Artificial Intelligence Project. Sentiment Analysis on tweets about Kashmir.

This was my AI project, in a team. We gathered tweets about Kashmir, in order to highlight how people felt about the terrority, and what their emotions were regarding it. We expected to see some animosity towards the subject because it has had such a controversial air to it. Still to this day people are conflicted from both Pakistan and India about this issue.

The dataset was obtained and then rated on a five point scale in order to distinguish the emotion that was felt with it by the twitter user. The data was then passed into a Naive Bayes classifier using a 70/30 split. and we obtained an accuracy which looking back on it in hindsight, probably wasn't the best ones you could obtain. It was a 33% accuracy, but we ran into some problems properly training and testing the data. We did not connect the nodes in the knime workflow correctly, unfortunately, because of a mistake. We did however manage to get a tagcloud which did showcase all the keywords associated with Kashmir.

You can further see our detailed work process in the pdf file present inside the folder. If you want to easily be able to see screenshots of our work and see a rundown of what we did, open up the report in the pdf format and you wont have to download it, the view will open up in your browser without hassle.