Advertisement Boundary Detection via Interaction Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods are inefficient in accurately determining advertisement boundaries in media content, leading to challenges in modifying or predicting advertisement placement opportunities, as they rely on incomplete or unreliable user interaction data.

Innovation Solution

A system that analyzes user interactions such as fast-forwarding, tuning out, and volume changes to identify scene changes and determine advertisement boundaries by assigning weights based on content genre, thereby accurately identifying advertisement start and end points.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional methods are used to determine advertisement boundaries, then the process is simpler, but the accuracy of advertisement location determination is poor

Engineering Contradiction:
Improveadvertisement boundary accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the advertisement detection process into multiple independent analysis components: scene change detection, user interaction event classification (fast-forward, tune-out, pause, volume changes), and weighted scoring mechanisms. Each component processes specific aspects of the data independently, then combines results to determine advertisement boundaries, improving accuracy while maintaining manageable system complexity

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary analysis layer that processes user interaction events and scene change data before determining advertisement boundaries. This intermediary layer applies weighted algorithms to intermediate data points, transforming raw interaction data into meaningful advertisement location indicators without requiring direct complex processing of all raw data

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If more user interaction data is collected to improve advertisement detection accuracy, then the precision improves, but the processing time and computational resources increase

Engineering Contradiction:
Improveadvertisement location accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent performs preliminary classification and filtering of user interaction events before full analysis. Scene changes are detected and categorized in advance, and user interactions are pre-sorted by type (fast-forward, tune-out, etc.), allowing the main advertisement detection algorithm to work with pre-processed data rather than raw data, reducing processing time while maintaining accuracy

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent applies weighted scoring where only certain user interactions and scene changes are given significant weight in determining advertisement boundaries. Not all data points are processed with equal depth - the system focuses computational resources on the most indicative signals, achieving accurate advertisement detection without processing every data point exhaustively

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3214821B1Determining advertisement locations based on customer interaction
Publication Date: 2019.11.06 COMCAST CABLE COMM LLC
  • EP3214821B1 patent drawingFigure 1
  • EP3214821B1 patent drawingFigure 2
  • EP3214821B1 patent drawingFigure 3

AI summary

A system may determine the locations of advertisements and advertisement placement opportunities based on user interactions. User interactions may correspond to scene changes in the content item. Events such as tune outs and fast forwarding may indicate that a content item has begun or ended, or that an advertisement section has begun or ended. By aggregating and analyzing user interactions and other events, the system may accurately determine which portions of the content item contain advertisements and advertisement placement opportunities.