What Can Hierarchies Do for Data Streams?

Much effort has been put into building data streams management systems for querying data streams. Here, data streams have been viewed as a flow of low-level data items, e.g., sensor readings or IP packet data. Stream query languages have mostly been SQL-based, with the STREAM and TelegraphCQ languages as examples. However, there has been little work on supporting OLAP-like queries that provide multi-dimensional and summarized views of stream data. In this paper, we introduce a multidimensional stream query language and its formal semantics. Our approach enables powerful OLAP queries against data streams with dimension hierarchies, thus turning low-level data streams into informative high-level aggregates. A comparison with STREAM shows that our approach is more flexible and powerful for high-level OLAP queries, as well as far more compact and concise.