Big-PubSub aims at bringing together the Data Management and the Publish/Subscribe communities. The data management community has traditionally studied continuous queries and Active Databases with the assumption that the number of such queries is relatively small, and recently had been focusing on the systems and algorithmic implications of addressing such problems when the streaming data become very large. On the other hand, the Pub/Sub community has studied how to reach the end users through a network of brokers. Recently, systems like Storm allow the efficient processing of high-throughput data streams. Despite the progress on both domains, there are several open research directions in the confluence of these domains. These research directions, which are the focus of this workshop, are motivated by the recent development and adoption of novel applications that connect users and provide services by analysing their data, in spaces such as social networks, smart cities, or healthcare. Research, work-in-progress (WiP), descriptions industrial applications or recent research initiatives, or position papers are particularly encouraged, on any topic related to Big Data and Publish/Subscribe Systems, including the following:
- How can a database system efficiently support millions of subscribers and publishers?
- How can both streaming and static data participate in subscription queries?
- Data management in social networks, crowdsourcing, or smart cities.
- Data management in the Internet of Things (IoT).
- Identifying and supporting event discovery on NoSQL and SQL databases, and complex event processing systems.
- Balance of work between back-end databases and brokers.
Dimitrios Gunopulos, University of Athens
Vagelis Hristidis, UC Riverside
Sharad Mehrotra, UC Irvine