Hierarchical Machine Learning for Media Recommendation Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current media content recommendation systems fail to effectively combine user engagement and satisfaction objectives, leading to suboptimal media content selection for users, as these objectives are not adequately integrated in existing hierarchical structures.
Innovation Solution
A hierarchical machine learning algorithm is employed, where multiple machine learning algorithms are arranged hierarchically to predict lower-level user engagement objectives and higher-level user satisfaction objectives, allowing for back-propagation training to adjust scores across the hierarchy, thereby improving media content recommendation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple machine learning algorithms are arranged hierarchically to predict both user engagement and satisfaction objectives, then the accuracy of media content recommendations is improved, but the device complexity increases
Solution Approach 1:
The patent divides the recommendation system into multiple hierarchical levels, where each level predicts a specific objective (e.g., completion probability at lower levels, satisfaction probability at higher levels). This segmentation allows each algorithm to focus on a specific task while maintaining overall system accuracy without requiring all algorithms to handle every objective simultaneously.
Solution Approach 2:
The patent implements a nested hierarchical structure where algorithms at higher levels take outputs from algorithms at lower levels as inputs. This nesting allows the system to build upon previous predictions, with each level adding a layer of refinement for different objectives, thereby improving overall recommendation accuracy through integrated multi-objective prediction.
2Reliability
If back-propagation training is used to adjust scores across the hierarchical algorithm levels, then the integration of user engagement and satisfaction objectives is improved, but the training time and computational resources increase
Solution Approach 1:
The patent employs back-propagation as a feedback mechanism that propagates error signals from higher-level objectives down to lower-level algorithms. This feedback loop allows the system to continuously refine predictions by adjusting weights based on actual user behavior data, improving the reliability of objective integration while managing training time through efficient weight updates.
Solution Approach 2:
The patent performs preliminary training of lower-level algorithms before higher-level algorithms, establishing a sequential training order. This preliminary action allows the system to build a solid foundation of engagement predictions before adding satisfaction objectives, reducing the overall training time required compared to simultaneous training of all objectives.
Data Source
AI summary
An electronic device generates a score for each objective in a hierarchy of objectives. Generating the score comprises, using a first machine learning algorithm, generating a score for a first objective corresponding to a first level in the hierarchy of the objectives and using an output of the first machine learning algorithm, distinct from the score for the first objective, as an input to a second machine learning algorithm to generate a score for a second objective corresponding to a second level in the hierarchy of objectives. The electronic device generates a combined score using the score for the first objective and the score for the second objective. The electronic device selects, automatically without user input, media content based on the combined scores for the plurality of media content items and streams, using an application of the media-providing service, one or more of the selected media content to a user.