IMI

Code for searching and annotating the NTU Multilingual Corpus. The code is under the MIT license. The databases may have their own licenses.

We will start with the code for searching.

If you use this, please cite:

Bond, Francis, Luís Morgado da Costa, and Tuấn Anh Lê (2015) IMI — A Multilingual Semantic Annotation Environment. In Proceedings of ACL-IJCNLP 2015 System Demonstrations, Beijing. pp 7–12

@inproceedings{bond-etal-2015-imi,
    title = "{IMI} {---} A Multilingual Semantic Annotation Environment",
    author = "Bond, Francis  and
      Morgado da Costa, Lu{\'\i}s  and
      L{\^e}, Tuấn Anh",
    booktitle = "Proceedings of {ACL}-{IJCNLP} 2015 System Demonstrations",
    month = jul,
    year = "2015",
    address = "Beijing, China",
    publisher = "Association for Computational Linguistics and The Asian Federation of Natural Language Processing",
    url = "https://www.aclweb.org/anthology/P15-4002",
    doi = "10.3115/v1/P15-4002",
    pages = "7--12",
}

There is code for exporting the corpus to XML here: https://github.com/lmorgadodacosta/NTUMC

Installation

Install passlib through apt-get (not sure why) sudo apt-get install python3-passlib python3-jinja2

Copy the files to /var/www/ntumc/

$ rsync -av IMI/www/* /var/www/ntumc/

Make $user:www-data the owner of /var/www/ntumc e.g. sudo chown -R bond:www-data /var/www/ntumc

$ sudo chmod a+rx /var/www/ntumc/cgi-bin/
$ sudo chmod a+rx /var/www/ntumc/cgi-bin/*.cgi
$ sudo chmod a+rx /var/www/ntumc/cgi-bin/*.py
$ sudo chmod g+w /var/www/ntumc/db/*.db
$ sudo chmod a+r /var/www/ntumc/db/*.db
$ sudo chmod g+w /var/www/ntumc/db
$ sudo chgrp www-data /var/www/ntumc/db
$ sudo chgrp www-data /var/www/ntumc/db/*.db

If you have not already enabled cgi:

$ sudo a2enmod cgid

Set up the logs:

cd /var/www/ntumc/cgi-bin/
touch  addss_error.log cgi_err.log
chmod a+w *.log

Put something like this in /etc/apache2/conf-enabled/httpd.conf

Add an admin db, wordnet db and corpora as needed (this code does not yet include all the bits)

### NTUMC
AddDefaultCharset UTF-8
SetEnv PYTHONIOENCODING utf8

# Uncomment to allow Python cgitb tracebacks in user-facing web browser.
# This is a workaround for later versions of Apache 2.2+ (around 2017)
# Not recommended in production settings. Ref: https://bugs.python.org/issue8704
#HttpProtocolOptions Unsafe


ScriptAlias /ntumc/cgi-bin/ /var/www/ntumc/cgi-bin/
Alias /ntumc/ /var/www/ntumc/html/

<Directory "/ntumc/cgi-bin/">
AddHandler cgi-script .cgi 
AllowOverride All
Options -Indexes +FollowSymLinks +ExecCGI
Order allow,deny
Allow from all
</Directory>

Note that the databases (in ntumc/db) and log file (ntumc/log/ntumc.txt) must be writable by the webserver (www-data by defualt in ubuntu).