Abstract
Limited number of input buttons on a mobile device, such as mobile phones and PDAs, restricts people's access to digital broadcast services. In this paper, we present a reinforcement learning approach to automatically navigating among services in mobile digital television systems. Our approach uses standard Q-learning algorithm as a theory basis to predict next button for the user by learning usage patterns from interaction experiences. We did the experiment using a modified algorithm in test system. The experimental results demonstrate that the performance is good and the method is feasible and appropriate in practice.
Original language | English |
---|---|
Title of host publication | ICEIS 2004 - Proceedings of the Sixth International Conference on Enterprise Information Systems |
Editors | I. Seruca, J. Filipe, S. Hammoudi, J. Cordeiro |
Publisher | SciTePress |
Pages | 140-145 |
Number of pages | 6 |
ISBN (Print) | 9728865007 |
DOIs | |
Publication status | Published - 23 Nov 2004 |
MoE publication type | A4 Article in a conference publication |
Event | 6th International Conference on Enterprise Information Systems, ICEIS 2004 - Porto, Portugal Duration: 14 Apr 2004 → 17 Apr 2004 |
Conference
Conference | 6th International Conference on Enterprise Information Systems, ICEIS 2004 |
---|---|
Country/Territory | Portugal |
City | Porto |
Period | 14/04/04 → 17/04/04 |
Keywords
- Automatic navigation
- Button prediction
- Exploration
- Intelligent user interface
- Reinforcement learning