Overview
To provide relevant services to specific users, the content to be delivered
must be selected according to the users' profiles. Since mobile users'
profiles, including their physical and personal situations, are subject
to change, the content for mobile users must be selected as close as possible
to the time when it is actually needed. For that purpose, we use a matching
engine that considers available content and the user profiles, and matches
them dynamically.
The matching engine must be able to handle a very large amount of content,
as well as many user profiles. Both the content and the users are constantly
changing. In addition, the matching process must be able to select appropriate
content efficiently. We have built a basic prototype and studying how to
improve its scalability and performance.
Research Topics:
- Languages for specifying content and user profiles.
- Methods for managing changing content and changing user profiles.
- Techniques for optimizing matching algorithms.
|
