Dynamic Ad Region Sentiment Analysis for Brand Safety
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advertisers face challenges in digital advertising as they are often unaware of opposing brand advertisements that may appear on the same webpage, leading to brand safety issues due to the dynamic nature of advertising regions, which are not adequately assessed before bidding.
Innovation Solution
A method using a machine learning classifier to analyze sentiments in dynamic advertising regions, determining similarity scores with the advertiser's brand sentiments, and associating webpages with approval or exclusion lists based on aggregate similarity scores to ensure brand safety.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If advertisers place ads in dynamic advertising regions without prior analysis, then advertising productivity is improved, but brand safety deteriorates due to unknown opposing brand advertisements
Solution Approach 1:
The system performs preliminary analysis of dynamic advertising regions before advertisers place their bids. The machine learning classifier scans and analyzes the content that will appear in dynamic ad regions, generating contextual information and sentiment scores in advance. This allows advertisers to make informed bidding decisions without manually analyzing each potential placement, thus maintaining high productivity while ensuring brand safety through prior knowledge of the advertising environment.
2Reliability
If the system analyzes all dynamic advertising regions before bidding, then brand safety is improved, but system complexity increases
Solution Approach 1:
The system replaces manual analysis mechanisms with an automated machine learning classifier. Instead of requiring human reviewers to manually examine each dynamic advertising region, the ML model automatically scans, analyzes sentiments, and generates contextual information. This substitution of mechanical/manual processes with automated intelligent systems reduces operational complexity while maintaining or improving brand safety assurance.
3Measurement precision
If the system provides detailed contextual information about dynamic content, then advertising decision quality is improved, but information processing time increases
Solution Approach 1:
The system generates contextual information about dynamic advertising regions in advance, before advertisers need to make bidding decisions. By performing the analysis beforehand and storing the results, the system eliminates the need for real-time analysis during the bidding process. Advertisers receive pre-computed contextual data including sentiment scores and content descriptions, enabling quick informed decisions without experiencing analysis delays.
Data Source
AI summary
Methods, systems, and media for providing content providers with contextual information associated with dynamic content are provided. In some embodiments, the method includes: accessing a webpage that contains at least one dynamic advertising region; receiving a plurality of brand sentiments associated with a first advertiser; identifying a position of the at least one dynamic advertising region in the webpage and at least one content item shown in proximity to the at least one dynamic advertising region; determining, using a machine learning classifier, (i) a plurality of sentiments for the at least one content item shown in proximity to the at least one dynamic advertising region, (ii) a plurality of similarity scores, wherein each similarity score is a probability that a sentiment from the plurality of sentiments for the at least one content item is similar to a sentiment from the plurality of brand sentiments, and (iii) an aggregate similarity score based on the plurality of similarity scores; and, in response to determining that the aggregate similarity score is within a first range of predetermined values, associating the webpage with an approval list associated with the first advertiser.