SMILE Twitter Emotion dataset

Dataset Overview

Dataset Name: SMILE Twitter Emotion dataset

Creators: Wang, Bo; Tsakalidis, Adam; Liakata, Maria; Zubiaga, Arkaitz; Procter, Rob; Jensen, Eric

Year of Creation: 2016

Content Description: The dataset contains tweets annotated with multiple emotions such as happiness, anger, sadness, providing rich resources for sentiment analysis tasks.

Dataset Download Link: SMILE Twitter Emotion dataset page

Data Processing

Preprocessing Tool: python preprocess.py

Preprocessing Output: Generates dataset_train.pt, dataset_val.pt, and data_info.json files for training and validation of BERT models.

Model Training

Model Used: bert-base-uncased model from the transformers library

Training Script: python train.py

Model Evaluation

Evaluation Script: python evaluate.py

Model Application

Example Code: python tweet = "I hate this movie" label = predict_label(tweet) print(f"Predicted label: {label}")

Citation Information

Citation Format:

@misc{wang2016smile, author = {Wang, Bo and Tsakalidis, Adam and Liakata, Maria and Zubiaga, Arkaitz and Procter, Rob and Jensen, Eric}, title = {SMILE Twitter Emotion dataset}, year = {2016}, publisher = {figshare}, doi = {10.6084/m9.figshare.3187909.v2} }