Disdain violations are sadly the same old thing in the public arena. Notwithstanding, web-based media and different methods for online correspondence have started assuming a bigger part in disdaining wrongdoings. For example, suspects in a few late disdain-related dread assaults had a broad online media history of disdain-related posts, proposing that web-based media adds to their radicalization. This needs to be prevented as a serious issue will arise in the behavior of the future generation. We desperately needed to prevent this, which led to the idea of SentiChecker.

What it does

SentiChecker is a web application that analyses tweets, finds out the sentiment in them, and provides a detailed report. It also identifies offensive words.

How we built it

We built it from scratch using HTML, CSS, Bootstrap, and JavaScript for the front end. Python and Flask for the backend. We used TwitterAPI for retrieving the tweets and Natural Language Processing (NLP) and Natural Language Tool Kit (NLTK) for sentiment Analysis.

Challenges we ran into

Pre-processing the data was a big challenge. Even though it could be done easily with poor results, we didn't want this to degrade the accuracy level of the results. So we took some extra measures to clean the tweet while keeping the context such as replacing emojis with their corresponding meanings.

Accomplishments that we're proud of

We are proud to have finished the project with such a high level of accuracy which turns out to be the reward for our hard work.

What we learned

Even though we learned a lot of technical stuff, we are more grateful to have learned something else. That humanity will thrive with goodness no matter how many weeds infect them.

What's next for EmoDetection

The next step for SentiChecker is the blocking of hate content and fixing the level of tolerance towards negativity which we think will benefit the whole digital world.

Share this project: