Semantic Fuzzy Search for Multidimensional Database Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Search engines in network databases struggle with discrete variables and Boolean logic, leading to inaccurate results by either excluding or improperly incorporating these variables, which are not suitable for handling open sets in multidimensional spaces.
Innovation Solution
Implementing a fuzzy search engine that converts search queries into sequences of numbers based on semantic content, allowing for open set criteria in a multidimensional space to identify and rank text strings by similarity, providing links to relevant media files.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional search engines use compact criteria for searching, then the search process is simple and fast, but the search accuracy deteriorates when handling discrete variables and Boolean logic
Solution Approach 1:
The patent transforms search queries from discrete text strings into continuous numerical vectors through embedding models. This parameter transformation allows the system to handle discrete variables (like Boolean logic and categorical data) by converting them into continuous multidimensional space, where similarity can be measured using distance metrics. The embedding process maps discrete concepts into continuous vectors, enabling accurate handling of previously problematic data types while maintaining search efficiency through vector operations.
Solution Approach 2:
The patent moves the search problem from one-dimensional text matching into N-dimensional vector space. By representing both queries and database entries as vectors in multidimensional space, the system can capture semantic relationships and handle discrete variables through geometric operations. This dimensional transformation enables the use of open set criteria (distance-based similarity) instead of rigid Boolean logic, improving accuracy while maintaining computational efficiency through vector operations.
2Device complexity
If discrete variables are excluded from search analysis, then the search process remains simple, but the search results become inaccurate and are disregarded by users
Solution Approach 1:
The patent replaces traditional mechanical text-matching mechanisms with neural network-based embedding models. Instead of using rigid string comparison and Boolean logic operations, the system uses learned vector representations that naturally handle discrete variables. The embedding models automatically process categorical data, Boolean values, and other discrete types by mapping them into continuous space, eliminating the need for special-case handling while improving result reliability.
3Ease of operation
If traditional Boolean logic is used for searching, then the search criteria are precise and clear, but the search results fail to capture semantic context and relevance
Solution Approach 1:
The patent transforms Boolean search criteria into continuous similarity scores through vector operations. Instead of rigid True/False matching, the system computes cosine similarity or distance metrics between query vectors and database vectors, producing graded relevance scores. This parameter transformation preserves the clarity of search intent while capturing nuanced semantic relationships that Boolean logic cannot express, preventing loss of semantic context in the search results.
Data Source
AI summary
A method for displaying a search result to a user in a client device is provided. The method includes receiving, in a server, a search query from a user, the search query including a string of text characters. The method also includes converting the search query into a sequence of numbers based on a semantic content of the string of text characters, the sequence of numbers defining a vicinity in a multidimensional space, identifying a text string associated with a point within the vicinity in the multidimensional space, ranking the text string according to a similarity value with the search query, and providing a link to a media file associated with the text string to the user as a search result. A system, a memory storing instructions which, when executed by a processor cause the system to perform the above method, and the processor are also provided.