Social Network Ad Targeting via TV Stream Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing social media platforms lack the ability to effectively correlate advertisements with video content, isolating social media content from the events and time-based media they refer to, which hinders targeted advertising efforts.
Innovation Solution
A system that monitors television streams to detect advertisements and segments video content, creating mappings between users and advertisements they are likely to have seen, allowing for targeted messaging based on demographic, content, and time criteria.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If social media content items reference time-based media events, then the content can be associated with specific TV shows and advertisements, but the content items remain isolated from the events they refer to
Solution Approach 1:
The patent introduces an intermediary system comprising a video processing module, event detection module, and mapping module that acts as a mediator between TV broadcast streams and social networking system. This intermediary correlates advertisement events with social media content by detecting airings in video streams, identifying referenced events in social media posts, and creating mappings between users and advertisements, thereby resolving the isolation problem without requiring direct integration between disparate systems
Solution Approach 2:
The patent replaces manual or mechanical correlation methods with automated electronic detection and processing systems. The video processing module automatically detects advertisement airings, the event detection module identifies referenced events in social media content, and the mapping module automatically creates correlations, substituting automated computational processes for manual association efforts
2Measurement precision
If advertisers send targeted messages to SNS users based on TV ad airings, then ad relevance improves, but the ability to accurately detect and map ad exposure becomes challenging
Solution Approach 1:
The patent implements feedback mechanisms where the mapping module continuously refines correlations between users and advertisements based on detected social media engagement with referenced events. The system monitors which users create or engage with content referencing specific TV shows or advertisements, and uses this feedback to improve the accuracy of ad exposure detection and targeting over time
Solution Approach 2:
The patent performs preliminary detection and mapping of advertisement airings and social media event references before initiating targeted messaging campaigns. The video processing module detects ad airings in advance, the event detection module identifies relevant social media content beforehand, and the mapping module pre-establishes user-ad associations, enabling accurate targeted messaging without requiring real-time measurement during actual ad delivery
Data Source
AI summary
Targeted messages are sent to users of social networking system (SNS) based on the detection of airings of advertisements in time-based media. This approach allows advertisers to leverage their investment in, for example, television advertising by sending advertisements to SNS users who have likely seen the advertisements within shows that they have expressed interest in in the context of a SNS. Time-based media streams are analyzed to detect the airings of advertisements within those streams. In one embodiment, SNS content items are received regarding individual users. Based on references in those content items between users and their interests, targeted messages may be sent to users. In another embodiment, targeted messages are sent to SNS users based on the airing of advertisements and targeting criteria provided by advertisers.