AI-Personalized Search Ranking Through Secondary Query Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current online search engines face issues with over-inclusive and irrelevant search results, distorted rankings, and lack of higher-level categorization, leading to inefficiencies in data transmission, storage, and user time consumption.
Innovation Solution
An AI-driven system that personalizes search results by generating secondary queries, utilizing AI servers like GPT-4 and embedding models to categorize and rank results based on user-specific parameters, providing relevant, higher-level classifications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If keyword-based search is used to retrieve documents and websites, then search results accurately correspond to search terms, but the results are over-inclusive and contain too many irrelevant results that users cannot review in reasonable time
Solution Approach 1:
The patent introduces an intermediary AI system that acts as a mediator between the keyword search and the user. This AI intermediary receives the initial search results, analyzes them using machine learning models, and filters/ranks them to present a refined subset to the user. This resolves the contradiction by maintaining accurate correspondence through the AI's intelligent filtering while reducing review time by eliminating irrelevant results.
Solution Approach 2:
The system performs preliminary action by pre-processing and pre-ranking search results using AI models before presenting them to the user. The AI system analyzes and evaluates results in advance, organizing them by relevance and importance, so that users receive pre-filtered, high-quality results ready for immediate review without needing to manually sift through large volumes of data.
2Ease of operation
If search results are ranked by third-party links or bidding processes, then results are returned in organized order, but rankings become distorted with sponsored results going to highest bidders rather than most relevant
Solution Approach 1:
The patent implements feedback mechanisms where the AI system continuously learns from user interactions with search results. By analyzing which results users click, spend time on, or disregard, the system refines its ranking algorithms to better reflect actual relevance rather than commercial bidding. This creates a feedback loop that progressively improves ranking accuracy while maintaining organized presentation.
Solution Approach 2:
The system uses self-service by employing AI models that automatically evaluate and rank results based on intrinsic quality metrics rather than external commercial factors. The machine learning system independently assesses result relevance, credibility, and usefulness, enabling the search engine to organize results by merit rather than by bidding, thus resolving the distortion caused by sponsored placements.
3Quantity of substance
If large databases are used to store and transmit comprehensive search results, then more complete information is available, but data transmission and storage requirements increase, overloading network and Internet systems
Solution Approach 1:
The patent applies extraction by using AI to identify and extract only the most relevant and valuable results from the comprehensive database, presenting a curated subset to the user. Instead of transmitting and displaying all available results, the system extracts the essential few that truly matter, maintaining information quality while dramatically reducing data transmission and storage requirements, thus lowering network and system load.
Solution Approach 2:
The system employs partial action by providing a carefully selected portion of the complete search database rather than the entire dataset. The AI determines the optimal number and type of results to present, delivering just enough information to satisfy the user's needs without the excess that would burden the system. This partial presentation maintains completeness of relevant information while minimizing energy consumption and system overload.
Data Source
AI summary
A computer server system and method are disclosed for personalization and customization of search results and search result rankings, such as for Internet searching. A representative server system comprises: a network interface to receive a primary query and transmit secondary queries and search results; and one or more processors configured to generate the secondary queries; to extract or transform responses into text variables; to use trained, supervised multi-class neural networks to classify the text variables to form initial classifications or categories and combine the initial classifications or categories to form resulting classifications or categories; to filter and rank the resulting classifications or categories; and to use the filtered and ranked resulting classifications or categories to generate and output the personalized search results and search result rankings, the personalized search results and search result rankings comprising one or more associated classifications or categories corresponding to the primary query.