Installing INCEpTION - an Open Source tool for Linguistic Annotation

In a previous post about Tarkeeb I mentioned the use of annotation software for rebuilding the Quran Corpus or building any Arabic corpus. I discussed software like Flat but then mentioned my preferred option for annotation, INCEpTION. Below I will explain how to install INCEpTION on Ubuntu 20.04(you should be able to adjust the commands to install it on most distros).

Installing Java

If you use a system container like LXC, the standard containers don't come with Java installed. This is a good thing, as it allows you to choose OpenJDK over Oracle's Java.

To install OpenJDK, do the following:

sudo apt update
sudo apt upgrade
sudo apt install openjdk-11-jdk

Verifying that Java is installed:

java --version

If it didn't work, visit here for debugging.

Downloading & Running INCEpTION

Create a directory for the standalone Java executable:

mkdir annotation/

Download INCEpTION from here. When I installed the software, it worked with version 0.18.0. However, version 0.19.3 should also work(I will update this article if it doesn't).

Save the inception-app-webapp-0.19.3-standalone.jar into the annotation/ folder.

On the Downloads page above, the instructions for running the web application are:

java -jar inception-app-webapp-0.19.3-standalone.jar

These instructions assume that you are running the executable locally and can double-click on it. This didn't work for me because I isolated the app into a container.

After a bit of yak-shaving, I learned that INCEpTION is built using the Spring Framework. To run a Spring web application as if it was a remote application and making it accessible outside a container, the following command needs to be used:

java -Djava.awt.headless=true -Dserver.port=8999 -Dserver.ip=0.0.0.0 -jar inception-app-webapp-0.19.3-standalone.jar

Binding to 0.0.0.0 and a specific port is self-explanatory. I cannot recall why, but running the application as headless was required to make it work. If you can run it without the headless command, please email me and explain how/why.

Using INCEpTION

The application should now be accessible on the container IP on the specific port selected. eg. 10.0.4.44:8999

Thereafter you can look at the User Guide for more information on how to use it.

I will be adding a future blog post about rebuilding the Quran Corpus with INCEpTION. At the moment I am manually adding some content to the corpus data. This data was added to the existing Quran Corpus and specifically relates to the Reference Nodes, Hidden Nodes and Empty Nodes.