Demographic Inference via Time-Sensitive Feature Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
There is a lack of reliable methods for predicting user demographics from viewing habits on communal television or media devices, which limits the ability to deliver targeted advertisements and content relevant to individual users.
Innovation Solution
A machine learning-based system that uses training data including household, person identification, program title, and watch time data to create predictive models, assessing features across specific time slices and determining user attributes based on viewing history, enabling accurate prediction of demographics such as age and gender.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional methods are used to predict user demographics, then predictions can be made when household information is available, but reliable prediction cannot be made for communal television devices with multiple users
Solution Approach 1:
The patent segments the viewing data into time-sensitive feature slices (e.g., morning, afternoon, evening slots) and uses sessionization to divide viewing history into discrete user sessions. This segmentation allows the system to distinguish between different users' viewing patterns on communal devices, enabling reliable demographic prediction for each user segment rather than treating the household as a single unit.
Solution Approach 2:
The patent introduces time as an additional dimension by creating time-sensitive features and time slots. Instead of analyzing viewing data in a static manner, the system incorporates temporal dimensions (time of day, day of week, viewing sequence) to create a multi-dimensional view of user behavior that enables differentiation between multiple users on communal devices.
2Measurement precision
If machine learning models are trained with comprehensive features, then prediction accuracy improves, but computational complexity and processing time increase
Solution Approach 1:
The patent segments the feature set into time-sensitive feature slices organized by time slots. This segmentation allows the model to process complex viewing patterns in manageable temporal units, reducing computational complexity while maintaining prediction accuracy by analyzing behavior in discrete time-based segments rather than as a monolithic feature set.
Solution Approach 2:
The patent performs sessionization and time slot creation as preliminary actions before model training. By pre-processing viewing history into structured sessions with time-sensitive features, the system reduces the complexity of the training process. The model receives pre-organized data with extracted temporal patterns, eliminating the need for complex real-time feature engineering during training and inference.
Data Source
AI summary
A system and method for predicting user demographics based on a user's television, or media, viewing habits using machine learning algorithms is provided. A method of predicting a user's demographics comprises acquiring training data including one or more household data, person identification data, program title data, or watch time data. The method includes assessing a set of features. In addition, the method includes training one or more models based on the training data and set of features. The method includes acquiring viewing history data associated with at least one user. The method further includes determining one or more attributes associated with the at least one user based on inputting the viewing history data into the one or more models.