This project is modification to the implementation shown in this article.
https://stackabuse.com/python-for-nlp-sentiment-analysis-with-scikit-learn/
Changes Applied:
Graphical reprenstation to study the details of various algorithms applied to the training data.
Adding Pickling concept to not go re-training on the data.
Using nltk libraries to churn more verbose data and fitering it more subtle manner.
Applying different sets of algortihms to study the trend of various behaviours.
Problem Statement:
This project mainly shows different airlines negative and positive reviews from the users commented on twitter handle in a single year. Comparative study is conducted to understand the inferences from the data. Used NLP to understand the sentiments of the tweets and making predictions on the unseen data, it could help airlines understand how many users are statisfied with their services and what improvement needs to be done for customer statisfaction.
Consists of Data Extraction:
Data Featuring:
Data Preprocessing:
Data Filtering:
Data Modelling:
Training the Classifiers:
Predicting:
Results in time taken and the accuracy gained in a graphical representation.