Fuzzy Logic User Identification for Targeted Ad Delivery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current targeted advertising in broadcast networks faces challenges in accurately identifying individual users within households, leading to inefficiencies in delivering ads to the appropriate audience, as household-based targeting often fails to account for multiple users and their varying demographics.
Innovation Solution
The use of machine learning, specifically fuzzy logic, to identify and classify current network users based on their inputs, such as click streams from remote controls, allowing for real-time and granular targeting of ads without the need for persistent user profiles, enabling accurate selection and insertion of ads to actively watching viewers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If household-based targeting is used, then asset delivery can be simplified, but user identification accuracy deteriorates because multiple users within a household are not distinguished
Solution Approach 1:
The patent segments the household audience into individual users by analyzing unique input patterns (remote control usage, viewing behavior, device interactions). This allows the system to treat each user as a distinct target for asset delivery while maintaining the simplified household-based infrastructure, thereby resolving the contradiction between operational simplicity and identification accuracy.
Solution Approach 2:
The patent introduces an intermediary classification system that mediates between the household level and individual user level. This intermediary layer processes and analyzes user input patterns to infer individual user characteristics without requiring direct identification, enabling accurate user targeting while maintaining system simplicity.
2Ease of manufacture
If conventional ratings-based targeting is used, then implementation is straightforward, but targeting effectiveness deteriorates due to lack of individual user insights
Solution Approach 1:
The patent enables the system to self-classify users by automatically analyzing their own input patterns and viewing behaviors. This self-service approach eliminates the need for complex external data collection and processing, maintaining implementation straightforwardness while significantly improving targeting effectiveness through individualized user insights.
Solution Approach 2:
The patent replaces the mechanical/ratings-based targeting system with an intelligent system that uses machine learning and pattern recognition. This substitution maintains ease of implementation by leveraging existing data collection mechanisms while dramatically improving targeting effectiveness through sophisticated analysis of individual user behavior.
3Measurement precision
If individual user identification is implemented, then targeting accuracy improves, but system complexity increases due to need for continuous user profiling
Solution Approach 1:
The patent applies partial action by focusing analysis only on specific, easily collectable user inputs (remote control commands, viewing duration, channel changes) rather than attempting to profile all possible user characteristics. This selective approach achieves sufficient targeting accuracy without the complexity of comprehensive user profiling systems.
Solution Approach 2:
The patent changes the parameters being measured from traditional ratings data to real-time user interaction patterns. By transforming the type of data collected and the methods of analysis, the system achieves individual user identification with manageable complexity, using straightforward parameter changes rather than fundamentally complex new systems.
4Productivity
If real-time user classification is performed, then asset delivery relevance improves, but processing resources required increase
Solution Approach 1:
The patent performs preliminary classification actions by pre-defining user categories and classification criteria before actual asset delivery occurs. This allows the system to quickly classify users in real-time by matching their input patterns against pre-established categories, significantly reducing the processing resources needed for real-time classification while maintaining high relevance in asset delivery.
Data Source
AI summary
A targeted advertising system uses a machine learning tool to select an asset for a current user of a user equipment device, for example, to select an ad for delivery to a current user of a digital set top box in a cable network. The machine learning tool first operates in a learning mode to receive user inputs and develop evidence that can characterize multiple users of the user equipment device audience. In a working mode, the machine learning tool processes current user inputs to match a current user to one of the identified users of that user equipment device audience. Fuzzy logic may be used to improve development of the user characterizations, as well as matching of the current user to those developed characterizations. In this manner, targeting of assets can be implemented not only based on characteristics of a household but based on a current user within that household.