DivineInspiration

Our application, Toneteller, uses state-of-the-art artificial intelligence to analyze text and accurately detect the tone of a message, including, happy, sad, angry, and more.

Comment

Inspiration

Have you ever been tired of trying to decipher the tone of an email or text message? It can sometimes be hard to know the emotion someone else on the other side of the screen is trying to convey to you. You might accidentally misinterpret their message leading to confusion and miscommunication. That is where Toneteller comes in!

What it does

Toneteller is a web application and chrome extension that uses machine learning and artificial intelligence to perform sentiment analysis on the text that a user has entered. Our application returns the emotion that the text is attempting to convey, giving the user more clarity regarding the intentions of the message.

How we built it

We built Toneteller using various tools and technologies:

To build the front end, we first developed the wireframes for the web application using Figma and implemented the web application using React.

To build the back end we used the Python library, Flask, to set up the web framework; Used the APIs and tools in the Python library transformers to train state-of-the-art pre-trained models, the models were then used to perform sentiment analysis on text; deployed the Python Flask server on Google Cloud Platform using Dockerfiles; implemented the Google Chrome extension using Chrome extension API, and pure Javascript, HTML, and CSS.

Challenges we ran into

During this hackathon, the greatest challenge all the team members encountered was inexperience. Two of the six team members had never down a hackathon before so the experience of being in their first hackathon was slightly overwhelming. Furthermore, these two team members had to learn React as well as how to build the front end of the application. Fortunately, they successfully built the front end of the application! Not only were two members inexperienced, but two members of the team were also isolated from the rest of their in-person teammates because they were remote. However, we overcame this challenge by being active on Discord and communicating often. All members of the team had to spend time learning more about the components of the project to which they had been assigned. For instance, one member was following tutorials on how to build the back end of the application and how to deal with servers. Overall, we are satisfied with how we have overcome all these challenges.

Accomplishments that we're proud of

Going into the hackathon, our team's goal was to learn more about how to make our idea come to life. Although we are proud of all our accomplishments, achieving this goal will be by far the accomplishment of which we are most proud. In the aforementioned challenges, we discussed how we overcame each and every challenge thrown our way so we are also extremely proud of that accomplishment. Of course, we are proud to have created the ToneTeller!

What we learned

We learned how to make Google Chrome extensions, how to use react and build a front-end application, more about servers and how to build a back-end application, flask, and how to do the analysis using machine learning, and optimization of the model. We also learned more about the topics covered in the workshops as well as more about job opportunities at the sponsor booths.

What's next for Toneteller

Toneteller has a bright future! We plan on refining the machine learning as well as getting more extensions into other apps like Slack and Discord. Also, Toneteller should be used as a tool! We would tell the world about it through advertising so that we could grow our user base. Hopefully, the new users would also spread the word about it so that we could grow exponentially.

Share this project:

Updates