Local-Search Ad Auction Algorithm for Non-Uniform Space Pricing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing online auction platforms struggle to accurately calculate the worth of non-uniform advertisement spaces, leading to inefficient bid pricing and advertisement placement, as they were designed for uniform advertisement spaces with predetermined characteristics.
Innovation Solution
The Search Advertising Auction (SAA) tool implements a local-search based algorithm that iteratively evaluates different advertisement configurations and uses a separable click-probability model to compute accurate bid prices for non-conforming advertisement spaces, optimizing both allocation and pricing within strict timing constraints.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional online auction platforms are used for uniform advertisement spaces, then bid pricing can be calculated accurately and efficiently, but they cannot handle non-uniform advertisement spaces with varying sizes and characteristics
Solution Approach 1:
The patent applies local quality by treating each advertisement space as a unique entity with its own specific attributes (size, position, content type) rather than assuming uniformity. The system calculates worth individually for each space based on its local characteristics, enabling accurate pricing for non-uniform spaces while maintaining computational efficiency through localized evaluation rather than global recalculation.
2Measurement precision
If complex algorithms are used to evaluate different advertisement configurations for non-uniform spaces, then more accurate bid prices can be computed, but computation time increases beyond real-time auction constraints
Solution Approach 1:
The system performs preliminary actions by pre-establishing the evaluation framework and worth calculation models for different advertisement space types before auctions occur. During the actual auction, it only needs to apply pre-computed evaluation criteria to specific configurations, dramatically reducing computation time while maintaining accuracy. This allows complex evaluation of non-uniform spaces to be completed within real-time constraints.
3Productivity
If uniform advertisement space assumptions are made, then auction processing is simple and fast, but the system cannot accommodate newer types of supplemental content with varying attributes
Solution Approach 1:
The patent segments the advertisement space evaluation into distinct components (size factors, position factors, content type factors) that can be independently calculated and combined. This segmentation allows the system to efficiently process each attribute separately and aggregate results, maintaining high processing speed while accurately evaluating diverse non-uniform advertisement spaces with varying characteristics.
Data Source
AI summary
Systems, devices, and methods generate a more accurate search advertising auction that considers heterogeneous attributes of advertisements for inclusions into an available advertisement space. Each advertisement considered for inclusion into the available advertisement space may be further analyzed to generate a truthful estimated price for reference during an online bidding for the advertisement space.