AdCOFE

Code for the paper Advanced Contextual Feature Extraction in Conversations for emotion classification

How to Run :

IEMOCAP (Run.ipynb)

1. Change the path to the .pkl files under the # Importing the Dataset heading.
2. Run the cell under the heading # Required Installs for the project
3. Run all the cells. You would get the output.

MELD (MELD_Run.ipynb)

1. Clone the project. Do not change the location of the train and test dataset.
2. If the location is changed, give the appropriate location in the pandas read cell.
3. Run all the cells. You would get the output.

Computational Resources :

This research was enabled in part by support provided by Compute Canada. For running our codes, we use the Compute Canada cluster, running on 14 CPU cores with 30,000M CPU memory and allocated time for the execution of the whole program as 150 minutes. The program terminates after 150 minutes or earlier with the output stored in a file. The GPU specifications used are as follows: Model: P100-PC21E-12GB, with two GPUs per CPU socket.