SemEval 2020 - Task 10: Emphasis Selection For Written Text in Visual Media

Team 'Corona Survivors':

• Rishabh Agarwal
• Sahil Dhull
• Vipul Singhal

Under the guidance of:

Prof. Ashutosh Modi

Installation

git clone https://github.com/SahilDhull/emphasis_selection.git
cd emphasis_selection
virtualenv project
pip install -r requirements.txt

Note: If not having virtualenv, install it using: pip install virtualenv. Download pre-trained GloVe embeddings from http://nlp.stanford.edu/data/glove.6B.zip

Problem Statement

The problem statement (Task 10 in SemEval-2020) is designing automatic methods for emphasis selection i.e. choosing candidates for emphasis in short written text.

More formally, Given a sequence of words or tokens C = , we want to compute a score for each which indicates the degree of emphasis to be laid on the word.

Directory Structure

• datasets/ -> contains the train and development sets

• BiLSTM_Attention_approach/ -> contains the model for 'BiLSTM + Attention' approach

• Transformers_approach/ -> contains model and ensemble files for Transformers approach

• requirements.txt -> contains versions of packages required

• eval_metric.py -> contains the function for Match_m as defined in task