Sentiment Analysis of New Sentences From Binary Classification

Project information

Summary: The aim of the project was to classify sentiments from the reviews of IMDB dataset, and evaluate the performance of classifier.

Tools: Jupyter Notebook, Python, Numpy, NLTK, Seaborn, Sklearn, and Pandas

Duties: It was an individual project. My role included performing feature engineering on the IMDB reviews dataset, classification of reviews and reporting of the findings. I performed detailed processing of the reviews such as removing punctuations, duplicates, and creating BoW representations using NLTK library.

Outcome: Achieved accuracy of 77% in predicting and classifying sentiments after performing detailed hyperparameter tuning.

Result: Obtained Distinction score for this project.