This directory contains three separate link data sets:

1. news_entities - generated manually from publically available webpages
                concerning entities in articles related to terrorism
2. hsiungspam - a link data set created from parsed spam files 
                collected from Paul Hsiung's mailbox.
3. archivespam - created from parsed spam files from
                 www.spamarchive.org

Each link data set contains three files:

*.names - contains all the names in a link data sets
*.links - contains all the links in a link data sets
*.alias - contains a subset of the ground truth alias

for example,
names can contain,
------------------
osama_bin_laden
ayman_al_zawahri
al_qaeda
bin_laden
...
------------------

links can contain,
------------------
link0, al_qaeda, osama_bin_laden, ayman_al_zawahri
link1, osama_bin_laden, bin_laden
...
------------------

alias can contain,
------------------
osama_bin_laden bin_laden the_emir
bush the_president george_bush
usa united_states_of_america
------------------

For further information, please contact

Paul Hsiung
my email is
1. concat my first and last name with no space in between.  lower case p and h
2. append _at_ gmail.com