🚩 A collection of NLP Challenge Kernels competed on Kaggle
- Data Preprocessing
- Remove Unwanted data from the tweets including urls, user names incorrect words
- Remove Stopwords and tokenize using HuggingFace BertWordPiece Tokenizer
- Exploratory Analysis
- What a disaster tweet looks like ?
- Plot Disaster Tweets most common words by category
- Classify Tweets
- Use simple pre-trained glove embeddings
- Try BERT (using HuggingFace Implementation)
- Try Roberta (using HuggingFace Implementation)
- Tweet Classification models results summary