Abstract : Sentiment analysis is the current popular method to know the business value in the market by analyzing the user feedback. Sentiment analysis helps the business to know the pulse of the customer and where to improve factors. Using Natural Language Processing(NLP) we can analyze the sentiments un- derlying the sentences whether positive or negative. Nowadays a lot of businesses are looking towards customer satisfaction factors and their opinions. Movie industry is one of them. In our project we have collected the reviews from the Internet Movie Database(IMDB) dataset from Kaggle to analyze the sentiments of the reviews. For text processing and modelling BoW(Bag of Words) and TFIDF vectorizer models are implemented and com- pared the results of both the methods. In our project BoW(Bag of Words) and TFIDF vectorizer models are implemented and to classify the given inputs we have used 2 classification algorithms SGD(Stochastic Gradient Descent) and Multinomial Naive Bayes compared the results of both the methods. The objective of the project is to classify the given review into positive or negative using Multinomial Naive Bayes and Stochastic Gradient Descent algorithms.The algorithms are implemented using sklearn library and the whole project is implemented using Python programming langauge. Both the algorithms achieved 80 percent accuracy and BoW model outperformed the TF-IDF model. Dataset is collected from kaggle open source repository.It contains 50k movie reviews each review is classified into positive or negative.IMDB is a most popular website for movie or celebrity content. And the project is deployed into the web application Python Flask frame work , HTML and CSS.
USER CREDENTIALS :
- SACHIT KUMAR TADISHETTY - 700734682 - sxt46820@ucmo.edu
- NAVEEN GOUD THANDARUPALLI - 700734683 - nxt46830@ucmo.edu
- PAVAN KUMAR NARAGA - 700733572 - pxn35720@ucmo.edu
- RAJSEH CHOWDARY KAKARLA - 700740728 - rxk07280@ucmo.edu