Synonym Identification via Co-occurring List Terms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current search engines face challenges in identifying effective synonyms for query terms, leading to suboptimal search results due to the reliance on costly thesaurus-based systems and user substitution analysis, which can create false synonyms and fail to account for context-dependent term usage.
Innovation Solution
A system that identifies potential synonyms by analyzing co-occurring terms in lists, such as HTML lists, and determines correlations based on their frequency of occurrence across documents from different authors, adjusting thresholds for evaluation and filtering out non-synonymous pairs to improve query modification and expansion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If thesaurus-based systems are used to identify synonyms, then synonym identification can be performed, but the system becomes costly to construct and is restricted to one language
Solution Approach 1:
The system automatically constructs synonym information by analyzing query sessions and document correlations without requiring manual thesaurus construction. The synonym data is self-generated through computational analysis of user behavior patterns and document co-occurrence statistics, eliminating the need for costly manual thesaurus development and maintenance
Solution Approach 2:
The system processes queries in multiple languages simultaneously by analyzing cross-lingual query sessions and document correlations. The same computational framework handles synonym identification across different languages without requiring separate language-specific thesauruses, making the system universally applicable to multilingual search scenarios
2Adaptability or versatility
If user substitution analysis is used to determine synonyms, then dynamic synonym identification is possible, but false synonyms are created that lead to unrelated query results
Solution Approach 1:
The system introduces document correlation as an intermediary validation layer between user substitution and synonym confirmation. Before accepting a user substitution as a valid synonym, the system checks whether the substituted terms appear together in correlated documents. This intermediary check filters out false synonyms that would otherwise be accepted based solely on user substitution patterns
Solution Approach 2:
The system uses query session feedback to continuously refine synonym identification. By monitoring successful and unsuccessful query refinements, the system learns which substitutions lead to relevant results and which create false synonyms. This feedback loop progressively improves synonym accuracy while maintaining adaptability to new query patterns
3Productivity
If query terms are automatically modified to include synonyms, then search effectiveness improves, but the system may include false synonyms leading to irrelevant results
Solution Approach 1:
The system applies synonym expansion partially by selecting only the most highly-correlated synonyms for inclusion in modified queries. Rather than expanding all possible synonyms, the system prioritizes those with strongest document correlation evidence, applying expansion action selectively to maintain result precision while still improving search effectiveness through relevant synonym inclusion
Data Source
AI summary
One embodiment of the present invention provides a system for detecting correlations between terms. During operation, the system identifies one or more lists contained in one or more documents and identifies two terms co-occurring in the lists. The system further determines a correlation between the co-occurring terms, and places the co-occurring terms in a correlated-pair list based on the correlation.