Sarcasm Target Detection

Steps to run implementation of base paper

1 - Prepare the dataset

a. Use "snippets.xlsx" from Dataset folder and upload it to colab notebook.
b. Use this shareable google drive link 'https://drive.google.com/drive/folders/17bSbi_sxoyWN1mGmwzwV4lje9Q3uBmbs?usp=sharing' and choose 'Add shortcut to drive'.
c. Use this shareable google drive link 'https://drive.google.com/file/d/1SO4Y13N9ges5A4UM4gezgJtQMy5IjPZe/view' and choose 'Add shortcut to drive', to include 'crawl-300d-2M.vec'.

2 - Run the code

a. Open google colab and upload the 'SMDM_CodeBase.ipynb' from the code folder.
b. Go to runtime and click 'Run all', which will run all cells in sequential order.
c. The first cell will ask for the dataset, Upload the 'snippets.xlsx' file.
d. Mount your google drive.

The code appproximately takes 40 minutes to run, once the code execution is finished, iterate to the last cell, to view the prediction accuracy over the training.

Steps to run improvement on base paper

1 - Prepare the dataset

a. Use "snippetsHindi.xlsx" from Dataset folder and upload it to colab notebook.
b. Use this shareable google drive link 'https://drive.google.com/file/d/1vtEb7E9mnSD1udqmUSHrgzqcoBMRQ0sA/view?usp=sharing' and choose 'Add shortcut to drive', to include 'gloveHindi.txt'.

2 - Run the code

a. Open google colab and upload the 'SMDM_CodeBase_IMPROVEMENT.ipynb' from the code folder.
b. Go to runtime and click 'Run all', which will run all cells in sequential order.
c. The first cell will ask for the dataset, Upload the 'snippetsHindi.xlsx' file.
d. Mount your google drive.

The code appproximately takes 15 minutes to run, once the code execution is finished, iterate to the last cell, to view the prediction accuracy over the training.