news-topic-understanding

Understand news articles' topics and entity names using neural network

How to run

1) Perceptron model

Run these commands in Terminal. They will apply Perceptron model to the Test data and output prediction to result folder.

• Newsgroup data:

  python perceptron.py newsgroups

• Propername data:

  python perceptron.py propernames

2) Maximum Entropy model

Run these commands in Terminal. They will apply Maximum Entropy model to the Test data and output prediction to result folder.

• Newsgroup data:

  python3 maximum_entropy.py newsgroups

• Propername data:

  python3 maximum_entropy.py propernames

3) Multi Layer Perception model

Run these commands in Terminal. They will apply Multi Layer Perceptron model to the Test data and output prediction to result folder. By default this will run for 1000 epochs.

• Newsgroup data:

  python multilayer_perceptron.py newsgroups

• Propername data:

  python multilayer_perceptron.py propernames