Made for python 3.X
assumes sentistrength is available in the current folder,
i.e.: the folder contains SentiStrengthCom.jar.
The language files should be in folders with 2-letter ISO
language codes.
- AR = Arabic
- CY = Welsh
- DE = German
- EL = Greek
- EN = English
- FA = Persian
- FR = French
- IL = Italian
- NL = Dutch
- PL = Polish
- PT = Portuguese
- RU = Russian
- SP = Spanish
- SW = Swedish
- TU = Turkish
Current language sets (except EN) drawn from https://github.com/felipebravom/StaticTwitterSent/tree/master/extra/SentiStrength
Sentistrength java file can be obtained from sentistrength.wlv.ac.uk/ by emailing Professor Thelwall (address on website)
Example use (single-core client):
>>> from senti_client import sentistrength
>>> senti = sentistrength('EN')
>>> res = senti.get_sentiment('I love using sentistrength!')
>>> print(res)
... {'negative': '-1', 'neutral': '1', 'positive': '4'}Example use (multi-core client):
>>> from senti_client import multisent
>>> ms = multisent('EN')
>>> texts = ['This is great!!'] * 10000
>>> res = ms.run_batch(texts) # push all texts and get the results in one go
>>> print(res[0])
... {'negative': '-1', 'neutral': '1', 'positive': '4'}
# OR, if you do not want all texts & results in memory
>>> res = ms.run_stream(texts) # pushes texts one-at-a-time for out-of-memory processing
>>> for i in res:
>>> print(i) # do what you want
... {'negative': '-1', 'neutral': '1', 'positive': '4'}Enjoy!