Inspiration

If we ever met an alien, our biggest wish would be to have a conversation! That's why we built a tool to converse in a way anyone can understand, regardless of language.

What it does

E.T. Translator takes a sentence, identifies the important words, and displays a series of images to convey universal meaning.

How we built it

Maud built the frontend using Java, and Finn built the backend using Python. We worked together to solve the hardest problems in both frontend and backend.

Challenges we ran into

  • Complex grammar structures (adjectives, verb tenses, singular and plural nouns).
  • Managing custom objects in frontend data-entry.
  • Downscaling image sizes to lower load-times.

Accomplishments that we're proud of

  • Complex GUI and original alien art by Maud!
  • Python CLI to interface with Java by Finn!

What we learned

  • How to insert HTML into a Java GUI.
  • How to analyze the grammar-structure of a sentence using NLTK (Natural Language Toolkit).
  • How to interface Java with Python through CLI calls.

What's next for E.T. Translator

  • Accounting for grammatical edge-cases (multiple adjectives in a row, homonyms, etc).
  • Upgrading image source to better represent the meanings of words.
  • Continuing to improve the GUI's data entry and display.

Built With

Share this project:

Updates