Media Recommendation Model Error Correction via Multi-Source Data Normalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional media asset recommendation systems rely on a single data set from a single content provider, limiting their accuracy and failing to fully capture user preferences across diverse media consumption patterns.
Innovation Solution
The system determines error values for training models by comparing user preference information from multiple data spaces, normalizing and aggregating data from different metrics to calculate improved media asset similarity and user enjoyment levels, using trainable parameters to adjust and refine recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional systems use a single data set from a single content provider to determine media asset recommendations, then the system complexity is low and ease of operation is maintained, but the recommendation accuracy is limited and cannot fully capture user preferences
Solution Approach 1:
The patent combines multiple data spaces from different content providers into a unified recommendation system. The system aggregates user preference information, media asset metadata, and interaction data from multiple sources, normalizing and integrating them to improve recommendation accuracy while managing the increased complexity through systematic data fusion approaches
Solution Approach 2:
The recommendation system is designed to universally process and integrate data from multiple content providers and data spaces. It implements multi-functional capabilities to handle diverse data formats, metrics, and user preference representations, enabling the system to adapt to various data sources while maintaining a unified recommendation framework
2Loss of information
If the system integrates user preference information from multiple data spaces with different metrics, then the comprehensiveness of user preference understanding improves, but the difficulty of detecting and measuring preferences increases
Solution Approach 1:
The system applies parameter changes by normalizing data from different metrics and scales into a unified framework. It transforms user preference information, interaction data, and media asset attributes from various data spaces into consistent parameter representations, enabling effective aggregation and comparison while preserving the essential characteristics of the original data
Solution Approach 2:
The patent introduces intermediary processing layers that mediate between diverse data sources and the recommendation engine. These intermediaries normalize and aggregate data from multiple data spaces, transforming heterogeneous information into a unified format that can be effectively processed while maintaining the integrity and meaning of the original user preference data
Data Source
AI summary
Methods and systems for determining an error value based on comparing an expected media asset similarity value corresponding to a first media asset and a second media asset, as determined using a model, to a media asset similarity value determined from user preference information associated with multiple data spaces. User preference information is received from two data spaces that are managed by different content providers. User preference information from the two data spaces is normalized and an indication of similarity between two media assets is determined. The indication of similarity is compared to an expected similarity value received from a model and an error value is determined based on the comparison of the expected similarity value and the similarity value.