WSA is scrapping the Web in order to give you its overall sentiment on a specific subject.
If needed you can install the required pip modules.
pip install -r requirements.txt
Also, in runtime.txt you have the python version used.
You can run wsa.py like this:
python wsa.py topic=cute_kitten
Or you can use the function inside like so (need to be in the same directory):
from wsa import wsa
sentiment = wsa("cute kitten", language="en", num=10)
The parameters language and num (number of webpages to scrap) are set by default to the values above.
The normalized sentiment will be around 0.
A positive result indicates a positive sentiment. A greater result means a more positive sentiment.
A negative result indicates a negative sentiment. A smaller result means a more negative sentiment.
In practice, you can take a 10 % margin to say the sentiment is neutral (when the result is over -0.1 and under 0.1).
Almost 7000 words are used to describe positivity and negativity.
The execution time can be significantly reduced by removing words. Of course, this will also impact the sentiment correctness.
Only english language is supported for now.
Some of the webpages will be skipped if they cannot be read.
The more specific you are in the topic the more relevant the sentiment will be.
Currently in alpha version 0.1.