Sentimenti

• 

  Screen at the start

• 

  Positive Sentiment

• 

  Neutral Sentiment

• 

  Negative Sentiment

Inspiration

I find it interesting to know what most people think about different things like companies, events and technologies, so I made my own app to find it all out!

What it does

Sentimenti app has a model that was trained on hundreds of tweets that contained people's sentiments. Model classifies each statement as positive, negative or neutral using Natural Language Processing. With Sentimenti, you can get to know what other people think about whatever you ask!

How we built it

I used Xcode as IDE for making my IOS-app, SwiftUI framework for building user interface for the app, NaturalLanguage and CoreML frameworks for building, training and testing my Machine Learning Model.

Challenges we ran into

1. I've spend a lot of time finding data sets and training the Model on this data set I chose. I had difficulties making the Model read data from .csv file and had to restructure the data set to make it appropriate to read.
2. I was confused on how to implement my model into the app and I ran into some errors in the process of integrating, but after a while I managed to figure it out.
   

Accomplishments that we're proud of

1. Trained my own Machine Learning Model with 70% accuracy of prediction.
2. Successfully integrated CoreML Model into my IOS-app.
3. Made my app working as I planned it.

What we learned

It was my first time working with CoreML and training my own model. I've learned a lot about Natural Language Processing because I was starting from scratch and knew nothing about building mobile apps using Machine Learning.

What's next for Sentimenti

Increasing the accuracy of predictions, building more complex UI and potentially making a multi-platform app.

Built With

• coreml
• swift
• swiftui
• xcode