Project 42

While there are countless tools for correcting grammar, there are none for tone, sentiment, or repetition.

Comment

Inspiration

As students, we need help editing our essays. While there are countless tools for grammar, there are none for tone, repetition, and an overall summary. Also, our teachers have to spend too much time grading essays and reading through each individual section to find the overall tone and meaning of the essay.

What it does

In out website and iOS app, the user inputs in an essay and it will return a summary, analyzes the sentiment, agreement, irony, and subjectivity of the content. It also classifies it under a category such as sports, technology, humanities, etc as well as shows repeated words.

How we built it

For the website, we used JavaScript for the backend and making the MeaningCloud analytical API function. We also used HTML and CSS for the frontend to make the website look and feel appealing to the user.

For the app, we created it using Xcode and programmed it using Swift. We implemented the IBM Watson Natural Language Processing for the sentiment analysis.

Challenges we ran into

We ran into issues making the APIs properly function with our apps and we had to work around previous APIs since the API keys for those specific APIs did not work. Also, we had to spend a long time parsing the JSON files to make sure the information analyzed from the app was readable and presentable. We also could not get Hasura to work at all even after consulting different people and resources.

Accomplishments that we're proud of

We are proud of implementing the APIs and properly making them work as well as being able to properly parse the JSON files to make the information readily presentable.

What we learned

We learned more about JavaScript, Swift, and using the MeaningCloud and IBM Watson API.

What's next for Project 42

Make the iOS app able to display a wider variety of results which are more accurate. Also, we want to implement Havenondemand API so that we can add further features such as OCR and Theme classification.

Built With

Share this project:

Updates