Enhancing Linear TV Channel Surfing: A Real-Time Personalized Ranking and Recommendation System with Dual Dynamic Queues
The proliferation of linear TV channels complicates content discovery, necessitating more efficient methods for viewers. This paper presents a real-time personalized ranking and recommendation system designed to enhance viewer satisfaction and interaction through dynamic channel surfing. Our approach uses a dual dynamic queue system, comprising Dynamic History Channel Queue (DHCQ) and Dynamic Future Channel Queue (DFCQ), to manage the viewer's interaction effectively. Leveraging advanced deep learning models, we generate “global” and “local” content embeddings and “user” embeddings to ensure real-time updates tailored to content and user time-sensitivities. A ‘look-ahead’ feature further enriches personalization by considering upcoming content on each channel. Preliminary user feedback indicates a strong preference for this system over traditional channel navigation methods, highlighting its potential to transform how viewers engage with linear TV. This paper underscores the system's significant contribution to improving linear TV content discovery and its implications for enhancing viewer satisfaction.
- Print ISSN
- 1545-0279
- Electronic ISSN
- 2160-2492
- Published
- 2024-07
- Content type
- Original Research
- Keywords
- linear tv, channel ranking, channel surfing, real-time personalization, dynamic queues
- DOI
- 10.5594/JMI.2024/HWAR6964