Multilingual-sentiment-covid19 Open-source Model - Free Deployment, Supports Multilingual Sentiment Classification of COVID-19 Tweets

Multilingual Sentiment Covid19

Developed by clampert

This is a sentiment analysis model trained on a large-scale multilingual dataset of COVID19-related tweets, supporting sentiment classification in multiple languages.

Text Classification

Transformers

OtherOpen Source License:Apache-2.0 #Multilingual Sentiment Analysis #Social Media Public Opinion #COVID19 Theme

Downloads 94

Release Time : 3/2/2022

Model Overview

This model is designed to analyze text sentiment, specifically optimized for tweets related to COVID19, capable of identifying both positive and negative sentiment tendencies.

Model Features

Multilingual Support

Supports sentiment analysis in multiple languages, including English, German, French, Thai, etc.

COVID19 Domain Optimization

Specifically trained and optimized for tweets related to COVID19 content

Large-scale Training Data

Trained on 18,437,530 multilingual tweet samples

Model Capabilities

Text Sentiment Classification

Multilingual Text Processing

Social Media Content Analysis

Use Cases

Social Media Analysis

Sentiment Trend Analysis on Pandemic-related Topics

Analyze changes in sentiment tendencies regarding COVID19 topics on social media

Can identify public emotional fluctuations towards the pandemic

Multilingual Public Opinion Monitoring

Monitor emotional reactions to specific topics across different language communities

Enable cross-language sentiment comparative analysis

🚀 Multi-lingual sentiment prediction trained from COVID19-related tweets

This project offers a multi - lingual sentiment prediction model trained on a large - scale dataset of COVID19 - related tweets, enabling sentiment analysis across multiple languages.

🚀 Quick Start

The model is trained on a large - scale (18437530 examples) dataset of multi - lingual tweets. These tweets were collected between March 2020 and November 2021 using Twitter’s Streaming API with various COVID19 - related keywords. Labels were auto - generated based on the presence of positive and negative emoticons. For dataset details, refer to our IEEE BigData 2021 publication.

The base model is [sentence - transformers/stsb - xlm - r - multilingual](https://huggingface.co/sentence - transformers/stsb - xlm - r - multilingual). It was fine - tuned for sequence classification with positive and negative labels for two epochs (48 hours on 8xP100 GPUs).

📦 Installation

No specific installation steps are provided in the original document.

💻 Usage Examples

Basic Usage

You can use the widget to test the model:

For English: Input text "I am very happy."
For German: Input text "Heute bin ich schlecht drauf."
For French: Input text "Quel cauchemard!"
For Thai: Input text "ฉันรักฤดูใบไม้ผลิ"

📚 Documentation

The model's repository can be found at [https://github.com/clampert/multilingual - sentiment - analysis/](https://github.com/clampert/multilingual - sentiment - analysis/).

Information Table

Property	Details
Pipeline Tag	text - classification
Language	multilingual
License	apache - 2.0
Tags	"sentiment - analysis", "multilingual"

🔧 Technical Details

The model uses a base model [sentence - transformers/stsb - xlm - r - multilingual](https://huggingface.co/sentence - transformers/stsb - xlm - r - multilingual) and is fine - tuned for sequence classification with positive and negative labels. It was trained for two epochs on 8xP100 GPUs for 48 hours.

📄 License

The model is released under the apache - 2.0 license.

📄 Citation

If you use our model in your work, please cite:

@inproceedings{lampert2021overcoming,
  title={Overcoming Rare - Language Discrimination in Multi - Lingual Sentiment Analysis},
  author={Jasmin Lampert and Christoph H. Lampert},
  booktitle={IEEE International Conference on Big Data (BigData)},
  year={2021},
  note={Special Session: Machine Learning on Big Data},
}

Enjoy!

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご