bert-base-uncased-triage

This model is a fine-tuned version of google-bert/bert-base-uncased for a 5-class triage classification task. It helps categorize student messages based on how they address specific learning objectives.

Classification Labels

0. ADDR_DIRECT: Message directly addresses the objective.
1. ADDR_PARTIAL: Message partially addresses the objective.
2. NOADDR_OFF: Message does not address the objective (Off-topic).
3. NOADDR_ON: Message does not address the objective (On-topic but irrelevant).
4. NOADDR_TANGENTIAL: Message is tangentially related.

Hyperparameters

{
    "learning_rate": 1.620896412491757e-05,
    "num_train_epochs": 4,
    "seed": 3,
    "per_device_train_batch_size": 4
}

Evaluation Results

The model was optimized for Macro-F1 Score on the test set to ensure balanced performance across unique objectives.

Classification Report (test set)

                   precision    recall  f1-score   support

      ADDR_DIRECT      0.949     0.771     0.851        96
     ADDR_PARTIAL      0.604     0.989     0.750        91
       NOADDR_OFF      0.968     0.732     0.833        82
        NOADDR_ON      0.933     0.922     0.927        90
NOADDR_TANGENTIAL      1.000     0.774     0.872        84

         accuracy                          0.840       443
        macro avg      0.891     0.838     0.847       443
     weighted avg      0.888     0.840     0.846       443

Confusion Matrix (test set)