|
CALL FOR PAPERS
Information Systems: Special Issue on Data Warehousing
Editors
|
|
Patrick O'Neil |
|
Kenneth A. Ross |
|
University of Massachusetts/Boston |
|
Columbia University |
|
poneil@cs.umb.edu |
|
kar@cs.columbia.edu |
Introduction.
A "Data Warehouse" is a database system that provides query capability
to analysts engaged in extracting strategic information about a
(usually commercial) enterprise. Data Warehouses can use a number of
different data models and user interfaces, including SQL queries on
Star Schemas, OLAP (ROLAP and MOLAP), and Data Mining, where detail
queries are performed automatically for high-level goals set by the
analyst. In most cases, the data about the enterprise held in the
Warehouse has been extracted from the enterprise's operational (OLTP,
Point-Of-Sale) system, and processed in various ways (brought to a new
data model, cleansed, indexed, preaggregated) to make queries more
efficient and responsive to analyst's needs. Data will often be
brought in from other sources as well.
A Data Warehousing system at an enterprise can have multiple
goals. One goal can be to answer the rather fundamental concerns of
senior management: "We sell various products in various markets and
measure our progress over time; how are we doing?" More complex
questions arise from this that can provide strategic advantage: "To
what can we attribute variations in sales by product, market, and time
period (as well as by placement on shelves, packaging, offered
promotions, etc.)? How can we build on this understanding to increase
our total sales?" Other goals include providing suppliers with good
information for reordering, detecting bottlenecks in complex supply
and demand chains, and understanding customer requirements.
Details of the Call.
Because of the widespread research activity in this topic, Information
Systems is planning a special issue on Data Warehousing Systems for
Winter 2000/1. Our goal is to collect papers with important new
insights and experiences, and put together a strong issue. Theoretical
papers should include a solid motivation for why the stated results
are applicable in real applications. Applied papers that give
insightful descriptions of working systems in actual use are
particularly solicited. Topics of interest include, but are not
limited to, the following.
| Novel Indexing for Performance |
Data Loading |
| Novel query processing algorithms |
Updating (Especially with following two topics) |
| Materialized views |
24X7 Availability |
| MOLAP, ROLAP, other models |
Stability (Repeatability of Queries) |
| Data Mining |
Schema Design |
| Data Cleansing |
|
Papers should be at most 30 pages long, double-spaced, in font size
10 or larger with one-inch margins on all sides. Five copies of each
paper should be sent to: Patrick O'Neil; Dept., of Math. and C.S.;
UMass/Boston; Boston, MA 02125-3393; USA. Alternatively, a paper in
PDF format can be sent to poneil@cs.umb.edu by email.
Note the following important dates:
