Disaster tweets classification is a kaggle competition. The aim is classifying if tweets are about disasters.
In the repository there are 2 notebooks:
-Disaster Tweets (Preprocessing): This notebook preprocesses the input data.
-Disaster Tweet Classification (ML): In this notebook I fine-tune a pre-trained BERT model from tensorflow-hub. Afterwards, I use it to predict the test samples.
NOTE: If Disaster Tweet Classification (ML) is not being displayed, You can check the colab notebook (exactly the same as the one that should be displayed here) below:
https://colab.research.google.com/drive/1q32mCpMSmK1ot1iz60Guaf0ltLwdyF7Z?usp=sharing
The data is from kaggle. In the notbooks there are links to the data.
The installed libraries can be founs in the requirements.txt pip install requirements.txt should install the required libraries. It is prefferable to run in a virtual environment.