sentiwordnet_ruby

Simple SentiWordNet 3.0 based tool for sentimental analysis. It is inspired from https://github.com/7compass/sentimental

How it works

Sentences are tokenized, tokens are assigned a numerical score provided by SentiWordNet 3.0. The total score is then used to determine the overall sentiment.

For example, If a sentence has a score of 0, it is deemed "neutral".

• Very positive scores are > 0.75
• Positive scores are > 0.25 & <= 0.75
• Weak positive scores are > 0 & <= 0.25
• Weak negative scores are < 0 & >= -0.25
• Negative scores are < -0.25 & >= -0.75
• Very negative scores are < -0.75

Installation

gem install sentiwordnet_ruby

Usage

# Load the default sentiment dictionaries
SentiWordNet.load_defaults

# And/or load your own SentiWordNet dictionaries
# Note:- Please remove the lines starting with '#' before loading a dictionary

SentiWordNet.load_senti_file('path/to/your/file.txt')



# Create an instance for usage:
analyzer = SentiWordNet.new
analyzer.get_sentiment "I love sentiwordnet"
#=> "very_positive"

analyzer.get_sentiment 'I like sentiwordnet'
#=> "positive"



# Get the numerical score of a string:
analyzer.get_score 'I love sentiwordnet'
#=> 0.9841496598639456

Credits

SentiWordNet lexical resource for opinion mining http://sentiwordnet.isti.cnr.it/

Gem functionality inspired from Seventh Compass, Inc.'s gem: https://github.com/7compass/sentimental