University of Ottawa - Carleton University
Ottawa-Carleton Institute for Computer Science (OCICS) Presentation
|
September 28, 2012 @ 10:00a.m.
Automatic Generation of a Warehouse Schema from a Conceptual Level Specification
|
Speaker: Jeyakanthan Mahesa
Location: 3101 CB (Canal Building) |
ABSTRACT
At present, organizations have access to various sources of data, both structured and unstructured, across various platforms such as databases, documents, and the web. These are not just limited to text, but also extend to images, videos and voice. If it is not yet apparent from the above, we are drowning in a deluge of data. A data warehouse provides an integrated repository of various heterogeneous data sources; where the ultimate goal is to help us connect the dots, turning data into valuable information.
Data warehouses have been around for some time now; we have all heard of the famous Inmon vs. Kimball approaches for years. Lately however, there has been concern about the impact of growing data on current design methodologies. The fundamental problem with current methodologies is that they are inherently bottom-up by nature. The typical process usually involves identifying the data sources, applying an ETL process to obtain the warehouse, and finally layering it with various business intelligence and data mining applications. According to some, this approach has little future since information derived from a data-driven model is merely information about the data set. The other approaches have been sidetracked by the influx of data, inevitably leading to bottom-up designs.
The purpose of our research is to introduce a paradigm shift by developing a top down methodology. This is a collaborative effort sponsored under the NSERC Business Intelligence Network.
|
| Return to Schedule |
|
