Dark Mode
Twitter Emotion Classification Data
Mental Health & Wellness
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
This Emotion Classification resource is designed to advance studies in natural language processing and textual emotion recognition. It contains text segments, mainly Twitter messages, each paired with a precise label indicating the underlying emotion. The data provides a rich foundation for exploring the nuanced emotional landscape of social media, supporting the development of models for emotion classification, sentiment analysis, and detailed text mining. The emotions are classified into six distinct categories: sadness, joy, love, anger, fear, and surprise.
Columns
- text: The text segment that provides the context, typically a Twitter message.
- label: A numerical classification that indicates the predominant emotion conveyed. The labels correspond to six categories: sadness (0), joy (1), love (2), anger (3), fear (4), and surprise (5).
Distribution
The primary data is found within the
emotions.csv file, usually presented in a standard tabular format. This resource totals 417,000 entries and occupies 42.11 MB. It consists of two columns, both of which are fully populated and exhibit 100% validity, with zero reported mismatched or missing entries. The text field contains approximately 394,000 unique values.Usage
- Sentiment Analysis: Uncovering the prevailing emotional sentiments expressed within English Twitter messages.
- Emotion Classification: Developing advanced machine learning models (such as LSTM) to accurately classify tweets into the six defined emotion categories.
- Textual Analysis: Exploring detailed linguistic patterns and specific expressions linked intrinsically to various emotional states.
- Multi-class Classification: Serving as a clear, labelled resource for training and experimenting with complex classification algorithms.
Coverage
The dataset is centred on text samples derived from the Twitter platform. The content is expressed primarily in the English language and is classified across six discrete emotional states (sadness, joy, love, anger, fear, and surprise). This dataset is static, with the expected update frequency being never.
License
Attribution 4.0 International (CC BY 4.0)
Who Can Use It
- Natural Language Processing (NLP) Specialists: Utilising the text for pre-processing and creating state-of-the-art emotion detection systems.
- Data Scientists and Machine Learning Engineers: For constructing and testing robust multi-class classification algorithms on social media data.
- Researchers in Mental Health and Social Science: Exploring the textual expression of defined emotional states on social platforms.
Dataset Name Suggestions
- Twitter Emotion Classification Data
- Six-Category Social Media Emotions
- Labelled NLP Emotion Corpus
- Text Emotion Mining Dataset
Attributes
Original Data Source: Twitter Emotion Classification Data
QUALITYLoading...