Image Content Keyword Extraction for Ad Relevance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining relevant keywords for content items on web pages, such as advertisements, often rely on textual content and do not effectively utilize image content, leading to suboptimal matching with user requests and reduced predictability in ad impressions.
Innovation Solution
A method that identifies and evaluates image content on web pages to determine keywords based on pixel data and labels associated with the images, using an evaluation engine to assess relevance and adjust existing keywords or suggest new ones for better matching with user requests.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If keywords are determined based only on textual content, then the method is simple and fast, but the relevance matching with user requests is suboptimal
Solution Approach 1:
The patent transitions from one-dimensional textual keyword extraction to two-dimensional analysis by incorporating image pixel data. The system evaluates both text and image dimensions to generate comprehensive keywords, using image processing to extract visual content information that complements textual data for better overall relevance matching.
Solution Approach 2:
The patent introduces an image evaluation engine as an intermediary component that bridges the gap between raw image data and keyword generation. This intermediary processes pixel data, identifies visual content, and translates it into meaningful keywords that work alongside text-derived keywords to improve matching accuracy.
2Measurement precision
If image pixel data is evaluated to determine keywords, then keyword relevance is improved, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary image evaluation and keyword extraction in advance, before the actual content matching process. By pre-processing images and generating keywords upfront, the system reduces real-time processing requirements and enables faster content delivery when users access the information.
Solution Approach 2:
The patent selectively applies image evaluation only to images that meet certain criteria (e.g., prominence thresholds, relevance to content type), rather than processing every image uniformly. This partial action approach optimizes resource allocation by focusing computational efforts on images most likely to contribute meaningful keywords.
Data Source
AI summary
Methods, systems, and apparatus, include computer programs encoded on a computer-readable storage medium, for determining keywords for an image that supports an overlay content item. A method includes identifying, using one or more processors, an image that is to support an overlay content item, the image being presented on a web site and including a portion that is designated as being enabled to receive and display the overlay content item; evaluating pixel data associated with the image including determining one or more labels that are associated with content included within the image; and determining one or more keywords for the image based at least in part on the one or more labels.