Online Analytical Processing Stream: Is it Feasible?

Document Type

Conference Proceeding

Publication Date



Real-time surveillance systems and other dynamic environments often generate tremendous (potentially infinite) volume of stream data: the volume is too huge to be scanned multiple times. However, much of such data resides at rather low level of abstraction, whereas most analysts are interested in dynamic changes (such as trends and outliers) at relatively high levels of abstraction. To discover such high level characteristics, one may need to perform on-line multi-level analysis of stream data, similar to OLAP (on-line analytical processing) of relational or data warehouse data.

With limited storage space and the demand for fast response, is it realistic to promote on-line, multi-dimensional analysis and mining of stream data to alert people about dramatic changes of situations at multiple-levels of abstraction?

In this paper, we present an architecture, called stream_cube, which, based on our analysis, is feasible for successful on-line, multi-dimensional, multi-level analysis of stream data. By successful, we mean that the system will provide analytical power and flexibility, derive timely and quality responses, and consume limited memory space and other resources.


Presented at the ACM SIGMOD Workshop on Research Issues in Data Mining and Knowledge Discovery, Madison, WI, June 2, 2002.