sloria / TextBlob

Simple, Pythonic, text processing--Sentiment analysis, part-of-speech tagging, noun phrase extraction, translation, and more.
https://textblob.readthedocs.io/
MIT License
9.05k stars 1.13k forks source link

New classifier most_informative_features #385

Open neomukul opened 3 years ago

neomukul commented 3 years ago

Hello Team,

I trained a new classifier with some example as shown in link: https://textblob.readthedocs.io/en/dev/classifiers.html#classifiers But I am getting the most_informative_features as: fidelity = True as pu : pos = 51.1 : 1.0 % = True calls : pos = 46.0 : 1.0 123 = True spark : pos = 46.0 : 1.0 200 = True outla : pos = 46.0 : 1.0 41 = True calls : pos = 46.0 : 1.0 45 = True calls : pos = 46.0 : 1.0 51 = True unkno : pos = 46.0 : 1.0 75m = True launc : pos = 46.0 : 1.0 93 = True falli : pos = 46.0 : 1.0 about = True says : pos = 46.0 : 1.0 accelerator = True a rev : pos = 46.0 : 1.0 access = True lists : pos = 46.0 : 1.0 acquires = True sees : pos = 46.0 : 1.0

Question: I am not able to interpret the result, can you help me because every word is TRUE and POS and have a number but there is no explanation of it. tx