Auto-Refresh Ad Detection via Weibull Distribution Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automatic refreshing of advertisements on web pages leads to wasted advertising budget as it cannot be distinguished from user behavior, and existing methods struggle to identify and prevent auto-refreshed ads, resulting in ineffective ad engagement and budget misallocation.
Innovation Solution
A method and system that analyze advertisement delivery information to identify deviations in display time patterns, fitting a Weibull distribution to determine if ads were refreshed automatically, allowing for differentiation from user interactions and providing advertisers with insights to optimize ad spending by inhibiting further ad displays on auto-refreshed pages.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If automatic refresh command is used to update advertisements, then advertisement freshness and timeliness are improved, but advertising budget is wasted due to loading impressions never viewed by users
Solution Approach 1:
The system implements feedback by monitoring advertisement display patterns and user behavior, then using this information to identify auto-refresh commands and adjust ad delivery accordingly. The feedback loop detects when ads are refreshed automatically versus user-initiated, allowing the system to learn and optimize budget allocation based on actual viewing patterns.
Solution Approach 2:
The system changes parameters by analyzing display time distributions and identifying deviations that correlate with auto-refresh commands. By detecting patterns in advertisement display durations and frequencies, the system can distinguish between automatic and user-initiated refreshes, then adjust ad delivery parameters to prevent waste.
2Adaptability or versatility
If auto-refresh command is used to serve multiple advertisements, then advertisement variety is improved, but user engagement time is insufficient due to very short refresh times
Solution Approach 1:
The system uses feedback to monitor how long users actually engage with advertisements before page refresh occurs. By analyzing the distribution of display times and comparing against expected user behavior patterns, the system can identify when refreshes are happening too quickly for meaningful engagement, then adjust delivery strategies accordingly.
3Productivity
If auto-refresh command is used to update advertisement impressions, then advertisement update frequency is improved, but brand consistency is lost due to replacement with competing company ads
Solution Approach 1:
The system implements feedback by tracking advertisement sequence patterns and detecting when auto-refresh commands cause unwanted brand switching. By monitoring the sequence and timing of ad displays, the system can identify patterns where competing ads are shown in rapid succession, then adjust delivery to maintain brand consistency while preserving update frequency.
4Measurement precision
If source code inspection is attempted to identify auto-refresh commands, then detection accuracy is improved, but system complexity increases due to obfuscated source code
Solution Approach 1:
The system replaces the mechanical approach of source code inspection with a behavioral analysis approach. Instead of parsing and interpreting HTML/JavaScript code (which is complex and unreliable due to obfuscation), the system monitors actual advertisement display behavior, timing patterns, and user interactions to infer the presence of auto-refresh commands, achieving accurate detection with simpler mechanics.
Data Source
AI summary
Methods, systems, and media for identifying automatically refreshed advertisements are provided. In some embodiments, a method for modifying advertisement spending is provided, the method comprising: receiving advertisement delivery information associated with a plurality of advertisements displayed on a web page; generating a distribution of an amount of time that the plurality of advertisements were displayed on the web page using the advertisement delivery information; identifying a deviation in the generated distribution; determining whether the deviation correlates to an automatic refresh command performed by one or more browser applications; and providing an indication corresponding to the plurality of advertisements that were displayed on the web page in response to the automatic refresh command based on the determination.