Image-Based Ad Selection for Low-Context Mobile Requests
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Computing systems face challenges in selecting relevant content items due to minimal context availability, particularly in mobile applications.
Innovation Solution
Utilizing image keywords derived from user-uploaded or selected images to identify and select ads, incorporating image analysis to determine similarity scores and adjust weights based on image age, ensuring context is provided even in low-context scenarios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If traditional ad selection methods are used with minimal context, then device complexity is reduced, but ad relevance to user interests deteriorates
Solution Approach 1:
The system performs preliminary image analysis and keyword extraction from user photos before ad selection occurs. Image keywords are derived and stored in advance, allowing the ad selection process to leverage pre-computed contextual information without adding complexity during the actual ad serving decision moment.
Solution Approach 2:
Image keywords serve as an intermediary between user content (photos) and ad selection. Instead of directly analyzing images during ad selection, the system uses extracted keywords as a mediator to bridge the gap between minimal context and relevant ad targeting, improving accuracy without requiring complex real-time image processing.
2Reliability
If image analysis is incorporated to derive keywords for ad selection, then ad relevance to user interests is improved, but processing time and computational resources increase
Solution Approach 1:
Image keyword extraction is performed in advance when users upload or capture photos, rather than during the ad selection process. This preliminary action stores contextual information that can be quickly retrieved and used for ad targeting, significantly reducing processing time when ads need to be served.
Solution Approach 2:
The system segments the ad selection process into distinct phases: image analysis and keyword extraction occur separately during photo upload, while ad selection uses these pre-computed keywords. This segmentation allows computationally intensive image processing to be decoupled from time-sensitive ad serving operations.
3Measurement precision
If image keywords are used with scoring weights based on image age, then ad targeting precision is improved, but system complexity increases
Solution Approach 1:
The system changes the parameter of keyword scoring by introducing time-based weight adjustments. Image keywords are assigned weights that decrease according to image age, allowing the system to prioritize recent user interests while still considering historical context. This parameter change improves measurement precision with a relatively simple temporal decay mechanism.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for content presentation. In one aspect, a method includes receiving a request for a content item to be presented to a user; for a collection of content items, identifying one or more candidate content items including using image keywords derived from images associated with the user to identify the one or more candidate content items; selecting a content item responsive to the content request from the candidate content items; and sending the selected content item for presentation to the user.