Dynamic Live Streaming Recommendation System
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional recommendation systems for live streaming fail to customize content recommendations based on user preferences, leading to biased exposure towards popular distributors and potential user boredom, as they do not dynamically update content offerings.
Innovation Solution
A method and system that identify live streaming programs based on user attributes and tags, generating subsets using different criteria to ensure personalized and varied recommendations, with periodic reordering and updating to prevent repetition and maintain user engagement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional recommendation systems prioritize popular distributors, then visibility for established content providers is improved, but user boredom increases and equal exposure for all distributors deteriorates
Solution Approach 1:
The recommendation system dynamically adjusts the criteria for selecting live streaming programs based on time-varying rules. It periodically changes the weighting between popularity-based criteria and diversity-based criteria, allowing popular distributors to maintain visibility while periodically promoting lesser-known distributors to ensure equal exposure opportunities.
Solution Approach 2:
The system implements periodic reordering of recommended content based on time-varying criterion rules. At certain intervals, it shifts from popularity-based ranking to diversity-based ranking, ensuring that different distributors receive exposure at different times, thus preventing user boredom while maintaining reliability for established distributors.
2Device complexity
If recommendation systems provide static content lists, then system complexity is reduced, but user engagement deteriorates due to repetition
Solution Approach 1:
The recommendation system transitions from static to dynamic content selection by implementing time-varying criterion rules that automatically adjust the composition of recommended live streaming programs based on the current time period, user behavior patterns, and distributor performance metrics, thereby maintaining user engagement without excessive complexity.
Solution Approach 2:
The system employs automated algorithms that self-adjust the recommendation criteria based on pre-defined time-varying rules and performance feedback, eliminating the need for manual intervention while maintaining high user engagement through continuous content variation.
3Manufacturing precision
If recommendation systems use single criterion filtering, then manufacturing precision of recommendations is improved, but adaptability to diverse user preferences deteriorates
Solution Approach 1:
The recommendation system segments the single criterion into multiple time-varying criteria that are applied at different intervals. It divides the recommendation process into phases, each with its own precision-optimized criterion, while the overall system adapts to diverse user preferences through the sequence of segmented criteria.
Solution Approach 2:
The system changes the parameters of the recommendation criteria over time, adjusting the weight and type of criteria based on user interaction patterns and time-of-day factors. This allows the system to maintain high precision for specific user segments at specific times while adapting to diverse preferences across different periods.
Data Source
AI summary
The present disclosure relates to a system, a method and a computer-readable tangible non-transitory medium for recommending live streaming data. The method includes identifying a set of live streaming programs according to an attribute of the user and a tag of each live streaming program in the set of live streaming programs, generating a first subset of live streaming programs from the set of live streaming programs according to a first criterion, and generating a second subset of live streaming programs from the set of live streaming programs according to a second criterion. The first criterion is different from the second criterion. The present disclosure can recommend contents in a more efficient way and increase user engagements.