Tweet-Movie-Review-Analysis

Machine Learning Project to categorize tweets and movie reviews into sentiments and topics using small and large datasets

Description

This project focuses on teaching the machine knowledge about certain topics that people in the real world review/tweet about. Each tweet/review has a sentiment and topic attached to it, be it positive or negative (sentiments), about elections, oscars or Super Bowl (topics). This project focuses on categorizing tweets/reviews after being trained from a given dataset of such categories. This effectively sorts the tweets/reviews to cater to certain interest areas of different people around the world. I was inspired by the need for social media platforms to constantly categorize and learn from the interest areas of users to cater to their specific needs. Social media now can understand what we like based on the videos we view or the article we read. Furthermore, this project uses a Tries data structure to store such information, which allows fast and time efficient access, add, remove and search operations of each data.

EazyML usage

I heavily depended on EazyML to check whether my large test datasets had any bias in them or not. It is difficult to tell how reliable the datasets from the internet are because of how large they are. I also used EazyML to check whether my prediction score was comparable to the EazyML score. My EasyML prection score was 0.72 where as my own prediction score was 0.68

Built With

• python