|
Activity Number:
|
311
|
|
Type:
|
Invited
|
|
Date/Time:
|
Tuesday, August 8, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #305366 |
|
Title:
|
Automated Metadata
|
|
Author(s):
|
Edward Wegman and Faleh Alshameri*+
|
|
Companies:
|
George Mason University
|
|
Address:
|
4201 Massachusetts Ave, NW, Washington, DC, 20016,
|
|
Keywords:
|
|
|
Abstract:
|
Homeland security implies searching massive databases for information involving possible terrorists and the threats they are likely to bring. Many of these databases include free-form text such as intercepted emails and transcripts of phone calls. The implication is that these massive databases are sufficiently large that they cannot be examined thoroughly by humans. Generally, metadata involve information about the format of the data, but not necessarily the actual content. The concept of automated metadata is to use data mining tools to extract features from the data and attach the features to the data as digital objects in the form of metadata. Thus, an investigator could search for specific datasets having some desired features. We have employed this notion with a dataset involving 16,000 articles gathered from CNN and Reuters. This work is joint with Faleh Alsham.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2006 program |
