Emo (Infer Underlying Emotion)
Textual dialogue along with previous context. Goal: Infer Underlying Emotions
By Huggingface Hub [source]
About this dataset
More Datasets
For more datasets, click here.
Featured Notebooks
- 🚨 Your notebook can be here! 🚨!
How to use the dataset
The dataset consists of three csv files: train.csv, test.csv and validation.csv, each containing text and labels for psychological conversations between two people, allowing for the identification of underlying emotions of each utterance. This is perfect for gaining insights into complex relationships between emotions and textual conversations. The data includes the full dialogue between two people, along with two previous turns of context and a label for each turn which is either Happy, Sad, Angry or Others.
Here are steps on how you could use this data set:
- Read in and explore the data – First off begin by reading in your CSV file or files if using both training and test sets into python such as Pandas to quickly familiarize yourself with your data set; looking at its size (number of columns/rows), datatypes (str/int), missing values etc . This will give you better understanding when manipulating it later on down the line
- Data preprocessing – Here we want to clean up our text by removing punctuation & other unnecessary characters; normalizing word lengths; parsing out whats being said from who’s saying it using machine learning algorithms like Natural Language Understanding(NLU); & lastly converting it all into a numerical format so that our model can understand it better
- Modelling – Now that our Neural Network has been properly trained we can begin applying various machine learning approaches such as deep Artificial Neural Networks (ANNs), Convolutional Neural Networks (CNNs) , Long Short-Term Memory networks (LSTMs) & many more! Given their complexity these models take in large amounts of data which makes them suitable well suited considering that they are quite complex while taking in very large amounts of information but have been proven successful at automatically extracting patterns from raw textual dialogues similar to what is found within this dataset
- Evaluation – No matter which approach was taken there should always be a way of evaluating results whether via manual inspection or quantitative metrics such as accuracy scores etc..Inspecting numerous examples manually allows you to become familiar with errors that might occur i therefore build up an intuition around what type errors needs attention first if any
Once completed these 4 steps provide us enough information
Research Ideas
- Developing automatic dialogue systems that can recognize the feelings and emotions expressed by conversations
- Building a sentiment analysis tool to extract emotions from conversations
- Training machine learning models to predict the emotion of each utterance in a conversation
Acknowledgements
If you use this dataset in your research, please credit the original authors.
Data Source
License
License: CC0 1.0 Universal (CC0 1.0) - Public Domain Dedication
No Copyright - You can copy, modify, distribute and perform the work, even for commercial purposes, all without asking permission. See Other Information.
Columns
File: train.csv
| Column name |
Description |
| text |
The text of the conversation between two people. (String) |
| label |
The emotion associated with each turn of dialogue (Happy, Sad, Angry or Others). (String) |
File: test.csv
| Column name |
Description |
| text |
The text of the conversation between two people. (String) |
| label |
The emotion associated with each turn of dialogue (Happy, Sad, Angry or Others). (String) |
Acknowledgements
If you use this dataset in your research, please credit the original authors.
If you use this dataset in your research, please credit Huggingface Hub.