measurement precision
If context analysis is added to determine user intent, then the accuracy of search results improves, but the complexity of the system increases
Solution Approach 1:
The patent segments the context analysis process into distinct, modular components: a 'context signal extraction module' that collects signals from user profile, browsing history, and current page; a 'concept disambiguation module' that processes these signals; and a 'concept selection module' that outputs resolved concepts. This segmentation allows each component to perform a specific function with well-defined inputs and outputs, reducing overall system complexity through modular design
Solution Approach 2:
The system performs preliminary context analysis and concept disambiguation before the main search and advertisement generation processes. By pre-resolving ambiguous keywords and selecting the most relevant concepts upfront, the system simplifies subsequent processing steps, as downstream components receive already-disambiguated concepts rather than having to perform their own complex analysis
Data Source
AI summary
Discovering relevant concepts and context for content nodes to determine a user's intent includes identifying one or more concept candidates in a content node based at least in part on one or more statistical measures, and matching concepts in a concept association map against text in the content node. The concept association map represents concepts, concept metadata, and relationships between the concepts. The one or more concept candidates are ranked to create a ranked one or more concept candidates based at least in part on a measure of relevance. The ranked one or more concept candidates is expanded according to one or more cost functions. The expanded set of concepts is stored in association with the content node.