Gesture Search Query Refinement via Character Recognition

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing search technologies do not effectively utilize user-drawn patterns on a display for real-time search query refinement, leading to inefficient search result updates and user experience.

Innovation Solution

A computer-implemented method and system that receives pixel maps of user-drawn characters, recognizes them, and updates search queries by combining new and previously recognized characters, eliminating infrequent combinations, to provide dynamic and relevant search results.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the search system processes and updates search results in real-time as users draw patterns, then search relevance and user experience are improved, but system response time and processing speed may be worsened

Engineering Contradiction:
Improvesearch result relevanceVSAvoidsystem response time
Core Design Contradiction:
Measurement precisionVSSpeed

Solution Approach 1:

The system pre-processes and stores character recognition models and search indexes in advance, enabling faster real-time query processing. By preparing recognition patterns and search databases beforehand, the system can quickly match user-drawn patterns against pre-computed results without performing heavy processing during the actual search interaction.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If the system recognizes and processes user-drawn patterns to generate search queries, then search accuracy is improved, but device complexity and processing overhead are worsened

Engineering Contradiction:
Improvecharacter recognition accuracyVSAvoidpattern processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The pattern recognition process is divided into discrete steps: capturing touch coordinates, generating pixel maps, comparing against character templates, and forming search queries. This segmentation allows each processing stage to be optimized independently and enables the system to handle complex recognition tasks through a series of simpler, manageable operations.

Inventive Principle:
Principle #1Segmentation

3Adaptability or versatility

If the system maintains and processes historical search data for query refinement, then search personalization is improved, but data management complexity and memory usage are worsened

Engineering Contradiction:
Improvesearch personalizationVSAvoiddata storage requirements
Core Design Contradiction:
Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The system extracts only the essential elements from historical search data—specifically previously recognized characters and search patterns—rather than storing complete search histories. By extracting and retaining only the critical information needed for query refinement, the system achieves personalization without proportionally increasing data storage requirements.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP2771816B1Gesture-based search
Publication Date: 2019.08.07 GOOGLE LLC
  • EP2771816B1 patent drawingFigure 1
  • EP2771816B1 patent drawingFigure 2
  • EP2771816B1 patent drawingFigure 3

AI summary

In general, the subject matter described in this specification can be embodied in methods, systems, and program products for performing searches with gesture-based input. A search system receives gesture data corresponding to one or more characters that have been drawn on a display of a client device. The search system recognizes the one or more characters that correspond to the gesture data. The search system formulates a search that includes the one or more characters as a query term. The search system communicates to the client device one or more search results for the search, and data identifying the one or more characters.