|
|
|
|
|
|
by rapid_snail
3993 days ago
|
|
|
I am currently experimenting with sentiment analysis using Python scikit-learn and NLTK. Working on a project that tries to predict if a user rated a business positively or negatively based on his review comment. I have access to only the 160 characters of the review (not the full text) along with the rating given by the user. With 10K samples of reviews and their ratings, I am currently at 70% accuracy in predicting if a review is positive or negative. |
|
|