Neural User Identification From TV Navigation Patterns

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing user authentication methods for shared audiovisual content devices, such as televisions, are inadequate for identifying individual users within a group, especially in the absence of active user input, and are not feasible for free-to-air television services, limiting personalized customization and usability.

Innovation Solution

A neural network-based method that analyzes user navigation patterns through audiovisual content, such as channel hopping sequences and time spent on channels, to assign unique identifiers to individual users within a group sharing a device, without requiring active user authentication or additional sensors.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional user authentication methods (smart cards, login screens) are used, then user identification is possible, but device complexity and ease of operation deteriorate due to requiring additional authentication devices or manual input

Engineering Contradiction:
Improveuser identification accuracyVSAvoidauthentication system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system performs user identification automatically without requiring user action. The neural network analyzes passive navigation data (channel changes, pause patterns, seeking behavior) to identify users, making the authentication process self-executing and eliminating the need for smart cards, remote control input, or login screens

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces physical authentication mechanisms (smart cards, biometric sensors) with a software-based neural network system that processes digital navigation patterns. This substitution eliminates hardware complexity while maintaining identification reliability through algorithmic analysis of user behavior sequences

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If active user input is required for authentication, then user identification is possible, but ease of operation deteriorates due to interrupting content consumption

Engineering Contradiction:
Improveuser identification accuracyVSAvoidcontent consumption continuity
Core Design Contradiction:
ReliabilityVSEase of operation

Solution Approach 1:

The identification system operates passively in the background, continuously analyzing navigation patterns without interrupting content playback. Users simply consume content while the neural network simultaneously processes their navigation behavior for identification purposes

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system maintains continuous content consumption while simultaneously performing user identification. The neural network processes navigation data in real-time during content playback, ensuring both uninterrupted viewing and continuous identification capability without requiring separate authentication phases

Inventive Principle:
Principle #20Continuity of useful action

3Measurement precision

If biometric sensors are added to the device, then user identification accuracy improves, but device complexity and manufacturing cost increase

Engineering Contradiction:
Improveuser identification accuracyVSAvoiddevice construction simplicity
Core Design Contradiction:
Measurement precisionVSEase of manufacture

Solution Approach 1:

The patent replaces physical biometric sensing hardware with a software-based neural network that processes digital navigation patterns. This eliminates the need for cameras, fingerprint sensors, or other biometric hardware, simplifying device construction while achieving comparable or superior identification accuracy through behavioral analysis

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

Instead of directly sensing physical biometric traits, the system creates and analyzes a digital copy of user behavior patterns through navigation data. The neural network learns and matches behavioral sequences, effectively copying identification functionality from physical biometric methods to digital pattern recognition without requiring physical sensors

Inventive Principle:
Principle #26Copying

4Reliability

If smart cards are used for authentication, then user identification is possible, but adaptability deteriorates because smart cards identify subscriptions not individual users

Engineering Contradiction:
Improvesubscription authenticationVSAvoidindividual user identification
Core Design Contradiction:
ReliabilityVSAdaptability or versatility

Solution Approach 1:

The system maintains subscription-level authentication through the existing smart card while adding individual user identification capability through neural network analysis of navigation patterns. Each user within a subscription receives personalized identification based on their unique behavioral characteristics, enabling both subscription management and individual customization simultaneously

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The neural network system serves multiple functions: it identifies individual users for personalized content recommendations, tracks viewing habits for analytics, and works with any authentication method (smart card, login, or none). This multi-functional approach replaces the single-purpose smart card system with a versatile platform supporting both subscription and individual user identification

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

Data Source

PatentEP4050505B1Methods, entities, computer programs and systems for training to assign user identifiers and to assign user identifiers in the consumption of audiovisual content
Publication Date: 2026.04.08 SKY ITALIA SRL
  • EP4050505B1 patent drawingFigure 1
  • EP4050505B1 patent drawingFigure 2
  • EP4050505B1 patent drawingFigure 3

AI summary

There is provided a method (and corresponding entity, computer program and system) for training a neural network to assign an identifier to a user comprised in a group of users who share use of a device for reproducing audiovisual content. The method comprising the steps of: for at least said user, acquiring (S10) a sequence of elements indicating information about use, each element of the sequence comprising temporal information indicating a time interval and reproduction information that refers to a respective time interval, wherein the reproduction information comprises a content indicator indicating that a unit of audiovisual content is reproduced by the device within said respective time interval; generating (S20) a data set comprising the sequence of elements relating to the user and an identifier of said user; training (S30) the neural network on the basis of said data set.