Classifying News Articles with Different Models and Techniques

Models From:

• SciKit Learn
  • Random Forest Classifier
  • MLP Classifier
• HuggingFace
  • DeBERTA
  • MiniLM

Model Links:

• DeBERTA
• Mini LM

Model Accuracies:

These models were tested on 10,000 articles each.

	SciKit Learn Models	DeBERTA-v3	Mini LM
Top 3 Categories	64.35%	50.70%	75.42%
Top 4 Categories	65.54%	48.50%	73.10%
Top 5 Categories	37.00%	36.60%	71.07%

Setup:

• Install Python version 3.8 or above
• Install Conda using pip install conda
• Install dependencies from the environment.yml file using conda env create -f environment.yml. This creates an environment called text-ML.
• Activate environment using conda activate text-ML or by going into settings for your code editor and selecting text ml from there, if your code editor has that feature.
• Download the dataset from Kaggle
• Note that when running a program on one singular news article, it automatically uses the best model, Mini LM

Example of Program Use: