Candidate features

We form a pool of candidate features by extracting keywords from the following web pages.
Influenza on Wikipedia
Swine inluenza on Wikipedia
Flu symptoms on NHS
Flu introduction on NHS
Flu treatment on NHS
Flu on BBC
MedicineNet's article about flu
Google Set for {flu}
Google Set for {flu, temperature, fever, cough and sore throat}

Tokenisation, name removal, stop-word removal and stemming (using Porter's algorithm) are parts of the preprocessing.

Here is the list with the 2675 candidate features we have extracted.

View the results of the experimental process