dblilienthal/Multiclass-Text-Classification-with-DistilBERT-on-COVID-19-Tweets
I implement a deep learning network to classify COVID-19 Tweets into 5 categories and 3 categories using DistilBERT (a lighter version of BERT) as an embedding layer along with an LSTM and Dense Layer. I Achieve 65% accuracy with 5 categories and 80% accuracy on 3 categories.
This tool helps social media analysts or public health researchers automatically sort COVID-19 related tweets. It takes raw tweet text and assigns it one of several sentiment labels, such as 'Positive,' 'Neutral,' or 'Negative.' This enables a rapid understanding of public sentiment without manually reading thousands of posts.
No commits in the last 6 months.
Use this if you need to quickly categorize large volumes of COVID-19 tweets by sentiment to gauge public opinion or track trends.
Not ideal if your primary goal is fine-grained analysis beyond general sentiment, such as identifying specific topics or factual inaccuracies within tweets.
Stars
13
Forks
4
Language
Jupyter Notebook
License
—
Category
Last pushed
Aug 20, 2021
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/nlp/dblilienthal/Multiclass-Text-Classification-with-DistilBERT-on-COVID-19-Tweets"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
raghakot/keras-text
Text Classification Library in Keras
shibing624/pytextclassifier
pytextclassifier is a toolkit for text classification....
celtics1863/envtext
中文环境领域文本分析包，纯神经网络架构，支持EnvBert,LSTM,RNN,word2vec等模型，支持自定义模型，下游任务包括分类，回归，多选，情感分析，命名实体识别等，专题包括气候变化文本...
Sshanu/Relation-Classification-using-Bidirectional-LSTM-Tree
TensorFlow Implementation of the paper "End-to-End Relation Extraction using LSTMs on Sequences...
Langboat/Mengzi
Mengzi Pretrained Models