Interactive AI Search Using Multi-LLM Query Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing product search systems, particularly in niche markets like semi-custom homes, struggle with integrating qualitative and quantitative criteria, leading to inefficient and overwhelming search results, especially for non-local buyers.
Innovation Solution
An AI-based system utilizing a network of Large Language Models (LLMs) to process user queries, separating qualitative and quantitative criteria, and generating structured and unstructured queries for comprehensive search results, enhanced by predictive models and visual input processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If major search engines aggregate products from numerous smaller suppliers, then search coverage is improved, but result overwhelming increases making it difficult for users to narrow down options
Solution Approach 1:
The search system segments the overwhelming results by automatically categorizing products into distinct categories and subcategories based on their attributes. It further divides the search process into multiple stages: initial broad search, intermediate filtering by user preferences, and final refined results, making the vast amount of data manageable and organized rather than overwhelming.
Solution Approach 2:
The system implements feedback mechanisms where it analyzes user interactions with search results and automatically adjusts subsequent search queries. It monitors user behavior patterns, preferences, and feedback signals to progressively refine search results, transforming a static search process into a dynamic, adaptive system that learns from user responses to reduce information overload.
2Ease of operation
If basic search features are provided on smaller supplier websites, then ease of use is improved, but search capability quality deteriorates
Solution Approach 1:
The search system acts as an intermediary between simple supplier websites and sophisticated search requirements. It connects basic search interfaces on supplier sites with advanced search capabilities through a mediating layer that aggregates, enriches, and processes search results, allowing users to access enhanced search functionality without requiring complex interfaces on individual supplier websites.
Solution Approach 2:
The system provides universal search capabilities that work across multiple supplier websites with different technical capabilities. It creates a unified search interface that handles diverse data formats and website structures, enabling sophisticated search functionality regardless of the underlying supplier website's complexity or technical sophistication.
3Measurement precision
If comprehensive search parameters including qualitative and quantitative criteria are added, then search accuracy is improved, but system complexity increases
Solution Approach 1:
The search system segments search parameters into distinct categories: quantitative parameters (price, dimensions, specifications) and qualitative parameters (design aesthetics, material quality, craftsmanship). Each parameter type is processed and filtered separately through dedicated algorithms, reducing the complexity of handling all parameters simultaneously while maintaining comprehensive search accuracy.
Solution Approach 2:
The system performs preliminary processing of search parameters by pre-categorizing, pre-validating, and pre-organizing search criteria before the actual search execution. It pre-processes product attributes, establishes parameter hierarchies, and prepares filtering rules in advance, reducing the computational complexity during real-time search operations while maintaining comprehensive search capabilities.
Data Source
AI summary
A system for interactive searches based on user queries data and a plurality of Large Language Models (LLMs) including a processor of a Human-Machine Interface (HMI) server node configured to host a network of LLMs and at least one machine learning module (ML) and connected to at least one user-entity node over a network and a memory on which are stored machine-readable instructions that when executed by the processor, cause the processor to: receive a search request input data from the at least one user-entity node; evaluate, by a first dedicated LLM, relevance of the search request input data by discerning between primary and secondary information; responsive to evaluation by the first dedicated LLM, derive classifying features from the primary and secondary information and generate a feature vector based on the classifying features; ingest the feature vector into the ML module configured to extract additional search parameters from a predictive search model based on historical search data associated with the at least one user-entity node; dissect, by a second dedicated LLM, the search request input data and the additional search parameters to separate the data into qualitative and quantitative criteria elements based on the primary and the secondary information; transform, by a third dedicated LLM, the quantitative criteria elements into structured queries for a database; process, by a fourth dedicated LLM, the qualitative criteria elements by searching through a semi-structured data repository; and synthesize, by the fifth LLM, processed search findings into a succinct human-language summary.