FAQs

1. What are the submission guidlines?

These will be published on the web site very soon.

2. What is XML::Simple and how do I use it?

Our Perl scripts for expanding keywords and interview data use Grant McLean's XML::Simple Perl module. At the time of this writing, recent releases of ActiveState Perl already include XML::Simple. Other releases might not include XML::Simple. It is a small module and generally easy to install if you need it. On one of our machines we simply downloaded the Simple.pm file and copied it into the XML directory in our Perl lib path. You should read the installation instructions if you need to install this module.

More information on XML::Simple (including detailed installation instructions) is available at:

http://homepages.paradise.net.nz/gmclean1/cpan/
http://search.cpan.org/~grantm/XML-Simple-2.13/lib/XML/Simple/FAQ.pod
http://search.cpan.org/~grantm/XML-Simple-2.13/

3. When are the French topics coming out?

French topics are complete and are now being checked.

4. What is the correct order for the auto-keywords description in the README?

AUTOKEYWORD2004A1 is from the University of Maryland
AUTOKEYWORD2004A2 is from IBM (please ignore "the top 16 concept keywords and the top 4 geographical keywords are included")

5. Why is topic number 2181 in Czech and Spanish evaluation sets but not in the English and French sets? What should we do with it?

Short Answer: Topic 2181 should be removed wherever it appears.

Long Answer: There are errors in three files:

Evaluation Topics/TREC/SpanishTopics.trec (ERROR: topic 2181 included, ACTION: remove topic 2181)
Evaluation Topics/TREC/CzechTopics.trec (ERROR: topic 2181 included, ACTION: remove topic 2181)
Evaluation Topics/XML/CzechTopics.xml (ERROR: 10 extra topics included, ACTION: all should be removed, they are: 1845, 1858, 1908, 1933, 2061, 2063, 2137, 2181, 2411, 9003)

6. Is the "Workspace" on the website active, if so, what are the username and password?

This part of the website is active but contains nothing at present, any ideas, let us know. We'll distribute the username/password to participants if we find a use for the "Workspace".