Hybrid Training Data for Accurate Brand Frequency Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing advertising systems struggle to optimize and personalize ad delivery across a diverse range of Internet of Things (IoT) devices, leading to suboptimal user experiences due to inadequate methods for brand detection and frequency management.
Innovation Solution
A system for programmatic generation of training data using a combination of human curation and AI model training, which includes receiving human input, generating hybrid training data, calculating brand-probability pairs, and performing frequency management to improve brand detection accuracy and regulate ad serving.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If programmatic training data generation is used, then productivity is improved, but measurement precision deteriorates
Solution Approach 1:
The system merges programmatic training data generation with human curation by combining automated brand detection algorithms with human reviewer feedback. The training module generates initial training data programmatically, then human reviewers correct and refine this data, creating a hybrid approach that maintains high productivity while improving measurement precision through human oversight.
Solution Approach 2:
The system implements feedback mechanisms where human reviewers provide corrections and refinements to programmatic brand detection results. This feedback loop allows the system to learn from human expertise and progressively improve its detection accuracy, resolving the contradiction between automated efficiency and precision.
2Reliability
If hybrid training data combining is used, then reliability is improved, but device complexity increases
Solution Approach 1:
The training system is segmented into distinct functional modules: a programmatic training data generation module that handles automated detection, a human curation module that processes human feedback, and a hybrid data combination module that integrates both sources. This segmentation allows each component to be optimized independently while working together to improve reliability without overwhelming complexity.
Solution Approach 2:
The system introduces an intermediary processing layer that combines programmatic and human-generated training data. This intermediary module acts as a mediator between automated detection and human curation, synthesizing both data sources into a unified training dataset that improves reliability while managing complexity through structured integration.
3Measurement precision
If quality score filtering is applied, then measurement precision is improved, but productivity decreases
Solution Approach 1:
The system applies quality score filtering selectively rather than universally. Instead of filtering all training data items, the system identifies and filters only those items that fail to meet minimum quality thresholds, allowing high-quality data to pass through efficiently while removing only the problematic portion. This partial action approach improves measurement precision without significantly impacting overall productivity.
Data Source
AI summary
Systems and methods for programmatic generation of training data, including: a training module configured to receive human curation input for brand entity detection, generate hybrid training data by combining programmatic and human generated data, and calculate brand-probability pairs by weighting detection results and human input to improve brand detection accuracy for frequency management; an online media service configured to serve training data to recipients during controlled experiments, calculate quality scores based on performance metrics and human input, and exclude low-quality training data from model training; a model training engine configured to train an artificial intelligence model for brand detection using the hybrid training data weighted by quality scores; and a frequency management service configured to execute the trained model on media items to identify brand identifiers with improved accuracy and regulate serving frequency of brand-associated content to recipients.