 |
 |
|
|
Our efforts in understanding the dissemination scenarios have revealed the need for customized dissemination. Our work investigates the use of a distributed publish/subscribe based infrastructure as a suitable platform using which customization needs can be captured and explored.
The underlying pub/sub model consists of a set of pub/sub servers (brokers) connected to each other through an overlay network. Publishers and subscribers connect to one of the brokers to send or receive publications and subscriptions. The main goal of most existing pub/sub approaches is to increase scalability of the pub/sub system through reducing each broker's load; for example, techniques have been developed to reduce the total cost of matching and routing events. A common approach is to construct a spanning tree on the pub/sub overlay network to forward subscriptions and publications. This allows avoiding diffusion of content in parts of the pub/sub network where there are no subscribers and prevent multiple deliveries of events.
However, in the case of emergency warnings, notification dissemination time, the time between publishing a notification and delivering it to all of the interested subscribers, also plays a critical role. We use content-based pub/sub as a communication infrastructure for a customized alert dissemination system. While existing content-based pub/sub systems can scale reasonably well and disseminate notifications with reduced communication cost, timely notification has not been addressed. The primary goal of our work is in timely (and customized) dissemination of notifications to interested receivers.
Publications:
HiPub: High Speed Content-based Publish/Subscribe
Hojjat Jafarpour; Sharad Mehrotra; Nalini Venkatasubramanian; Raymond Klefstad;
Technical Report, 2006-06-15
Customized Dissemination in the context of emergencies
A. Ghigi;
Thesis, 2005
In this work, we explore customized delivery of information in a client/server context. Applications that need to disseminate dynamic information from a server to various clients can suffer from heavy communication costs. Customized information needs of clients can influence whether information must be pushed to the client or pulled from the server. Data caching at a client can help mitigate these costs, particularly when individual push-pull decisions are made for the different semantic regions in the data space. The server is responsible for notifying the client about updates in the push regions. The client needs to contact the server for queries that ask for data in the pull regions. We call the idea of partitioning the data space into push-pull regions to minimize communication cost data gerrymandering. In this study, we present solutions to technical challenges in adopting this simple but powerful idea. We give a provably optimal-cost dynamic programming algorithm for gerrymandering on a single query attribute. We propose a family of efficient heuristics for gerrymandering on multiple query attributes. We handle the dynamic case in which the workloads of queries and updates evolve over time. We validate our methods through extensive experiments on real and synthetic data sets.
Publications:
Achieving Communication Efficiency through Push-Pull Partitioning of Semantic Spaces to Disseminate Dynamic Information
Amitabha Bagchi; Amitabh Chaudhary; Michael Goodrich; Chen Li; Michal Shmueli-Scheuer;
Technical Report, 2006-03