Multi-Bid Ad Selection via Graph Connectivity Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional online systems' auctions for sponsored content do not account for varying compensation willingness from different users, leading to inefficient selection processes.
Innovation Solution
The online system maintains multiple ad requests with bid amounts specifying compensation for each user, creating a graph of connections to select the most appropriate ad request based on bid amounts and determining the price by analyzing recent connections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional auctions select sponsored content based on maximum bid amount only, then the selection process is simple, but it does not account for varying compensation willingness from different users leading to inefficient selection
Solution Approach 1:
The patent segments the bid amount into multiple components: a base bid amount and additional bid amounts associated with different alternative users. This segmentation allows the system to consider varying compensation willingness from different users while maintaining a structured auction mechanism that processes each component systematically through graph-based connectivity analysis.
Solution Approach 2:
The patent transitions from a single-dimensional bid amount to a multi-dimensional bid structure by introducing additional bid amounts for different alternative users. This dimensional expansion enables the system to evaluate compensation willingness across multiple user dimensions simultaneously, improving selection efficiency without overwhelming complexity.
2Productivity
If the system maintains multiple ad requests with multiple bid amounts for each user, then selection efficiency improves by considering different compensation values, but the complexity of managing and processing these bid amounts increases
Solution Approach 1:
The patent introduces a graph data structure as an intermediary to manage and process the complex relationships between ad requests, users, and bid amounts. The graph visualizes connectivity between entities, transforming the complex management task into a systematic graph traversal and analysis process that identifies optimal ad requests based on bid amounts and user relationships.
Solution Approach 2:
The patent changes the parameter representation from simple scalar bid amounts to structured objects containing multiple bid amounts associated with different alternative users. This parameter transformation enables systematic processing through graph algorithms while maintaining the ability to consider varying compensation values for efficient selection.
3Measurement precision
If conventional auctions use simple maximum bid selection, then processing is fast and simple, but pricing does not reflect the true competition and collaboration dynamics between users
Solution Approach 1:
The patent performs preliminary actions by pre-establishing the graph structure and pre-processing bid amount data before the actual selection process. This allows the system to quickly traverse and analyze pre-organized data structures during selection, achieving accurate pricing that reflects user competition and collaboration dynamics without excessive processing time.
Solution Approach 2:
The patent replaces the simple mechanical maximum bid selection mechanism with a graph-based analytical system. This substitution enables more precise pricing by analyzing connectivity and bid amount relationships across the graph, capturing true competition and collaboration dynamics while maintaining reasonable processing speeds through efficient graph algorithms.
4Measurement precision
If the system analyzes all bid amounts and connections to determine pricing, then pricing accuracy improves, but the computational burden and processing time increase
Solution Approach 1:
The patent applies partial action by selectively analyzing bid amounts and connections based on graph connectivity thresholds and relevance criteria. Rather than processing all possible bid amounts uniformly, the system focuses computational resources on the most relevant connections and users, achieving sufficient pricing precision without excessive computational energy consumption.
Solution Approach 2:
The patent incorporates feedback mechanisms that allow the system to adjust its analysis depth and scope based on preliminary results and computational constraints. This feedback loop enables the system to optimize the balance between pricing precision and energy consumption by adapting the extent of bid amount and connection analysis to actual needs and available resources.
Data Source
AI summary
To select content for presentation to a viewing user, an online system maintains multiple bid amounts associated with various content items. Content items are each associated with multiple bid amounts, with each bid amount specifying an amount of compensation to the online system and identifying a user, so the bid amount identifies an amount of compensation to the online system for selecting the content item in place of content items associated with the identified user. Based on users identified by bid amounts in various content items, the online system generates connections between content items and determines a group of content items including content items connected to each other content item in the group. Using connections between content items in the group, the online system selects a content item and determines a price charged to a user associated with the selected content item.