Semantic Search Interface Using Filter Sets for Accuracy

Solution Overview

Problem

Conventional search engines, including those using keyword search logic, often fail to provide accurate results as they do not account for user context and the underlying meaning of data, leading to resource-intensive semantic searching in large document spaces.

Innovation Solution

A method and system for semantic searching that receives user input including seed data and a semantic input, generates a filter set of documents, and provides it to a semantic search engine to retrieve results that are semantically similar, thereby limiting the search to a subset of the document space for efficient and accurate results.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If semantic searching is performed on a large document space, then search accuracy is improved by considering user context and underlying meaning, but resource consumption and processing time increase significantly

Engineering Contradiction:

Improvesearch accuracyVSAvoidresource consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent segments the large document space into multiple smaller subsets using filter sets generated from user feedback. Instead of performing semantic searching across the entire document space, the system divides it into manageable segments (filter sets) that are more targeted and relevant to the user's information need, thereby reducing computational resources while maintaining search accuracy

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system performs preliminary actions by generating filter sets from user feedback before executing the main semantic search. These filter sets pre-process and organize document subsets in advance, so when the semantic search is executed, it operates on already-prepared, relevant document segments rather than the entire document space, reducing real-time computational burden

Inventive Principle:

Principle #10Preliminary action

2Measurement precision

If semantic searching is performed on a large document space, then search accuracy is improved by considering user context and underlying meaning, but processing time increases

Engineering Contradiction:

Improvesearch accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent segments the large document space into multiple smaller subsets using filter sets generated from user feedback. Instead of performing semantic searching across the entire document space, the system divides it into manageable segments (filter sets) that are more targeted and relevant to the user's information need, thereby reducing computational resources while maintaining search accuracy

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system performs preliminary actions by generating filter sets from user feedback before executing the main semantic search. These filter sets pre-process and organize document subsets in advance, so when the semantic search is executed, it operates on already-prepared, relevant document segments rather than the entire document space, reducing real-time computational burden

Inventive Principle:

Principle #10Preliminary action

3Productivity

If keyword search logic is used, then processing speed is maintained and resources are conserved, but search accuracy deteriorates as results may not represent what the user was searching for

Engineering Contradiction:

Improveprocessing efficiencyVSAvoidsearch accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent introduces filter sets as an intermediary between keyword search and semantic search. The filter sets are generated from user feedback and act as a mediator that guides the semantic search engine to focus on relevant document subsets. This intermediary structure allows the system to maintain processing efficiency by limiting the search scope while improving accuracy through context-aware semantic analysis on targeted segments

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS10235455B2Semantic search system interface and method

Publication Date: 2019.03.19 INNOGRAPHY

AI summary

A method for semantic searching includes receiving a user input including seed data and a semantic input at a search system. The method further includes automatically generating a filter set based on the user input, where the filter set including a plurality of documents that correspond to the seed data, and includes providing the filter set and the semantic input to a semantic search engine. The method also includes receiving a set of semantic search results from the semantic search engine based on the filter set and the semantic input. The set of semantic search results corresponds to a sub-set of the filter set that is semantically similar to the semantic input.