The Data folder includes 15594 pieces of training texts, a csv file for their labels and 5 unlabeled texts, provided by the Kaggle contest as the link below.
https://www.kaggle.com/c/feedback-prize-2021
Visualization.ipynb presents some basic features of the dataset, done by Yuchen Liu.
BERT_baseline.ipynb is finished by Yuchen Liu, providing fully functional word level solver of this problem. Since the Bert model limits input size to 512 tokens, the longer part of some texts are truncated.
LSTM_baseline.ipynb is finished by Yuchen Liu, providing fully functional sentence level solver of this problem. This model takes sentences as input, categorically classifying them without context. This notebook provides some simpy pre-processing and adjustable training parameters. LSTM.py is modified from it and run on his personal computer. Since the nature of RNN, PC is faster and than colab environment.
BERT_experiment_XinyuZhao.ipynb is a sentence level model trained by Xinyu Zhao.
longformer_ner.ipynb is a word level model trained by Ting-Po Huang.
seq2seq.py: implement LSTM with prior knowledge (not finished)