About the Project
1) The project simply classifies whether a provided text, and/or emoji, is positive, negative, or neutral using the Multinomial Naive Bayes algorithm. 2) Python libraries used are sklearn, nltk, and pandas 3) The dataset used is extracted from kaggle. So, you can train the machine using different datasets, and the code will remove attributes that are not relevant for sentiment classification. 4) As one can already guess, the project's accuracy is dependent on the data used to train it. From our test, this ranges from as low as 65% to as high as 83%.
To run the application:
1) Execute every block of code in the project.ipynb file. The last code block will launch the web application, which uses Flask. 2) Simply enter any text and/or emoticon in the input box and press the classify button to classify the text