Semantic Search Weighting via Lexical Database
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional syntactic searches in corpora of terms face challenges with homonymic terms, irrelevant results, and failure to retrieve related terms due to reliance on syntactic similarities, especially when dealing with multi-term search expressions and target expressions, which complicates semantic similarity determination.
Innovation Solution
Assigning concepts to each term in a lexical database and calculating semantic similarity between search and corpus expressions by determining distances from common super concepts, weighting terms by significance, and aggregating similarities to rank relevant results based on a predefined threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional syntactic search methods are used, then search speed and simplicity are maintained, but search precision and relevance deteriorate due to homonymic terms and inability to capture semantic relationships
Solution Approach 1:
The patent introduces a lexical database as an intermediary between the search expression and corpus expressions. This database stores pre-computed semantic relationships (synonyms, homonyms, hypernym-hyponym relations) that mediate the matching process, enabling semantic search without requiring complex real-time semantic analysis
Solution Approach 2:
The system performs preliminary actions by pre-processing and storing semantic relationships in a lexical database before search execution. Concepts, synonyms, and hierarchical relationships are established in advance, allowing the search algorithm to simply query pre-computed data rather than performing complex semantic analysis during search operations
2Measurement precision
If semantic similarity calculation is performed on multi-term expressions, then search relevance is improved, but computational complexity and processing time increase significantly
Solution Approach 1:
The patent segments multi-term expressions into individual terms and evaluates semantic similarity term-by-term. Instead of calculating semantic similarity for the entire multi-term expression as a unit, the system breaks it down into component terms, calculates similarity for each term independently using the lexical database, and aggregates the results. This segmentation dramatically reduces computational complexity while maintaining semantic accuracy
3Reliability
If strict syntactic matching is applied, then false positives from homonymic terms are reduced, but relevant results using synonyms or related terms are missed
Solution Approach 1:
The lexical database serves multiple functions: it stores synonyms, homonyms, hypernym-hyponym relationships, and other semantic relationships in a unified structure. This universal database handles diverse semantic relationships (synonymy, homonymy, hierarchy) through a single mechanism, enabling the search system to adapt to different types of semantic variations without requiring separate processing paths for each relationship type
Data Source
AI summary
A method and system for conducting semantically weighted searches in a governed corpus of terms is provided. A search expression having a plurality of terms for performing a search in the governed corpus of terms is received. The governed corpus of terms comprises a plurality of corpus expressions each comprising a plurality of terms, each term within the governed corpus of terms being associated precisely with a single concept within a lexical database. At least one concept of the lexical database is assigned to each term in the search expression based on a syntactical analysis. A semantic similarity is calculated between pairs of concepts of the search expression and one of the corpus expressions. A total semantic similarity is calculated between the search expression and the one of the corpus expressions by aggregating the semantic similarities of the pairs of concepts based on an order of significance of the terms.