Dynamic Ad Selection via Content Sentiment Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for selecting advertisements during media programs often fail to consider the relevance of advertisements to the actual state of the media content, such as subjects or sentiments, leading to suboptimal engagement with viewers or listeners.

Innovation Solution

A system and method that identifies the creator of a media program by transcribing and analyzing its content, including acoustic features, and matches it with brand advertisements using machine learning models to select relevant ads based on similarity, allowing for real-time adjustment of media content to maximize engagement.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If advertisements are selected based on predicted attributes of viewers or listeners, then advertising reach is maximized, but relevance to actual media program content is not considered

Engineering Contradiction:
Improveadvertising reachVSAvoidrelevance to media program content
Core Design Contradiction:
Quantity of substanceVSLoss of information

Solution Approach 1:

The system continuously monitors the actual state of the media program (topics, sentiments, emotions) and uses this feedback to dynamically adjust advertisement selection. This closed-loop approach ensures ads remain relevant to current program content while maintaining broad reach through real-time optimization.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The advertisement selection system transitions from static pre-scheduling to dynamic real-time selection. The system adapts advertisement choices based on changing program states, viewer responses, and contextual factors during live broadcasts, allowing simultaneous optimization for both reach and relevance.

Inventive Principle:
Principle #15Dynamics

2Ease of operation

If advertisements are aired at desired times regardless of actual program state, then scheduling simplicity is maintained, but engagement with advertisement content is suboptimal

Engineering Contradiction:
Improvescheduling simplicityVSAvoidengagement with advertisement content
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The system automatically monitors program state, evaluates advertisement relevance, and makes selection decisions without manual intervention. This self-service approach maintains scheduling simplicity while dramatically improving engagement through context-aware ad placement.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system pre-loads and prepares multiple candidate advertisements with associated metadata and relevance criteria before program segments begin. This preliminary preparation enables rapid real-time selection without compromising scheduling efficiency or engagement optimization.

Inventive Principle:
Principle #10Preliminary action

3Device complexity

If traditional advertisement selection methods are used, then system complexity is low, but ability to match advertisements to creator and content is poor

Engineering Contradiction:
Improvesystem complexityVSAvoidability to match advertisements to creator and content
Core Design Contradiction:
Device complexityVSAdaptability or versatility

Solution Approach 1:

The system introduces an intelligent intermediary layer between program content and advertisement selection. This intermediary analyzes program state, creator attributes, and advertisement characteristics to make matched connections, significantly improving adaptability without requiring complete system redesign.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system employs a multi-functional platform that simultaneously performs program analysis, creator profiling, advertisement evaluation, and real-time selection. This universal approach handles diverse content types and advertisement formats through unified processes, scaling adaptability without proportional complexity increases.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentUS11463772B1Selecting advertisements for media programs by matching brands to creators
Publication Date: 2022.10.04 AMAZON TECH INC
  • US11463772B1 patent drawing
  • US11463772B1 patent drawing
  • US11463772B1 patent drawing

AI summary

Advertisements for brands that are to be aired during media programs are selected based on similarities between the brands and creators of the media programs. As a media program is being aired, data representing content of a media program is processed to identify words being spoken or sung during the media program and sentiments associated with the media program. The media program is classified based on the words and sentiments, and a classification of the media program is compared to attributes of advertisements to determine which of the advertisements is best suited for airing during the media program. Additionally, a set of words that, if spoken or sung during a media program would establish conditions favorable to a given advertisement may be identified and provided to a creator.