Audience identification
Need
Identify user profiles based on the history of TV programs viewed.
Approach
Develop a smart algorithm that estimates the probability of each segment for every view. We implemented an intelligent model to calculate the likelihood of each segment watching a specific television and designed a dynamic threshold for accurate segment assignment.
Result
A server-side solution that processes anonymous profiles, enriches them through advanced enrichment techniques, and identifies audience segments with an error margin between 0% and 3%, ensuring highly accurate viewer profiling.
Categories
Type
AI & Data Science
Need
Benefit
Business function
Industry
Telco, Media & Entertainment
, Advertising
, Utilities
, Publishers
, Finance
, Tourism & Events
, Education
, Retail Distribution
Business Model
Business Size
