/sentiment_analysis

Primary LanguagePython

Sentiment analysis

Script analyzes given sentences & provide the calculated sentiment value, & then store in an excel file.
The initial data to be analyzed (sentences) are to be read from an excel file.

All input file names must be exact same as denoted here

File name to be read: ml - Copy.xlsx
Note: This file should be in exact name & format internally to get the correct result.

then the results are written to file: ml_scored.xls

Example files included

Working

  1. run settings.py from /support directory.
  2. download all the nltk packages by selecting 'all' from the window.
  3. train the models using any of the desired trainers named as trainer#.py in the root directory. -- Note: All the trained modles will be saved as byte files in 'trained' directory.
  4. run main.py .

Output will be written to a newly created file in the root directory as ml_scored.xls

Current test results

*Best results were gained after using trainer2.py for initial training.
(Use trainer sentiment_intensity.py from 'alternate' directory if neither of trainer#.py outputs desired results.)