We go on to implement 'Emoji Powered Capsule Network to Detect Type and Target of Offensive Posts in Social Media' with aid from the official GitHub repository. Since there are three different subtasks like
- Offensive language identification
- Offensive language categorization
- Offensive language target identification,
for the sake of simplicity and to accommodate within 8 pages, we go on with the implementation of Capsule Network and CNN over the subtask 1. The same can be extended for other tasks with the aid of the OLID dataset that is attached to this repository. I have attached the Colab notebooks for your reference. I have remodelled the capsule network to suit my needs and have written preprocessing and evaluation functions to suit my personal needs / expectations of obtaining results in this project.