Conversational Keyword Extraction for Context-Aware Ad Targeting

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional systems fail to provide context for user interests beyond keywords searched or accessed, leading to irrelevant advertisements and degraded user experience.

Innovation Solution

Capture and analyze voice data to extract keywords using a 'sniffer' algorithm, identifying trigger words to associate positive or negative interests with users, and utilize these keywords for targeted advertising and recommendations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If conventional systems track keywords entered by users or content accessed to determine user interests, then advertising targeting can be implemented, but the information is limited to topics users specifically search for with no context provided, leading to irrelevant advertisements

Engineering Contradiction:

Improvecontext informationVSAvoidsystem complexity

Core Design Contradiction:

Loss of informationVSDevice complexity

Solution Approach 1:

The patent introduces voice data as an intermediary source that bridges the gap between limited keyword searches and comprehensive user interest profiling. Voice data captures natural conversations, preferences, and context without requiring active user searching, thereby enriching the information available for advertising targeting while maintaining system feasibility through existing voice processing technologies

Inventive Principle:

Principle #24Intermediary (Mediator)

2Adaptability or versatility

If users search for gifts for others or browse content against their preferences, then keywords are associated with users, but this results in irrelevant or upsetting advertisements that degrade user experience

Engineering Contradiction:

Improveadvertising relevanceVSAvoiduser experience

Core Design Contradiction:

Adaptability or versatilityVSEase of operation

Solution Approach 1:

The system analyzes voice data to detect user preferences, dislikes, and contextual information, then uses this feedback to refine advertising content. By continuously processing voice inputs and adjusting ad delivery based on detected preferences and negative reactions, the system adapts to user needs while preventing the delivery of irrelevant or upsetting advertisements

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

Instead of inferring user interests solely from active searches or clicked content, the patent inverts the approach by analyzing passive voice data - conversations, ambient speech, and natural language expressions. This inversion allows the system to understand user preferences before they actively search, preventing irrelevant ad delivery while improving advertising relevance

Inventive Principle:

Principle #13The other way round (Inversion)

Data Source

PatentUS12525246B2Keyword determinations from conversational data

Publication Date: 2026.01.13 AMAZON TECH INC

AI summary

Topics of potential interest to a user, useful for purposes such as targeted advertising and product recommendations, can be extracted from voice content produced by a user. A computing device can capture voice content, such as when a user speaks into or near the device. One or more sniffer algorithms or processes can attempt to identify trigger words in the voice content, which can indicate a level of interest of the user. For each identified potential trigger word, the device can capture adjacent audio that can be analyzed, on the device or remotely, to attempt to determine one or more keywords associated with that trigger word. The identified keywords can be stored and/or transmitted to an appropriate location accessible to entities such as advertisers or content providers who can use the keywords to attempt to select or customize content that is likely relevant to the user.