modified functionalities

[asiffarhankhan]

Change Log:

• Implemented nltk for training data
• removed urllib
• removed textblob
• added dynamic URL input
• changed exception to suit newspaper
• populated positive headlines in positive_headlines.csv and negative in negative_headlines.csv
• Program shows news author now

• If there are any clarifications that is needed to be explained regarding the working. please ask as few changes like use of classifier may not be apparent to everyone.

[asiffarhankhan]

ALSO I have added a file named test. It contains a function that can be used to train the classifier...suppose a user finds the sentiment to be unsatisfying he can re label the news into +ve/-ve so the headline along with the new sentiment gets added to the training data (positive_headline.csv/negative_headline.csv) to improve future results,

Although there are some drawbacks

• multiple data; if added with same values may reduce accuracy
• user may spoil the data by adding their biased opinions in data
• too many 'user added data' in either file may result unbalanced positive/negative training data

Due to these reasons the file is added separately and can be deleted in the next commit if not required