Automated Thumbnail Selection for Text Ads

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Selecting suitable images to accompany textual content items, such as advertisements, is a time-intensive manual process that can be inefficient and resource-intensive, especially when no images are specified by the content provider.

Innovation Solution

An automated method that identifies and selects relevant non-textual content items, like images, from landing pages, advertising entities, or verticals based on relevance measures, and stores this data for display with textual content items, allowing for hierarchical prioritization and collision avoidance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual image selection is performed, then image relevance to textual content can be ensured, but time consumption and resource investment increase significantly

Engineering Contradiction:

Improveimage relevanceVSAvoidtime consumption

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary actions by pre-fetching images from landing pages and pre-computing their relevance scores to textual content before actual display needs arise. This allows the automated system to have relevant images ready when needed, eliminating the need for manual real-time selection while maintaining high relevance quality.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system enables self-service by automatically selecting and matching images to textual content items without requiring manual intervention. The automated relevance computation and selection process allows the system to serve itself, eliminating the time-consuming manual search and review process while maintaining consistent quality standards.

Inventive Principle:

Principle #25Self-service

2Quantity of substance

If no images are specified by content providers, then resource investment by advertisers is reduced, but the quality and relevance of displayed content decreases

Engineering Contradiction:

Improveresource investmentVSAvoidcontent quality

Core Design Contradiction:

Quantity of substanceVSManufacturing precision

Solution Approach 1:

The system introduces an intermediary automated image selection service that bridges the gap between advertisers who don't specify images and high-quality content display requirements. This intermediary service automatically fetches and selects appropriate images from landing pages, ensuring content quality is maintained without requiring additional resource investment from advertisers.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system changes the parameter of image selection from manual advertiser specification to automated algorithmic selection based on relevance computation. This parameter change allows the system to maintain high content quality (relevance) while reducing the resource investment required from advertisers, as the automated system handles the selection process efficiently.

Inventive Principle:

Principle #35Parameter changes

3Productivity

If automated image selection is implemented, then time and resources are saved, but system complexity increases

Engineering Contradiction:

Improveselection efficiencyVSAvoidsystem complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The system segments the image selection process into distinct modular components: image fetching from landing pages, relevance computation between images and textual content, and final selection based on computed scores. This segmentation allows each component to be independently optimized and managed, reducing overall system complexity while maintaining high productivity.

Inventive Principle:

Principle #1Segmentation

4Measurement precision

If multiple images are selected from landing pages, then more relevant images can be provided, but processing time and computational resources increase

Engineering Contradiction:

Improverelevance accuracyVSAvoidcomputational resources

Core Design Contradiction:

Measurement precisionVSPower

Solution Approach 1:

The system applies partial action by selecting only the top N most relevant images based on computed relevance scores, rather than processing and selecting all available images. This partial selection approach maintains high relevance accuracy by focusing computational resources on the most promising candidates while avoiding the excessive processing that would result from evaluating all possible images.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS9183577B2Selection of images to display next to textual content

Publication Date: 2015.11.10 GOOGLE LLC

AI summary

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for selecting thumbnail images to display with text advertisements. In one aspect, a method includes identifying landing page non-textual content items from a landing page to which the textual content item links and which a user device requests when the textual content item is selected at the user device; for each landing page non-textual content item, determining a relevance measure that measures the relevance of the landing page non-textual content item to the content of the landing page; selecting one or more of the landing page non-textual content items for display with the textual content item based on the relevance measures of the landing page non-textual content items; and storing, in a data storage system, data associating the selected landing page non-textual contents with the textual content items.