Skip to content

Opinion mining for provided data from various NLTK corpus to test/enhance the accuracy of the NaiveBayesClassifier model.

Notifications You must be signed in to change notification settings

rohitthapliyal2000/Sentiment-Analysis-NLTK

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

23 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Sentiment-Analysis-NLTK

Basic Algorithm:

NaiveBayesClassifier was used as the opinion classifier.

The dataset included a list of sentences.

The sentences was converted to form bag of words.

75% of the corpora was used as the Training set.

Rest 25% was used to test the Classifier module.

Modifying the algorithm:

  • Using first 'N' frequent words can increase the accuracy.

    Note : N should be changed along a certain range to check the peak point.

  • There can be sentences like :

    This is not good

    Bigrams (pair of words) can be used to include {'not', 'good'} in the bag of words.

    Using bigrams can exponentially increase the algorithmic accuracy.

  • For opinion mining, including emoticons can make the Classifier a lot better.

  • Pickle can be used to save the training model once and for all.

    The algorithm can be trained once and classifying the data will take no time henceforth.

Analysis of movie_reviews corpora:

Accuracy on modified algorithm: 72.8

Accuracy with algorithm including bigrams: 84.8

Analysis of twitter_samples corpora:

Accuracy with algorithm excluding emoticons: 76.9

Accuracy on modified algorithm: 97.35

Accuracy with algorithm including bigrams: 99.1

About

Opinion mining for provided data from various NLTK corpus to test/enhance the accuracy of the NaiveBayesClassifier model.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages